Report No. 25665-PK

Pakistan
Public Expenditure Management
Strategic Issues and Reform Agenda
(Volume I)

January 28, 2004



Poverty Reduction and Economic Management Sector Unit
South Asia Region




Document of the World Bank

                                         CURRENCYEQUIVALENTS

                                         CurrencyUnit =Pakistan Rupee
                                               US $1 =PKR 57.8

                                                FISCALYEAR
                                                  July 1-June30


                                        ACRONYMSAND ABBREVIATIONS

ADB    Asian DevelopmentBank                             MOF       Ministry of Finance
ADF    Asian DevelopmentFund                             MTBF      Medium-Term Budget Framework
AGP    Auditor General of Pakistan                       MTIP      Medium-Term Investment Plan
AGPR   Account General PakistanRevenues                 NDP        NationalDrainageProgram
CBR    CentralBoard of Revenue                          NEPRA      National ElectricPower Regulatory Authority
CCAs   Canal Command Areas                               NFC       National Finance Commission
CFAA   CountryFinancial AccountabilityAssessment        NHA        National Housing Authority
CGA    ControllerGeneral of Accounts                     NRB       National ReconstructionBureau
CWIQ   CoreWelfare Indicator Questionnaire              NSS        National SavingSchemes
DFID   Department for InternationalDevelopment(UK)       O&M       Operations and Maintenance
DPCO   DebtPolicy Coordination Office                    OFWM      On Farm WaterManagement
DSCs   Defense SavingsCertificates                       OGDC      Oil and Gas Development Corporation
EAD    Economic AffairsDivision (in MOF)                 PAAS      PakistanAudit and Accounts Service
ESR    Education SectorReform                            PACs      Public Accounts Committees
ESW    Economic & Sector Work                            PFC       ProvincialFinance Commission
FATA   FederallyAdministeredTribal Areas                 PIHS      Pakistan IntegratedHousehold Survey
FBS    Federal Bureau of Statistics                      PIA       Pakistan InternationalAirlines
FIP    FinancialImprovementPlan                          PIBs      Pakistan InvestmentBonds
                                                                   Project for Improvementin Fiscal Reportingand
FPSC   FederalPublic ServiceCommission                   PIFRA       Auditing
GOP    Government of Pakistan                           PR         Pakistan Railways
GST    General SalesTax                                 PSDP       Public SectorDevelopmentProgram
HIPC   HighlyIndebted Poor Countries                     PTC       Pakistan TelecommunicationsCorporation
ICG    Internal Cash Generation                          PV        Present Value
IDA    InternationalDevelopment Association              RICs      Regular IncomeCertificates
IMF    International MonetaryFund                        SAP       Social Action Program
IPPS   Independent Power Producers                       SBP       StateBank of Pakistan
I-PRSP InterimPovertyReduction StrategyPaper             SMCs      SchoolManagementCommittees
IRSA   IndusRiver SystemAuthority                        SOEs      StateOwned Enterprises
KESC   Karachi ElectricitySupplyCompany                  s s c s   Special SavingsCertificates
LBOD   Left Bank Outfall Drain                          WAPDA      Water and Power DevelopmentAuthority
MDGs   Millennium DevelopmentGoals


                                   Vice President: Praful Patel, SARVP
                                CountryDirector: John W. Wall, SACPK
                                   SectorDirector: Sadiq Ahmed, SASPR
                                  SectorManager: Ijaz Nabi, SASPR
                                  TaskManagers: Paul Wade, SASPR; Hanid Mukhtar, SASPR
                                                   Parvez Hasan, Consultant

                                                    TABLE OFCONTENTS



EXECUTIVE SUMMARY ...........................................................................................................                                         i

INTRODUCTION......................................................................................................................                                xvii

CHAPTER1:EXPLORING THE DIMENSIONSOFTHE CRISISINPUBLICEXPENDITURES                                                                                          ...........1
 Introduction................................................................................................................................................ 1
 Historical Perspective ................................................................................................................................              2
 Restructuringand Refocusingan Extended Public Sector .......................................................................                                       11
 Key Gaps inPublic Goods .......................................................................................................................                    13
     Health ...............................................................................................................................................         13
     Education...........................................................................................................................................           14
     Irrigation ...........................................................................................................................................         14
     Rural Water Supply ...........................................................................................................................                 15
     Transport..................................         ....................................................................................................       15
                                                                                                                                                                     16
 The Strategic Agenda for Public Expenditure Management....................................................................
     Power.................................................................................................................................................
                                                                                                                                                                     16

CHAPTER2: ENSURINGFINANCIAL                                DISCIPLINE AND CREATINGFISCALSPACE....................                                                    18
 Introduction..............................................................................................................................................          18
 Tum Around inDebt Position..................................................................................................................                        18
 Baseline Scenario .....................................................................................................................................            23
     Slowing the Rate ofBorrowing ..........................................................................................................                        24
     IncreasingRevenues.............................                                                ............................................................    25
     Public Enterprise Losses and Other Conti                                                       ............................................................ 26
     Defense Spending ...........................................................................................                   ..............................  30
     The Cost ofBorrowing ...................................................................................                       ..............................  31
     DomesticDebt Costs.........................                 ............................................................................................ 32
     The Cost of External Borrowing........ ............................................................................................                             34
     Structural Reforms and Growth Revival ........................................................................................... 36
 Alternative Scenarios ...............................................................................................................................               39
     Low CaseScenario ............................................................................................................................                   40
     High Growth Scenario..............                   ...................................................................................................        42
 Conclusions..............................................................................................................................................           43

CHAPTER3: REORIENTING                          PUBLIC SECTOR PRIORITIES                            ........................................................ 45
 Introduction..............................................................................................................................................          45
 Fundingthe Poverty Reduction Strategy .................................................................................................                             46
 Remedying the Neglect o f Education ............................................................
 Improving Health and Population Outcomes..................................................
     Health Sector Reforms ...............................................................................
     The Way Fonuard....................................................................
 AcceleratingWater Resource Development............................................................................................                                  55
 Transforming the Power Sector ...............................................................................................................                       61
 Conclusion............................................................................................................................................... 64

CHAPTER4: IMPROVINGOFPUBLICEXPENDITURES  EFFECTIVENESS                                                                    ................................... 66
 Introduction.............................................................................................................................................. 66
 Responsiveness to Citizens ...................................................................................................................... 66
 Reachingthe Poor.................................................................................................................................... 66
 Improving BudgetaryProcesses...............................................................................................................                  67
     Recent Initiatives ...............................................................................................................................       68
     Medium-Term Budgetary Framework...............................................................................................                           68
 Re-invigorating the PlanningProcesses...................................................................................................                     73
     StrengtheningExpenditure and ProgramMonitoring.......................................................................                                    75
 Redefiningthe Role ofthe State .............................................................................................................. 77
     Reducing the Role of the Government in Producing and Providing Private Goods and Services....78
     FurtheringDevolution.......................................................................................................................              82
 Federal-Provincial Responsibilities ......................................................................................................... 87
 Inter-Government Financial Flows ..........................................................................................................                  87
 Improving Govemance and FightingCorruption.....................................................................................                              88
 Strengthening FinancialAccountability................................................................................................... 91

CHAPTER 5: STRENGTHENINGTHE CIVILSERVICE.................................................................                                                     94
 Key Issues ................................................................................................................................................. 94
 Structure. not size. i s the problem............................................................................................................             95
     Bottom heavy .....................................................................................................................................       97
     Rigid .................................................................................................................................................. 97
 Mismatch Between Skills and Rewards................................................................................................... 98
 Persistent Patronage ...............................................................................................................................         100
 A Pragmatic Approach...........................................................................................................................              101
     Seize the emerging consensus..........................................................................................................                   101
 Managing the Transition........................................................................................................................              102
     Address pay .....................................................................................................................................        102
     Completedevolution........................................................................................................................ 103

REFERENCES/BIBLIOGRAP                  ............................................................................................................... 106

Annex A: Debt ReductionandManagement Strategy
Annex B: Medium-TermFramework
Annex C: Simulations BasedonDynamic Benefit Incidence ofPublic EducationExpenditures inPakistan

STATISTICAL     ANNEXES

TABLES
Table 1.1: Non-Interest andNon-Defense Expenditure for Selected Countries.........................................                                 1
Table 1.2: Investmentas a percentage o f GDP...........................................................................................            3
Table 1.3: Fiscal Indicators as a percentage o f GDP..................................................................................             5
Table 1.4: Consolidated Federal and Provincial Total Expenditure in 1999/00 Prices ..............................                                  7
Table 1.5: Consolidated Federal and ProvincialTotal Expenditure ...........................................................                        7
Table 1.6: PlannedPrivatization o f State Enterprises, November 2002-June 2004 ..........................                                   :......12
Table 1.7: Pakistan inComparison with Regional Countries inHealth and Population Outcomes .........13
Table 1.8: Present Status o f Health & PopulationOutcomes and MDGTargets for Palustan.................13
Table 1.9: Public Expenditure on Educationfor Selected Counties (% o f GDP)....................................                                    14
Table 1.11:Additional Coverage Required for Rural Water Supply (% o fpopulation served) .................14
Table 1.10: Canal Water Diversion ...........................................................................................................
                                                                                                                                                   15
Table 2.1: Extemal and Public Debt Burden............................................................................................              19
Table 2.2:   Summary o f Key Macroeconomic Variables inBaseline Scenario. 1998/99-2006/07 ...........23
Table 2.3:   3ase Case Scenario: PrincipalBudget Aggregates..................................................................                     23
Table 2.4:   Losses o f State OwnedEnterprises (Rs.Million) ....................................................................                  30
Table 2.5:   ?ahstan:Real Borrowing Cost on Public Debt. FY1996-2002 ..............................................                               32
Table 2.6:   Summary o f Key MacroeconomicVariables inL o w Case. 1999/00-2006/07 ........................                                        41
Table 2.7:   Palustan: L o w Case Scenario. Principal Budget Aggregates...................................................                        42
Table 2.8:   Summary o f Fiscal Space Alternatives over the Medium Term..............................................                             43
Table 3.1: 1ESR Targets by Sub.Sector. 2001-2005 ..................................................................................               48
Table 3.2:  1Water Availability, 1960/61 - 1999/00 ................................................................................... 56
Table 3.3: 1Electricity Tariffs (WAPDA System), FY96-2003 ................................................................                         64
Table 3.4: 1[llustrative Allocation o f Additional Fiscal Space...................................................................                 65
Table 4.1: 1Functional Responsibilities o fDistricts, Tehsils and Union Councils ....................................                              84
Table 4.2: 1Key Governance Indicatorsin South Asia............................................................................... 91
Table 5.1: 3vilian wage bills inPalustan (current) by comparison with selected counties ...................97
            (
Table 5.2:   <L
             Sanctioned Posts by Province (selected years) ........................................................................                97
Table 5.3: 1[mpact ofthe January 2002 Pay Reforms................................................................................                  99
Table 5:4: 1Proportion o fmonetary allowances infederal civil service wage bill...................................                               102

FIGURES

Figure 1.1:    Trends inSelectedExpenditures in 1999/00 Prices ...............................................................                      6
Figure 1.2:    Trends inSelectedExpenditures as a percentage o fGDP......................................................                           6
Figure 1.3:    General Administration Expenditures in 1999100Prices .......................................................                         8
Figure 1.4:    Law & Order Expenditures in 1999/00 Prices........................................................................                   8
Figure 1.5:    EducationExpenditures in 1999/00 Prices............................................................................. 8
Figure 1.6:    Health and Population Expenditures in 1999/00 Prices .........................................................                       8
Figure 1.7:    Agriculture Expenditures in 1999100Prices...........................................................................                 8
Figure 1.8:    Irrigation Expenditures in 1999/00 Prices..............................................................................              8
Figure 1.9:    General Administration Expenditures as % o f GDP ..............................................................                      9
Figure 1.10: Law & Order Expenditures as % o f GDP...............................................................................                   9
Figure 1.11: EducationExpenditures as % o f GDP.................................................................................... 9
Figure 1.12: Health and Population Expenditures as % o f GDP ................................................................                       9
Figure 1.13: Agriculture Expenditures as % o f GDP..................................................................................                9
Figure 1.14: Irrigation Expenditures as % o f GDP..................................................................................... 9
Figure 1.15: Spending on Education (Per Facility)................................................................................... 10
Figure 3:1: Budgetary Expenditure on Education...................................................................................                  49
Figure 3.2:    Per capita Benefit Incidence of Primary and Secondary Education.....................................                                50

Figure 3.3: Distribution o f Public Expenditure for Primary and Secondary Education.............................                                         50
Figure 3.4: Benefit Incidence of Public Primary Education (Urban vs Rural) ..........................................                                    50
Figure 3.5: Benefit Incidence of Public Secondary School (Urban vs Rural)............................................                                    50
Figure 3.6: Public Sector Expenditures......................................................................................................             57
Figure 3.7: Public Sector Expenditures as % o f GDP................................................................................                      57
Figure 3.8: Public Sector Expenditures as % o f total PSDP ...................................................................... 57
Figure 3.9: Depletion o f ExistingStorage and PlannedAugmentation .....................................................                                  59
Figure 4.1: DecisionTree for Identifying Core Governmental Tasks........................................................                                 80
Figure 4.2: The 2-Step Logic of Decentralization.......................................................................................                  82
Figure 5.1: The Main Components o f Public Sector Employment inPalustan..........................................                                        96
Figure 5.2: Composition o f Public Employment.........................................................................................                   97

BOXES
Box 1.1: Privatization................................................................................................................................... 4
Box 2.1: Refining  the Debt Strategy .......................................................................................................... 21
Box 2.2: Determinants o f the Rate of Public Debt Accumulation .............................................................                             22
Box 2.3 : Contingent Liabilities .................................................................................................................. 28
Box 2.4: Beneficiaries o f National Saving Schemes ................................................................................. 33
Box 2.5: Growth Revival and Poverty Reduction: Private Investment and Public Spending...................38
Box 3.1: The Rich-Poor Gap and Public Expenditureon Education.........................................................                                   50
Box 4.1: Budgetary and Financial Management Reforms.........................................................................                             69
Box 4.2:                                                                                                                                                 70
Box 4.3: DFIDAssistance to Strengthenthe Medium-Term Budget Framework .....................................
          Steps inPreparation o f an MTBF ...............................................................................................
                                                                                                                                                         72
Box 4.4: Planning Processes: Sources o f Weaknesses ...............................................................................                      74
Box 4.5: Options for Public Financing of Private Delivery o f Education..................................................                                79
Box 4.6: Public-PrivatePartnership inTuberculosis Control ....................................................................                           81

Box 5.1: Monetary Allowances and inkindBenefits..............................................................................
Box 4.7: Corruption inPalustan ................................................................................................................ 90
                                                                                                                                                         100
Box 5.2: Incomplete Devolution.............................................................................................................. 105

                                      ACKNOWLEDGEMENTS

This report was prepared by a core team including Parvez Hasan (Team Leader), Hanid Mukhtar, Paul
Wade and Nicholas Manning. Shahid Javed Burluand MohsinKhan served as Peer Reviewers. The field
work for the report was done in three missions during the period October 2002 to February 2003. The
basic draft o f the report was completed in early May 2003 and shared with the government. Discussions
with high-level government officials took place inlateAugust 2003. The draft has beenrevisedinlight of
the government's comments and has been updated to reflect information available up to middle
September 2003, includingmacroeconomic data provided to the September IMF mission.

Volume I1onAccelerated Water Development was preparedby Usman Qamar and Walter Garvey. Sector
contributions from Tahseen Sayed Khanand Ambar Narayan (Education), Inaam ul Haq (Health), Rashid
Aziz (Power), and Navaid Qureshi (Transport) are gratefully acknowledged. The report also benefited
from advice given by Tariq Hussain on the civil service, water and education sections o f the report.
Shanza Khan provided valuable research assistance. Shaheen Malik helped with data problems and
prepared the Statistical Annex. Ms. Fritschel contributed to the editing o f this report. Thelma Rutledge
processed the report and was responsible for its production.

Duringthe preparation, the team worked closely withthe government authorities. Inthisregard, the team
acknowledges the active and constructive participation from the Planning Commission, Ministry o f
Finance, especially the Debt Office and the Budget Wing, and the Economic Affairs Division. The Bank
team's discussions with the key counterparts, Secretaries Finance, Planning, and Economic Affairs
division, Director General Debt Office, and Additional Secretary Budget, at the various stages o f the work
starting from the concept paper finalization were instrumental in clarifying the main issues and
developing the analysis and the conclusions presented in the report. The team also benefited from
discussions with the heads o fplanning and finance departments o fthe governments o f Punjab and Sindh.

 The team leader made a Power Point presentation in February 2003 on Ensuring Fiscal Discipline and
Creating Fiscal Space to the government's economic team, headed by the Federal Finance Minister Mr.
Shaukat Aziz and including Dr.Ishrat Hussain (Governor, State Bank o f Pakistan), DrShahid Amjad
Chaudhry (Deputy Chairman, Planning Commission) Mr. Navid Ahsan (Secretary Finance), Dr.
Mutawaklul Kazi (Secretary Planning), Dr. Waqar M. Khan (Secretary Economic Affairs Division), Dr.
Ashfaque Hasan Khan (Director General Debt Office), Javed S. Malik (Additional Secretary Budget,
Ministry of Finance) and Wajid Rana (Joint Secretary, Ministry o f Finance). The Bank team benefited
greatly from the discussion of the principal issues involved in creation o f additional fiscal space that
followed the presentation. The team's work on civil service reforms was very much helped by extended
discussion in a brainstorming session on the subject at the Establishment Division Those attending from
the government side included Mr. Javed Hasan Ali (Secretary Establishment Division), Mr.Tariq Saeed
Haroon (former Secretary Establishment Division), Mr. Tariq Sultan (Principal, Pakistan Administrative
Staff College) and Messrs Mueen Afzal and Saeed Qureshi (former Secretary Generals Finance).

Finally, the team thanks John Wall, Abid Hasan, Ijaz Nabi, and Sadiq Ahmed for continued support and
guidance throughout the course o f preparation o f this report.


                                        EXECUTIVESUMMARY

1.       Pakistan appears to be at an important economic turning point and there are good prospects o f
sharply reversing the poor economic and social performance witnessed for more than a decade. The
comprehensive and forcehl reformagenda, which the government has been pursuingduringthe last three
years, and i s reflected in its debt and poverty reduction strategies, has already improved the financial
position very considerably and has helped to revive growth. For the first time inmany years, there i s no
threat o f a financial crisis hanging over Pakistan's head. Looking ahead, the main challenges now are to
increase economic growth from the current annual rate o f about 5 percent to close to a sustainable growth
rate o f 6 percent or higher over the medium term, substantially reduce poverty incidence, and begin to
narrow Palustan's social gap, while maintaining macroeconomic stability.

2.       The achievement o f Palustan's economic and social goals would require a strong recovery in
private sector investment supported by further improvements in general investment climate, deepening
and widening o f the financial sector, better incentives for exports and direct foreign private investment to
promote both technological change and export orientation. Whereas in the past aggregate investment
levels have been constrained by resource availability, inparticular low national savings, this has recently
changed. Higher worker remittances, the sharp reduction in the government revenue deficit (before
grants), higher external grants, lower external interest payments, and the reversal o f capital flight have all
helped to push the national savings rate to the average of more than 19 percent of GDP inFYs 2002-03,
its highest level ever. A part o f the improvement--- compared to the national saving rate of 14 percent of
GDP during FYs 1999-2000 - is due to transient factors such as reversal o f capital flight and
exceptionally high level of external grants. Still, provided confidence in the currency can be maintained
and the improvement in public savings continues, there i s every reason to hope that much o f the
improvement in national savings rate will prove to be durable. Thus, at least from the point of view o f
resource availability, the stage i s set for sustained and strong increases in the level o f investment. The
more important issues that are likely to influence future private investment are the investor perceptions
about political stability, the new government's commitment to sustained economic reform and its ability
to bring about improvements in governance and delivery o f public goods and services. Non-economic
factors such as the regional security situation and the threat o f terrorism will also continue to have a
bearingon private investment trends.

3.       While private investment must provide the bulk o f the growth in investment, it must be
accompanied and supported by significant improvement in both the quality and level o f public spending
to overcome the key gaps in the delivery o f public goods, including infrastructure. The low level o f
human resources development inPalustan i s limitingproductivity growth and hindering competitiveness.
The possibilities o f increasing value added in traditional manufactured exports and developing new
service and technology related exports would improve with higher levels o f s h l l development. The high
economic costs o f port handling, road and rail transport, and electricity are also sources o f
discouragement to the private sector.

4.       Better and more effective delivery of public investment and public services in key areas
must, however, contend with the reality, at least in the medium term, o f a slow-growing resource
envelope for the public sector. Our indicative macroeconomic projections suggest, that even with a
moderate but steady growth in government revenues, the level o f public spending would not likely
increase from its present (and recent) level o f around 22 percent o f GDP during 2004-7 (compared with
25 percent inthe 1980s and the first halfof the 1990s), ifthe necessary further progress towards reduction
in public debt burden has also to be ensured. Fortunately, there is considerable scope for both more
effective use o f public resources and restructuringo fpublic spending.

5.       The economic and social reform program has been tackling the various constraints on progress
towards a viable and vibrant public sector in a comprehensive and bold fashion for the last four years.
There has been notable progress and distinct success in a number o f areas, such as reducing
macroeconomic imbalances and public debt burden, accelerating the efforts towards privatization, and re-
orienting public expenditures towards social services and poverty groups. A start has also been made in
improving governance and strengthening o f institutional capacity, though the reform measures are yet to
yield concrete results interms o f improvement indelivery o fpublic services.

6.       Against the above background, the objective o fthis report i s to assist the government inexploring
ways to enhance resource mobilization and to further intensify and strengthen public expenditure policies
and management. The report focuses on three aspects o f public expenditure management: the institutional
reforms that are needed to make more effective use o f public spending, economic and fiscal policies that
would make possible a steady and significant enlargement o f the fiscal space, defined as non-interest,
non-defense spending (also excluding public enterprise losses) as a percentage o f GDP, and the issues in
re-orientation o f public expenditure priorities to make optimal use o f the fiscal space that could become
available. The scope of the work and analyses was shared and discussed with the government counterparts
at different stages o f the task. Related work to strengthen public expenditure management and service
delivery capacity at the provincial and district levels i s also underway.

7.       The main finding o f the report i s that the next few years represent a period o f great opportunity
for Pakistan to accelerate its economic and social development. Efforts to strengthen governance, improve
delivery o f education and health services, and last but not least, make economic and social policies pro-
poor by making public institutions stronger and the civil service better paid and more accountable will be
central to the public sector playing its role in increasing growth and reducing poverty. The relaxation o f
the financial constraints inthe public sector can strongly reinforce these efforts. With the maintenance of
the momentum o f economic and fiscal reforms and effective implementation o f the priorities outlined in
the govemment's poverty reduction strategy, there could be a sharp increase not only inthe level but also
the quality o f social and economic development spending over the next five years, insharp contrast to the
stagnation in outlays and their declining effectiveness during the last decade. The creation o f additional
fiscal space, however, would depend critically on hrther success in raising the level o f government
revenues, reducing the cost o f government borrowing, limitingthe public enterprise losses, and continued
restraint on defense spending in relative terms. With policy slippages inthese areas, the additional fiscal
space could virtually disappear, making it difficult to attain and sustain high growth and meet social
goals. Even if the fiscal space i s created, it could be wasted either because o f unproductive expenditures
or lack o f progress towards more effective use o f otherwise highpriority spending. The challenge for the
new government is to build on the improvements in economic management o f the last few years and to
avoid mistakes that often led to very ineffective use o fpublic resources inthe past.

Challenges of Improved Public Sector Management

8.       That Pakistan under-performs other countries with similar per capita incomes in delivery o f basic
social goods i s well lmown. What i s perhaps less commonly realized i s that the gaps in public goods
delivery have been widening during the last decade especially in relation o f delivery o f services to the
poor. Furthermore, the inadequacy o f public investment in key areas is takmg a heavy toll in efficiency
losses for the economy and hindering private investment. The depth o f the crisis in public spending in
Palustan can also be gauged from the fact that consolidated public expenditures (excluding interest and
defense) averaged little over 10 percent o f GDP in the late 1990s compared to the average o f about 15
percent o f GDP in the 1980s. Even though non-interest, non-defense spending rose sharply in FYs 2002-
03, Pakistan's spending levels compare unfavorably with other developing countries including South
Asian counties




                                                       11

9.       While public development and social spending must recover further from the extremely depressed
levels inthe last decade and must play a significant part in sustaining highgrowth and reducing poverty,
the biggest economic and social gains must come fi-om improved efficiency inthe use ofpublic resources.
Improving the effectiveness o f spending not only through a better allocation o f resources and a more
careful choice o f priorities but also through better implementation, more efficient delivery o f services,
plugging o f leakages and waste is a key challenge. This requires, above all, restoring the capability and
the authority of public institutions especially those concerned with economic and financial management
and the delivery o f social services.

10.       The following will be critical for success in refocusing the public sector and increasing its
effectiveness:

         Improved governance and more effective delivery o f law and order, justice and social services
         through effective devolution not only from the provinces to the districts but also from the Federal
         government to the provinces. This would need to be accompanied by suitable financial transfer
         arrangements and strengthening o f the civil service through improved compensation and greater
    .    accountability.
         Creationo f fiscal space consistent with further debt burdenreduction
         Using the fiscal space to adequately finance the government's poverty reduction strategy,
         maintenance o f existing public assets and selective important long-tenn investments especially in
         water

EnsuringFinancialDisciplineand CreatingFiscal Space

11.       Creation o f adequate fiscal space i s a necessary but not sufficient condition for better public
sector management. But opening up o f the fiscal space, ifproperly utilized, can certainly increase the
room for maneuver for improving governance (through more adequate funding o f law and order and
better civil service salaries) and re-balancing spending (by increasing allocation for operation and
maintenance o finfrastructure and more adequate fundingo fnon-salary expenditures inthe social sectors.)
Pakistan i s fortunate that it would not be so deeply constrained financially as in the past decade in its
efforts to expand the level o f its public spending in economic and social fields. With continued reform on
the lines mentioned below, the non-interest, non-defense spending will rise, in the base case scenario, to
14.3 percent o f GDP by 2006-07 compared to only 10.9 percent in 2001-02 (and an estimated 12.0
percent in 2002-03) consistent with a significant further reduction in public debt burden. In the l o w
scenario, it increases only to 11.3 percent o f GDPby 2006-07 and the debt burdenreduction i s modest.

12.      The major elements that contribute to the increased fiscal space under the base scenario are the
reduction in interest payments (2.9 percent o f GDP), higher government revenues (1.3 percent o f GDP),
and reduction in enterprise losses (1.4 percent o f GDP. It i s noteworthy that significant progress towards
raising revenues and reducing interest payments as a percentage o f GDP was already made in 2002-03,
the first year o f the scenario. Inthe low scenario, the government revenue declines slightly as a percent
o f GDP over 2002-07, interest payments fall more moderately, and public enterprise losses remain high.
The interesting point is that even with higher fiscal deficits (including grants) o f 4.3 percent o f GDP in
2004-7, compared to the baseline scenario deficit o f 3.0 percent o f GDP, the fiscal space under the low
scenario i s only a fraction o f what could materialize under the continuedreform scenario.

13.      A scenario with more fiscal space than inthe base case is certainly possible. Achieving this will,
however, require three things: (i)even more determined effort to lower the real cost o f domestic
borrowing, (ii)total elimination o f public enterprise losses, and (iii)strong improvements in the
effectiveness o f government spending and institutional capacity to optimally utilize the additional fiscal
space on poverty and growth related spending.


                                                       ...
                                                       111

Reducing costs of Government Borrowing

14.      With the recent sharp reversal o f the trend towards real depreciation o f the rupee, increased flow
o f concessionary assistance, very favorable terms o f Paris Club rescheduling, and the government actions
to retire high cost external debt, the real cost o f external borrowing has come down sharply. Pakistan's
case for continued soft blend of external assistance i s strong because o f the high level o f public external
debt and the strength of the reform program. The challenge i s to preserve and deepen the reformprogram
to ensure continued flexible assistance from donors at low cost. If the momentum o f quick disbursing
concessionary assistance can be maintained, and the real exchange rate remains stable, it should be
possible to keep the real cost o f external borrowing to 1-2 percent per annum inthe near future.

                                                                        a
15.      The real cost o f domestic borrowing, however, remains high. The government has taken several
strong measures in recent years to address some o f the weaknesses o f NSS instruments. It eliminated the
access o f institutional investors to the schemes in early 2000. Recently it took commercial banks out
because o f abuses o f the system. It made, after M a y 1999, a number o f successive reductions in the
administratively set interest rates on new issues since and has during the last year or so linked the NSS
rates to the market benchmark Palustan Investment Bonds (PIBs), which have been revived as an
important instrument o f debt management. Still real long-term rates o f return on 10-year certificates for
widows and pensioners o f 6 percent (assuming inflation rate o f 4 percent per year) and on Defense Saving
Certificates (DSCs) o f 4.5 percent on what are risk free assets are high by international standards; the
reductions in nominal interest rates have been largely offset by a drop inthe rate o f inflation.. Ideally it
would be desirable to phase out the use o f NSS instruments and rather fund government deficits
domestically by selling PIBs. However, phasing out the NSS instruments would meet with substantial
public opposition and i s not practical partly because the government sees some o f the recently launched
schemes under NSS as a safety net to pensioners, widows and small savers. But the Government must
continue with progress on reforming the NSS while educating the public on the true opportunity cost and
problems o f the schemes. The successful launch o f the PIBs has not only set a market-based long-term
benchmark yield but has also helped the government to finance its budgetary requirements with stable
long-term debt. In any event, recourse to PIBs should be further enhanced, the on-tap feature o f NSS
instruments should be eliminated, and the margins over PIB yields should be phased out.

Revenue mobilization

16.      Pakistan's revenue mobilization has been low, bothby international standards and compared with
its potential. In FYs 2000 and 2001 total revenue collection was the equivalent o f 16.2 percent o f GDP;
tax revenue was only 13 percent of GDP and had actually stagnated for several years.' However, revenue
collections duringthe last two years increased sharply to 17.9 percent o f GDP in 2002-03 reflecting both
the success o f tax collection efforts and the revival of the economy. The structure o f taxation has, also
improved considerably in recent years, and the heavy dependence on foreign trade taxes has been
reduced. Still, revenue mobilization falls short o f that o f many developing countries at comparable
income levels in Afnca, Asia, and Latin America.2 In light o f the weaknesses in the current revenue
mobilization system, there i s considerable scope for enhancing collection and broadening the income tax
base. For instance, while there are only 1-1.5 million income tax filers, there are 3 million cell phone
subscribers, over 10 million electricity consumers, 3-4 million gas consumers and 2-3 million car owners.




'Note that Pakistanhas undertaken considerable trade liberalization inrecent years with an initial import tariff revenue loss.
*Countries at comparable income levels with higher revenue/GDP ratios include Bhutan, Botswana, Kenya, Maldives, Namibia,
Nicaragua, Sri Lanka, Vietnam, and Zambia. All of these countries had revenue mobilization in excess o f 20 percent o f GDP,
with the exception o f Sri Lanka and Zambia.


                                                              iv

17.     Inour view, it shouldnotbe difficult, inlight of current efforts, to increase revenuemobilization
further to 18.3 percent o f GDP by FY07, the assumption in the base case. But the suggested level o f
revenue mobilization will merely regain the level o f the early 1990s. Therefore, to underpin strong public
finances and to adequately fund poverty-oriented expenditures, determined efforts are necessary to
complete institutional and policy reform especially the fundamental restructuring o f CBR inthe next few
years. Inthe area o f tax policy, reforms should continue to aim at continuing to reduce the number o f tax
exemptions (for customs duties, income and withholding tax) and make sales taxation more effective.
Reforms are also needed for tax on income from financial and land assets. In particular the current
practice that allows a single person to hold assets under different names (and more easily evade taxation)
should be abolished. On the non-tax revenue side it will be necessary to seek greater cost recovery in
higher education, health (subject to considerations about access to basic services for the poor), and water
usage for irrigation. Sustained increase in revenue mobilization also requires reversal o f the declining
trend in provincial tax collection, which i s only 0.5 percent o f GDP. Tax authority powers o f the
provincial as well as local governments should be reviewed and incentives for sub-national level resource
mobilization need to be strengthened as a part o f larger discretionary resource transfers to the provincial
and local governments that appear necessary (see para. 45 below).

Public enterprise losses

18.      Though there has been progress duringthe last two years in eliminating the operating losses o f a
number o f public enterprises notably Pakistan Railways, PIA, and most state-owned banks, the losses in
the power sector are continuing at a highlevel. The WAPDA and KESC losses in2002-03 were close to 1
percent of GDP. The continued large losses o f public owned power entities i s the most troubling element
of Palustan's fiscal position and the only one that has not shown improvement inrecent years. While the
urgency o f need to privatize KESC and parts o f WAPDA has grown because o f continued losses and
inadequate resources for investment as well as the costs to other industries from unreliable power supply,
the progress towards privatization i s the slowest inthe power sector, though WAPDA has moved to create
separate corporate power entities.

19.      The power sector reforms are fundamental. WAPDA needs to decisively reduce line losses,
enforce bill collection, if needed by cutting o f f non-payers more aggressively, and ensure timely
implementation o f tariff adjustments determined by NEPRA. The govemment should help WAPDA in
resolving the issue o f nonpayment by FATA o f its utility bills since this seems to require a political
solution. The cost o f pursuingnoncommercial objectives as imposed by government mandate should be
fully financed by explicit grants in the fiscal budget. Fortunately, in the years beyond FY 2004, power
sector costs will be favorably influenced by the rising share o f hydro- electric generation due to
commissioningo f Ghazi Barodaproject and some reduction inIPPpayments.

Defense

20.      In the 1990s the real spending on defense increased little and as a percentage o f GDP, it has
declined steadily from the peak o f 6.9 percent in the second half o f the 1980s. Despite the spurt in
spending in 2001-02 due to regional tensions, defense spending remained at 4.7 percent o f GDP (4.1
percent excluding defense pensions which are shown separately since 2000-01).

21.      The government feels that in current circumstances assuming further reduction in defense
spending as a percentage o f GDP would not be realistic. While easing o f tensions on the Easternborder i s
very likely, the situation on the Western border as well as combating terrorism in the country will
continue to put pressure on the defense expenditures. The report, therefore, assumes that in the base case
scenario defense spending (excluding pensions) in FY04, would be as inthe federal budget approved by
the National Assembly and incoming years would remain at 4 percent o f GDP.



                                                      V

Reorientingpublic sector priorities

22.      Proper use o f the considerable fiscal space that can be created under the continued reform
scenario will be critical. The financing o f the poverty reduction strategy should be the first charge on the
additional fiscal resources that are likely to become available. But it should also be possible to increase
funding for maintenance and undertake selective new public investment in infrastructure. As the past
experience suggests, the availability o f the fiscal space alone i s no guarantee that it will be utilized for
highpriority spending. Indeedthe danger isthat it couldleadto arelaxationofthe financial discipline and
undertaking o f low priority spending. The only way to guard against this danger i s for the government to
exercise self-discipline and to strictly adhere to well laid out processes and procedures for approval o f
projects especially large projects. The low scenario will largely mean business as usual and will neither
enable reversing much o f the decline inthe quality o f public goods witnessed during the last decade nor
ensure sufficient progress towards Millennium Development Goals (MDGs), though governance
improvementscould still ensure better performance than inthe past.

Fundingthe povertystrategy

23.      The original indicated funding for poverty reduction (4.2 percent o f GDP in 2003-04) in the
Interim Poverty Reduction Strategy Paper (I-PRSP) appears to be totally inadequate in relation to
ambitious targets especially in education. On the one hand, there i s need to set more realistic but still
ambitious social targets consistent with MDGs, on the other hand, the financing needs o f meeting these
targets require more careful assessment. It i s also recommended, that priority poverty spending should be
defined more broadly to include law and order maintenance, access to justice, and rural electrification.
The governance problems resulting from inadequate spending on police and the justice system impact the
poor the most. Similarly, it i s the poor in the rural areas who most lack access to electricity; o f the
roughly 40 percent o f Pakistani households that do not have electric connections most live in the rural
areas and are poor.

24.      Using the broader definition o f poverty related spending suggested here, the actual priority
poverty spending in 2001-02 was about 4.5 percent o f GDP. Our very rough estimates suggest that this
spending should increase to about 7 percent o f GDP by 2006-07. Larger education and health spending
would probably take up half o f this increase in order to achieve goals set; the rest being accounted for
largely by higher allocations for law and order, provision o f justice and rural and urban infrastructure
including rural electrification. It i s encouraging that the recently completed PRSP (December 2003)
provides both for a broader definition of poverty spending suggested above and proposes to increase this
spending steadily to 6.8 percent o f GDP by 2007-08.

Remedyingthe neglect of education

25.      The lessons learnt from the experience during the last decade in which expansion o f primary
education spending resulted in only moderate benefits are already being reflected in approaches to
primary education. Less emphasis i s put on building new schools and more attention i s being given to
improvements inquality through hiring o fnew teachers, greater accountability o f teachers, better physical
facilities in schools and higher non-salary spending. Human Development Foundation set up by the
Government o f Palustan as an autonomous corporation i s emphasizing enrollment processes, non-formal
education especially for recent dropouts, and innovative approaches to capacity building.But it i s not at
all clear that programs, new initiatives, institutions and funding commitments are in place on a scale to
achieve the key targets inprimary education: 35-40 percent growth inprimary enrollments over the next
five  years, narrowing o f gender and geographical differences, significant improvements in quality,
reduction indrop out rates and expansion o f programs o fnon-formal education.




                                                      vi

26.      Four sets o f issues relating to the social sector appear especially important. First, devolution o f
responsibilities for education and health to the local government holds the greatest promises o f improved
effectiveness o f delivery of social services. But as yet there are a large number o f unresolved issues inthis
area. Second, though enrollment levels in Pakistan are among the lowest in the developing world and
gender disparities and rural-urban gaps are among the widest, there are few incentive programs that have
been used around the world to stimulate demand for education and encourage enrollments. Third, the
effective public-private sector partnerships that will tap into the strength o f the private sector have yet to
be developed. The private sector i s the most dynamic segment of education. It accounted for over 40
percent o f the growth in enrollments in the primary education in the last decade. The average teacher
salary inprivate school i s actually lower than in public schools but the quality o f teaching appears to be
better because o f much greater accountability. Last but not least, more detailed and realistic estimates o f
public funding requirements are needed not only for the primary education but also for the entire
education sector. The government plans for reversingthe decline inhigher education deserve highpriority
andneed to be adequately funded.

27.      Clearly increasing public funding alone will not help meet the urgent challenges inthe education
sector. But higher public spending i s a necessary if not sufficient condition for improving educational
attainments. Detailed work i s necessary to firm up the education funding requirements including for the
secondary and higher levels. Our tentative estimates suggest the need o f gradually raising public
education expenditures from 1.8 percent o f GDP in2001-02 to 2.5 percent in2006-07.

28.      Within the recommended overall increase ineducation expenditures, a more rational allocation o f
spending i s needed. This includes increasing resources for: non salary recurring expenditures for primary,
middle and secondary school improvements, especially textbooks at the primary level; provision o f
missingfacilities inexisting infrastructure, especially at the elementary and secondary level; provision o f
quality services such as teacher training; increasing resources for new infrastructure where there i s a need
at i.e., middle and secondary level of education; girls' incentiveprograms; and demand side interventions.

29.      The fundamental problems that the government has to resolve to get better outcomes ineducation
are institutional. Inthe absence of fundamental reform, additional funding by itself will achieve little.
Many o f these problems, accountability, responsibility, service orientation, and incentives that reward
people for doing their job well, are generic to the entire public sector, but they manifest themselves most
visibly and with greatest impact in the education sector because of the size and multitude o f actors that
constitute the system.

Accelerating Water Development

30.      Because o f great neglect of water investments in the recent past, a substantial acceleration in
overall public spending in the water sector appears justified. WAPDA and Planning Commission must be
commended for their initiatives in reviving water sector investments. There are, however, several
problems with the thrust of the present program that require careful consideration and review by the
Government o f Palustan. Many of the issues involved, the tackling o f which can yield a more optimal
and balanced expansion o f irrigation water related public spending, are discussed at length in Volume I1
        .
o f the Report. These can be summarized as under

        .. Need for definition o f a clearer vision which integrates long-term agriculture, irrigation water
           and hydro-electric development
           Excessive emphasis on extension o f irrigation facilities some o f whom may prove marginal and
           all o f which will yield economic benefits only after a number o fyears
           Inadequate attention to the urgent needs o f rehabilitation o f existing irrigation assets and on-
           farm development


                                                       vii

       ...  Inadequate attention to the political and technical issues which may stand inthe way o f an early
            resolution o f a muchneeded additional storage on Indus
            The need for harmonization o f ambitious water sector plans with overall resource availability
            and to avoid pre-emption o f urgent social investments
            Balancing institutional development needs with higher water sector investment including
            Federal- Provincial co-ordination, and greater investment inplanning and investigations.

31.      A balanced program (including priority activities) covered under the poverty reduction strategy
can help initiate the transformation o f the water and agnculture sector even inthe medium term. The near
trebling of the real spending in water over 2002-07 will take up 20-25 percent o f the additional fiscal
space that i s expected to be created under the base scenario and will sharply reverse the declining trend o f
water sector spending. Needless to say, such large spending carries risks and could waste resources ifit i s
not accompanied by much tighter control on program and project priorities, stronger implementation and
institutional arrangements than exist at present. A fundamental strengthening o f the sector planning,
research and development, project implementation environment, and monitoring mechanisms including
much greater inter-provincial and Federal-Provincial coordination i s also necessary to facilitate decision
making on the more difficult longer-termissues.

Transformingthe Power Sector

32.      While the Government has taken a number o f steps to promote private investment in the power
sector, the desired goal to ensure the development o f the sector without recourse to the GOP budget,
except for hydro-electric development and possibly transmission, i s no where in sight. The continuing
financial distress inKESC, notwithstanding substantial injection o f public monies to fund past losses, and
intemational conditions make it highly unlikely that KESC can be privatized in the near future. It also
appears somewhat unlikely that the newly corporatized distribution and generation companies under
WAPDA will actuallybe privatized on a timetable to fundpower sector investments during2003-07.

33.      Consequently, there i s a disconnect between large national power sector requirements and the
amounts that can realistically be expected from the development budget and own resources o f WAPDA.
Whereas the TYPP has notionally allocated Rs.90 billion for WAPDA mainly from external loans,
WAPDA's proposed program including Rural Electrification for 2003-07 i s Rs.261billion. So there will
be a large additional requirement, after providing for WAPDA's own resources, which will have to be
mobilized from the budget andor borrowing inthe market with government guarantee. KESC investment
needs will be over and above this. The extra requirement for the power sector investment could wipe out
one-third of the fiscal space that i s supposed to be available under the baseline scenario.

34.      While the system o f automatic adjustments in electricity tariffs to reflect costs must be further
strengthened and strictly enforced, there may be only limited scope for increasing real, as distinct from
nominal, electricity tariffs3. A detailed sector plan for the power sector for the period 2004-08 needs to be
developed on an urgent basis and harmonized with the development o f a rolling three-year plan and the
MTBF. In the absence o f a plan, the power sector could derail government's economic and social
priorities. Some of the elements o f the power sector review should include the following:
              ...  A realistic butpro-active time table for privatizationo f corporate entities inthe sector
                   Carefulreview o f the present plans o f WAPDA especially for generation
                   Reinforcedefforts to reduce technical losses including theft
                   Identification o f key actions to achieve cost cutting
                   Clear identification o fpower sector financial requirements under a reform scenario

  According to recent World Bank estimates, 10 -15 percent increase in nominal tariffs over the next few years will be
inescapable even with aggressive cost reduction and speedy progress towards reduction of theft.



                                                             V l l l
                                                              ...

EffectivenessofPublic Spending

35.     The effectiveness o f public expenditures has several dimensions; strategic approaches to public
interventions; soundness, transparency and accountability o f budgetary and planning processes; careful
screening o f projects, speedy and effective implementation o f programs and service delivery; and the
existence o f strong monitoring and evaluation systems. The presence o f functioning institutions and a
framework o f defined distribution o f responsibilities at various levels o f government are the cross-cutting
issues which influence all aspects o f effectiveness o f public spending. While the mechanics o f better and
more efficient public service delivery are important, a real test o f effective service delivery i s whether it
meets the expectations o f the citizens especially the poor. Pakistan's record in effectiveness o f public
spending has deteriorated for several reasons. The public sector became overextended, poor govemance
and declining strength o f public institutions led to large leakages from expenditures, strict discipline in
choice o f new large projects was not observed. Many o f these weaknesses are being set right.
Privatization i s being pursued strongly and public sector role in commercial activities i s being de-
emphasized. Devolution i s being stressed as away to move services closer to the people and to increase
accountability o f public officials. Fiduciary controls have been strengthened. But inmany cases, notably
privatization and devolution, reform implementation i s far from complete. In other areas, such as
improving budgetary and planning processes and strengthening o f civil service, a full reform agenda has
yet to be developed. Sustained improvements in the management o f public expenditures would require
not only successful implementation o f the measures inhand but also a clear vision o f the future changes
that are required. Some o f the key elements that would shape the effectiveness o f public spending are
discussed below.

Accelerating Privatization

36.      Successive recent governments since the early 1990s have already opted for a muchmore focused
role o f the state and much greater role for the private sector. A great deal o f progress in furthering
privatization has been made in the last three years. The need i s to speedily complete the privatization
plans that are underway and to extend the program inpower, transportation, and insurance.

Public-PrivatePartnerships

37.     InPakistanthepossibilityof inductingtheprivate andnon-govemment sector inservice delivery
throughpublic-privatepartnership has not been fully explored.

38.      This has become especially important because o f the emerging possibility o f larger fiscal space
but continuing capacity problems within the government. Education and health appear to be the most
promising sectors for collaborationpartly because national prioritiesjustify a rapid expenditure o f public
spending on them. Tertiary water management i s another area where efficiency o f water use can be
increasedand localresources mobilizediffarmer associations can be inducted inthe process.

Devolution

39.     The Government o f Palustan has during the last two years taken a major initiative to devolve
powers to local levels o f the government especially the district level. The basic idea i s to move the
responsibilities for service delivery closer to the people and to increase the accountability o f the officials.
While the devolution program provides the best opportunity in years for improving the structure of
governance and improved effectiveness o f public services, the system needs considerable fine-tuning and
improvement to achieve these objectives. Some o f the areas that require early attention o f the government
are:




                                                      ix

             .....Extendingadministrative and fiscal autonomy:
                  Extendingfacility-level oversight
                  Reducing institutional complexity
                  Improving capacity
                  Improvingfunding

40.      The funding issues are the most urgent and immediate. A significant portion o f the fiscal space
that i s emerging must be made available to the district governments in part through Provincial Finance
Commission Awards and partly through incentive mechanisms which would help the local governments
to prepare and implement their policies and plans in accordance with the national and provincial
priorities, without compromising the autonomy o f the local governments.

41.      As the system beds down, it will be important to harness the potential that it offers for
strengthening user control over local policy and over the supply o f local services. This will require that
government take a distinctly pragmatic approach to the development o f Citizen Community Boards,
amending their structure and their govemance arrangements as practical experience i s gained. More
strategically, devolution may require some adjustments in the new political environment. The provinces
might wish to pursue somewhat diverse approaches to the precise pattern and speed o f devolution. This
should not be discouraged by the Federal government as long as the essential spirit o f the devolutionary
processes i s preserved and there i s clarity about the functions o f government at various levels.

Federal- Provincial Responsibilities

42.      There i s significant scope for devolving additional functions from the federal to the provincial
governments. The simplest way to proceed on this front i s to have a serious review o f the actual federal-
provincial responsibilities in relation to the letter and spirit o f devolution especially o f those service
delivery functions that have been assumed by the Federal Government duringthe last three decades either
because o f financial constraints or lack o f technical and administrative capacity (actual or perceived) at
the provincial level.     Devolution o f some o f the functions presently carried out by the Federal
Government would reduce the provincial government concern o f becoming redundant in the wake o f
large devolution to the district governments, will lift the morale o f the provincial civil service and would
enable the federal govemment to focus more intently on its key strategic functions. In most o f the
concurrent list areas, the federal government should remain concemed with providing strategic direction,
setting national minimumstandards and providing financing to the provinces conditional on their compliance
with these standards.

43.      Over-centralization o f functions i s also reflected in the planning process, which still exhibits a
strong top-down format. To some extent it i s unavoidable because o f the central role o f foreign assistance.
Butefforts shouldbemade to eliminate those schemes that are regional innature from the FederalPSDP.

Inter-Government Financial Flows

44.      The role o f provinces in developing realistic budgetary frameworks and viable poverty reduction
strategies can hardly be over-emphasized. We have argued for much greater devolution o f responsibility
to the provinces in order to underpin and strengthen the devolution to the district level. But even with the
present rather ad hoc distribution o f responsibilities, it i s clear that it i s the provincial/ district level
spending on education, health, law and order, provision o f justice, irrigation, rural and urban water
supply, androads that will account for the bulko fthe increase inreal public spendinginthe near future.

45.      The share o f provincial/ district level public spending i s likely to rise from around 26 percent o f
the total public spending inFY03 to over 33 percent in FY07. Interms o f GDP, provincial/ district level



                                                         X

spending should increase from 5.5 percent to 7.5 percent over the period, if our assumptions about the
fiscal space and priorities materialize. This has major implications for allocation o f national financial
resources. At present, the sub-national govemments are facing a severe fiscal squeeze. Whereas the
National Finance Commission (NFC) has been seriously considering a significant increase inthe share o f
provinces inthe federal divisible poll from the present 37.5%, additional flow o f funds would be required
to achieve the targets o f the PRSP. Clearly, mechanisms will need to be developed to allow a large
portion o f the fiscal space created at the federal level to be passed down to the sub-national levels. This
will require two things: (1) new NFC and PFC arrangements which assign much greater formula-based
transfers to the lower levels o f govemments than at present and (2) puttinginplace a system o f financial
incentives and special transfers to the provincial and local govemments so that national priorities can be
protected consistent with decentralization and devolution. Such transfers can be linked to provincial
revenue efforts and demonstration o f capacity to effectively utilize resources.

Improvingthe Effectivenessof PlanningandBudgetaryProcedures

46.     The introduction o f WRSP and MTBF provide important opportunities for moving away from
the excessive focus on the development budgets and to a more comprehensive approach to public
expenditure management. But several planning and budget process issues will need to be revisited and
resolved before the MTBF can become a truly operational document. The relative responsibilities o f a
range of public sector entities inrelation to various elements o f a MTBF estimation o f resource envelope,
costing o f sector programs, choice o f inter sector priorities, adoption o f specific targets etc. need to be
looked at afresh and agreed upon because they are in a state o f flux. The arrangements for putting
together a MTBF will also need to be harmonized with the preparation and implementation o f the final
PRSP. These entities include the federal and provincial govemments, the sector ministries/ departments,
Planning Commissiodprovincial planning departments, and the Ministry o f Finance/ provincial finance
departments

47.     Given that Pakistan i s a federation, it i s critical for improved budgetary outcomes and enhanced
effectiveness o f public expenditure that the MTBF process i s "rolled out" to all the provinces (and
ultimately to all the districts).

48.     The introduction o f the new processes, i.e. MTBF and PRSP, also provide an important
opportunity for strengthening overall planning in Pakistan and improve policy co-ordination among the
key economic ministries/departments both at the center and in the provinces and between the federal
government and the provinces. This will not, however, happen automatically. The Planning Commission
and the planning departments should lead the effort to integrate recurrent and investment expenditures in
a coherent medium-term budget fkamework. The preoccupation o f planners with the development budget
i s hamperingthe pursuito fbalanced strategies and speedy social development.

Strengthening CivilService

49.     In Pakistan the strength of public institutions has been greatly eroded over time. Many inter-
related factors have contributed to this decline: increasing centralization and non-transparent decision
making, lack o f accountability, the dilution o f the role o f merit in selection and promotion of civil
servants, undue political interventions in normal functioning o f government departments and agencies,
declining technical and professional competence o f officials, low or declining compensation, and a
culture of risk aversion behavior. The broader issues o f accountability and over-centralization are now
being tackled as a part o f the reform process. Some steps have also been taken towards civil service
reform.




                                                       xi

50.      But clearly fundamental problems with the civil service, low compensation for the higher levels,
low level o fprofessional competence due to low investment intraining, still excessive weight to seniority,
and perverse incentives and risk averse behavior, have not disappeared. The government recognizes this
and i s examining a number of proposals, to further strengthen merit, to streamline cadres and to increase
both opportunities and incentives for greater professional development. These issues have acquired
urgency with the implementation o f devolution, which requires re-definition o f the roles and
responsibilities o f district, provincial and federal levels o f government.

51.      This report sets out the need for administrative reform, but argues against an over-ambitious or
over-elaborate set o fproposals. History suggests that such proposals will generate many opportunities for
lengthy reports, but will lead to little sustained progress. Instead, it argues for three steps. First, it
suggests that there i s an opportunity to move ahead rapidly on the creation o f a National Executive
Service that would greatly increase incentives both for improved performance and upgrading o f s h l l s
through learning and education. The two critical elements o f this reform, on which there i s seemingly a
substantial consensus, are the much greater emphasis on merit and performance for selection to top
grades, including a highbar for promotion beyond grade 18, and opportunities for higher education and
learning for civil servants especially early in their careers. Second, it suggests sharp pay increases (and
monetization o f many o f the large in lundbenefits) for higher levels o f civil servants so that the serious
decompression o f pay that has been going on for decades can be reversed. Finally, it notes that
completing administrative devolution, by transferring budgetary responsibility for salaries to the district
governments and by progressively passing appointment, promotion and transfer (APT) authorities to the
districts, will address a necessary, if not sufficient, condition for local governments to value
professionalism more than patronage.

52.      The opening of the fiscal space can provide opportunity for reversing the squeeze inreal salaries
and de-compressing the compensation differentials and further consolidating monetary allowances and in
hnd benefits into base pay. However, before a new compensation policy can be formulated the future
pension liability must be assessed and actions taken to avoid a crisis in the future. The initiation o f work
inthisregardinthe MinistryofFinance is very welcome.

53.      Provided future pension liability can be contained and civilian employment growth can be kept
down to say 1percent per annum, there should be room for increasing real salaries significantly over the
next five years. The challenge will be to use the fiscal space, on the one hand, for restoring compensation
closer to private sector for the higher grades, and on the other to adjust pay scales for groups for instance
primary school teachers, policemen, lady health workers who are particularly poorly paid. Flat rate
increases, that do not decompress salaries or provide motivation to important occupational groups, will
not serve the objectives o f the civil service reform.

54.      Similarly, expansion o f civilian government employment will be necessary but should be
extremely selective and driven by identified needs in light o f the goals set in the poverty reduction
strategy papers at various levels o f government. The de-compression in compensation that i s being
recommendedmay be politically difficult but i s urgently needed ifhigh-level managerialand professional
slulls are to be attracted to the public sector. A move towards more adequate compensation should be
accompanied by and make possible other reforms such as much greater selectivity inpromotion beyond a
certain level, strengthening of incentives for learning and slull upgrading, and selective induction o f
higherlevelprofessionals from outside.

Fiduciary controls

55.      The government has recently embarked on a comprehensive reform program to restructure its
financial management system. Although significant steps have been taken in this regard, the government


                                                        xii

must carry the program forward and involve the subnational governments, public corporations, and
autonomous bodies more intently in the reform process. The continuation, expansion, and deepening o f
the reform program can ensure substantial improvement in the government's financial management
system within the next three to five years. The CFAA-Phase-I1 has outlined (among other things) the
following key actions for improving the financial management system.

56.       These actions include (1) enhancing the transparency o f the budget by disclosing fixed-cost and
contingent liabilities and the transparency of the accountability systemby publishingannual and quarterly
financial reports o f all levels o f government; (2) developing provincial budget implementation systems
along the lines o f the federal government's Revised System o f Financial Management and Control and
malung the federal and provincial systems fully operational; (3) establishing a strong internal control
mechanism for each level o f government; (4) building the capacity o f the auditing and accounting
establishments and o f the PAC; and (5) renewing the government's commitment to implementation o f
PIFRA.

Conclusion

57.      With political stability, reduction in regional tensions, improved governance and a continued
forceful pursuit o f the economic and social reform agenda, Palustan has good prospects o f raising the
annual GDP growth to around 6 percent inthe next three to four years and possibly higher in later years
and sustaining it at a high level, consistent with progress in debt reduction and a more self reliant and
equitable pattern o f development than inthe past. With the decline inthe population growth rate expected
to continue, sustained per capita GDP growth rates in the later part o f this decade of about 4 percent per
annum certainly appear feasible. This would represent a dramatic contrast to the average annual per
capita increases o f little over 1 percent in the latter part o f the 1990s and would contribute strongly to
poverty alleviation and employment creation.

58.      A center- piece of the growth revival must be the recovery inthe fixed investment rate from 13-
14 percent o f GDP in recent years to around 17 percent by 2007 led mainly by the private sector. The
implied high growth in investment over the next few years would require further efforts to improve
investment climate both for the domestic and the foreign investors. Better public resource management
can assist the growth inprivate investment in several ways, by improving governance especially law and
order and access tojustice, by promoting slulls development, providing needed infrastructure, and last but
not least by strengthening bureaucracy and malung public institutionsmore responsive to citizens.

59.      With continued reform, the restructuring o fpublic expenditures away from interest payments and
public enterprise losses will create substantial fiscal space for increasing social and development
spending. This will provide a unique opportunity to fund Pakistan's poverty reduction strategy and at the
same time narrow the gaps in key public goods. Social spending targeted on the poor deserves special
priority because poor have seriously lagged behind in access to basic services. It should also be possible
to selectively increase infrastructure spending and also fund governance and civil service improvements
more adequately. However, without addressing the institutional issues that are at the heart o f the
effectiveness o f public spending, the opportunity afforded by likely creation o f fiscal space would be
wasted. Some o f the elements o f public expenditure management that can make possible more effective
public spending are: continued focus on accelerating privatization plans, making devolution work at the
district level, devolving more responsibilities and finances to the provinces, adoption o f medium-term
budgetary frameworks at all levels o f government, moving away from excessive focus on development
spending, developing public-private partnerships in delivery o f social services, and last but not least
strengthening the civil service by improving incentives, expanding opportunities for learning and more
transparent accountabilities.



                                                       ...
                                                      XI11

60.      Even though there i s a clear need for sharp expansion o f both private investment and non-
interest and non-defense public spending, it i s the qualitative aspects o f this expansion that will, in the
ultimate analysis, assure the productivity improvements vital for growth and poverty reduction. Even with
continued reform momentum relative levels o f total investment and non-interest and non-defense
spending do not fully recover to the level of the early 1990s. Higher growth and more positive social
outcomes must come mainly from the expected improvements in the quality and effectiveness o f public
spending and a private sector that rises to the challenge o f diversifying export and industrial structures,




                                                      xiv

             PUBLICEXPENDITUREMANAGEMENT:STRATEGICAGENDA

               Challenges & Objectives                               Key Recommendations


A. Reduction in public debt to sustainable levels,    1  Contain the overall fiscal deficit (including external
    i.e. limiting public debt to no more than 400        grants) to 3 percent o f GDP on average during
    percent o f government revenue by 2006107.           FY04-07.

A I . Increase government revenue (to over 18         .  Accelerate CBR reforms, including a complete
      percent o f GDP by 2006107, and to an average      reorganization over a three-year period (FYs 04-06)
      of 17.7 percent o f GDP per annum during           and fkndamental        changes in human resource
      2003107).                                          management.

A2. Reduce financial losses o f State Owned           1  Speedily complete the privatization plans that are
                                                      ...
       Enterprises.                                      underway and to develop a realistic but proactive
                                                         timetable for privatizationo f WAPDA and KESC.
                                                         Reinforce efforts to reduce technical losses in the
                                                         power sector.

A3. Reduce cost o fpublic borrowing.                     Reduce the margin o f NSS instruments over the PIB
                                                         yield.
                                                         Endthe on-tap feature o fNSS instruments.

                                   II.Reorienting PublicSectorPriorities
B. Improve social indicators and narrow the gender    1  Fund the poverty reduction strategy adequately by
                                                      .
    gap by improving the effectiveness o f social        raising priority poverty spending by 2% o f GDP to
    spending, and allocating more resources to the       about 7% by 2007.
    sector.                                              Substantially increase allocation for education and
                                                         health focusing on both quantitative and qualitative
                                                         improvements.
                                                         Devise incentive programs to stimulate demand for
                                                         education and encourage enrollments, especially
                                                         among the poor.
                                                         Develop public-private partnership in social sector,
                                                         especially inremote and underserved areas.

C. Enhance agricultural productivity through             Develop a clear vision, which integrates long-term
    accelerated development o f water sector.         .
                                                      1




                                                         development o f agriculture and power sectors with
                                                         investments inirrigation sectors.

                                                      1  Develop a water sector investment strategy which
                                                         strikes appropriate balance between expansion of
                                                         irrigation   facilities  and:  (i)development       of
                                                         additional water storage facilities; (ii) needs
                                                                                                   O&M
                                                         o f the system; (iii) investments inthe drainage sub-
                                                         sectors;  and (iv) investment in water saving
                                                         schemes, e.g. lining o f water courses and leveling
                                                         land under OFWMprojects o f the provinces.

                                                      1  Implement the institutional plans for raising the
                                                         effectiveness o f spending on irrigation, drainage and
                                                         water storage.




                                                   xv

            Challenges & Objectives                                 KeyRecommendations
I
11                       III.ImprovingEffectiveness of PublicExpenditure
  D.Improveoutcomes o fpublic spending at lower unit  1 Develop and adopt a rolling Medium-Term Budget
                                                        Framework (MTBF) at the federal level and in
                                                        provinces, based on sector strategies and hard
                                                        budget constraints and integration o f development
                                                        and recurrent expenditures.

                                                      1 Reinvigorating the planning processes by moving
                                                        away     from excessive    focus   on development
                                                        spending and integrating planning and monitoring
                                                        with the preparationo fMTBFs.

           govemance through finther devolution.      1 Complete devolution to district level by extending
                                                        adrmnistrative and fiscal autonomy,       improving
                                                        capacity, and expanding funding.

                                                      1 Undertake devolution from the federal to provincial
                                                        governments        with      appropriate     funding
                                                        arrangements.

                                                      1 Establish mechanism, including financial incentives
                                                        (e.g. matching grants), through which national and
                                                        provincial priorities could be internalized by the
                                                        local govements.
II                                I K Strengthenin,   the Civil Service

                                                      1 Move ahead rapidly on the creation o f a National
                                                        Executive Service that would greatly increase
                                                        incentives both for improved performance and
                                                        upgrading o f skills through learning and education.

                                                      1 Use the fiscal space, on the one hand, for restoring
                                                        compensation closer to private sector for the higher
                                                        grades, and on the other to adjust pay scales for
                                                        groups, for instance primary school teachers, lady
                                                        health workers who are particularly poorly paid.

                                                      1 Avoid flat rate increases that do not decompress
                                                         salaries  or provide motivation for       important
                                                        occupational groups should be avoided.

                                                      1 Give the FPSC (and PPSCs) an unambiguous
                                                        responsibility for oversight o f the recruitment
                                                        process at all grades (even if it is decided that it
                                                         should not itself undertake direct recruitment for
                                                        positions below BPS16).




                                                    xvi

                                                 INTRODUCTION

1.        Pakistan has been grappling with the serious problems o f slow growth, heavy debt, social gap and
widespread poverty for some time. These problems have their roots ineconomic and social policy as well
as institutional weaknesses. There was extraordinary and persistent reliance on external resources for
financing economic development, almost relentless growth inpublic spending until at least the mid-1980s
and the consequent persistent high levels of fiscal deficits leading to unsustainable levels o f public debt.
Though the reliance on public sector was very heavy, the strength o f public sector institutions was eroded
over time, reflecting both the growing weaknesses in the civil service and undue political interference in
such matters as choice of projects and pricing decisions o f public corporations. Even though public
spending was large, social sector investments were greatly neglected. Palustan's social gap i s well known.
Not only did social indicators not improve in line with the growth in GDP but also the gap between
Pakistan and comparable low-income countries has tended to widen (see Figure 1).

2.        As the debt burden rose and the financial constraints deepened, the investment rate showed a
steady decline reflecting not only political instability but also a lack of confidence in the currency. The
impact o f reduced investment on the economy was aggravated by poor governance and continued slides
inthe institutional capacity inthe government, which adversely affected the quality o fpublic and private
investment as well as the effectiveness o f public spending. GDP growth averaged only 3.3 percent per
annum inthe second halfo f the 1990s.

                                               Table 1.1:GrowthTrends


                        Indicator(%)                                AnnualAverage for
                                                  1980s   1990191-1994/95 1995/96-1999/00 2000/01-2002/03
            CompoundgrowthrateofrealGDP            6.5           4.9              3.3               3.7
            Inflation (periodaverage)              7.2           11.5             7.9               3.4




3.        The Musharraf Government, which assumed power in October 1999, has been giving high
priority to revival of economic growth, reduction o f poverty, and speedy human development. It has
pursued a comprehensive and forceful reform agenda aimed at restoring macroeconomic stability,
improving governance, accelerating investments in education and health, improving the climate for
private investment, and last but not least malung econqmic and social policies pro-poor. The Govemment
o f Palustan's (GOP) determination to restore financial stability i s reflected in its Debt Reduction and
Management Strategy adopted in March 2001.4 The fairly ambitious targets o f public and external debt
reduction are intended to be made a part of a Fiscal Responsibility Ordinance (approved by the Federal
Cabinet) and are to be monitored by a high level Debt Office. The GOP's development agenda and
reform program were detailed in the I n t e r i d Poverty Reduction Strategy Paper (VPRSP) issued in
November 2001. In December 2003, GOP issued the final PRSP, which elaborates and substantially
strengthens the poverty reduction strategy. A great deal o f success was already achieved inFYs 2001 and
2002 in attaining macroeconomic stability, reviving growth, building up foreign exchange reserves,
reversing the decline inpublic spending and increasing pro-poor expenditures.




  "Debt Reduction and Management Strategy", Govemment of Pakistan, Ministry of Finance, March 2001 (see Annex). This
strategy was approvedby the cabinet following the recommendationsof the Debt ReductionandManagement CommitteeReport
(hereto referredas theDebt Committeereport). The Committee was set up inJanuary 2000 and submitted its report inearly2001.

                                           Figure 1:TrendsinSocial Indicators

 `overtyHeadcountfell in Pakistanduringthe 1980sbut stopped
 allingin the 1990s, leaving45 millionpeoplebelow the povertyline.
                                                     @Pakistan                                                                               I3Pakistan
                                                     ElBangladesh       140                                                                    Bangladesh
    60                                               .India
                                                     ONepal                                                                                    India

 $
 h                                                                      120--

    50
                                                                        100--
 3
 ft
    40
 8                                                                      SO -.

 3
 E
 e 30                                                                   60 --
 G
 3
 g2o                                                                    40 -.

 4                                                                      20 -.
 s   10

                                                                         0 7
     0                                                                         Fi'in?avSchool     Proportionof girls in Infant Mortality Rate Matmal Momlit

              1985-1986          1995-1996          1999-2000                  E n r o l k t rate all Primary student?       @er 1,WO)          @R10,000)
                                                                                  loaeent)              h r e m t )


 Pakistan'soverallspendingon educationhasfallen, while that of
 comparatorcountrieshas risen.                                        FertilityRatiohas begunto fall in Pakistan- but it hasfallen
                                                                      muchfaster,earlier in comparatorcountriesand is amongthe
                                                                      highestinAsia.

                                                      Bangladesh          8.0                                                               ElPakistan
                                                                                                                                            BBangladesh
                                                                          7.0

                                                                          6.0
                                                                      E

                                                                      d   5.0

                                                                          4.0
                                                                      e
                                                                      8   3.0
                                                                      12.0

                                                                          1.o

                                                                          0.0

            1997         1998       1999      2000         2001                      1960            1970             1980            1990         2000



4.         Buildingon the analysis inthe Debt Committee Report, the WRSP and the World Bank's recent
economic work referred to above, the present report, which i s prepared in close collaboration with the
govemment o f Pahstan,' focuses on the central issues in public expenditure management that will be
critical not only for accelerating growth but also for malung growth more pro-poor. There can be little
doubt that export led growth by the private sector i s crucial for Pakistan. But private sector efforts must
be strongly supported by improved public sector management. The broader issues o f macro-economic
imbalances, political instability and poor governance have adversely affected the climate for private
sector investment especially in the manufacturing sector. However, it i s equally true that during the last
decade or so deeply constrained public sector spending for social and economic development, continuing


 The Review Team intends and hopes that the same spirit o f cooperation and collaborationbetween the World Bank and the
Govemmentwill continueduringthe disseminationof, andfollow-up work to, this report.


                                                                 xviii

lags in social and human development, poor allocation o f resources, and growing inefficiency in the use
o f capital and delivery o f public services have all compounded the trend towards reduction in growth and
a stagnation in poverty levels. A sharp turn around in Pakistan's economic and social performance
requires, above all, a strong recovery in private sector investment. However, it all also requires a
significant increase inpublic economic and social development spending, a better management o f scarce
public resources, and a resolutionof the problem o f large loss malungpublic enterprises.

5.        This is the first ina series ofreports dealing with public expenditure management issues planned
for 2003-05. It is highly selective in coverage.6 No attempt i s made to cover all major sectors or all
fiduciary issues in expenditure control and management. In a number of public expenditure areas, the
base of knowledge i s already good and critical policy actions have already been taken. The approach here
i s (1) to refine and synthesize the agreed-uponparameters o fpublic expenditure management as they have
emerged from the government's development policy agenda and considerable World Bank economic and
sector work, and (2) to focus on public expenditure issues o f strategy, planning, and implementation that
remain either understudied or unresolved.




  The World Bank has been involved--albeit somewhat unsuccessfullyin terms of its impact--in analyzing public expenditure
issues inPakistan.Some recent Public ExpenditureReportsinclude the following: "Pakistan ReformingPunjab's Public Finances
and Institutions," August 2001, World Bank; and "Pakistan Public ExpenditureReview: Reform Issues and Options," October
1998, World Bank.




                                                          xix


 CHAPTER 1: EXPLORING                    THE DIMENSIONS THE CRISIS INPUBLIC EXPENDITURES
                                                                     OF

Introduction

1.1     This chapter outlines the various dimensions of the crisis inpublic expenditures inPalustan: the
low and stagnant revenue base, the high burden o f public debt, the great squeeze on public resources
available for economic and social spending in the 1990s, the relative neglect o f social sectors, and the
ineffective use o f public spending resulting from both declining capacity o f public institutions and
governance problems. It traces the historical origins o f present-day problems and gives an overview o f the
key gaps inprovision o f public goods. Finally, it presents a strategic agenda o fpublic sector management,
which i s the focus o fthis report.

1.2     It is well h o w n that Pakistan under-performs other counties with similar per capita incomes in
delivery o f basic social goods. What i s perhaps less commonly realized i s that the gaps in public good
delivery have been widening during the last decade especially in relation o f delivery o f services to the
poor. Furthermore, the inadequacy o f public investment in key areas i s taking a heavy toll in efficiency
losses for the economy and hindering private investment. While a higher level o f social and economic
development spending i s no guarantee for growth revival and poverty reduction, it is highly unlikely that
these will occur without more spending. The depth o f the crisis inpublic spending inPalustan can also be
gauged from the fact that consolidatedpublic expenditures (excluding interest and defense) averaged little
over 10 percent o f GDP inlate 1990s compared to the average o f about 15 percent o f GDP inthe 1980s.
'Even though non- interest non-defense spending rose sharply in 2001-02 and 2002-03, Palustan's
spending levels still compare unfavorably with other developing countries including South Asian
countries (see Table 1.1).

                    Table 1.1:Non-Interestand Non-DefenseExpenditurefor SelectedCountries
                                                            (Percent of GDP)

                                                1990
      Argentina                                  9.0                            13.2                            12.7
      Bangladesh                                 15.1                           12.6                            11.8
       Chile                                     16.2                           17.4                            21.1
       Egypt                                    20.4                           24.3                             19.4
       India                                    21.0                            17.8                            19.8
       Indonesia                                 14.7                           12.1                            13.0
      Malaysia                                  20.5                            15.9                            17.5
      Nepal                                      n.a.                           14.3                            13.8
      Pakistan                                   14.5                           11.7                            10.3
      Philippines                                11.2                           12.9                            14.3
       Sri Lanka                                20.9                            19.3                            15.9
      Thailand                                   9.8                            13.2                            15.5
      Turkey                                     12.2                           17.3                            22.2

      Source: Staff estimates, Govemment Financial Statistics, Economic Survey and World Development Indicators, Selected Issues.
      Data for Indiahas been obtained for Govemment Authorities and interest expenditures have been calculated for both the central
      and the state governments.
      Note: Data inthe 2000 column for Egypt and Malaysia are for the year 1997,while those for Indonesia and Turkey are for the
      year 1999.




'This decline is somewhat exaggeratedbecause in the early 1990s, the definition of development spending was revised to
  exclude spendingfinanced fromown resources ofpublic corporations. The latter constitutesabout 1-2 percentof GDP.

1.3      Beside inadequate resources devoted towards social and economic development, weak
governance led to low efficiency o f public expenditure, which contributed significantly to the poor
development outcomes. Over time, the role o f the government has become increasingly out o f touch with
reality.   Rigid internal structures made it impossible for the government to change its role and
organization to meet challenges o f running a modem day government. Rather than focusing on
increasingly complex tasks o f efficient and effective management and regulation o f the economy, and
providing good governance, the government continued to devote a large part o f its efforts and resources to
direct provision o f private goods and services, an area where it had neither the advantage nor the need to
be involved.

1.4      The transition from the government's traditional functions to a modern government was badly
hampered by a continuous decline in the quality and authority o f the civil service. Falling real income,
coupled with the waning authority o f the civil servants in runningthe affairs o f the government, made it
difficult for the government to attract the best and the brightest into the civil service. The system o f
recruitment, placement and promotion, which got increasingly distanced from merit, caused further
damage to efficiency o f civil service. The problem was compounded by a system, which was unable to
provide incentives for improving the individual skills, productivity or performance. A sharp decline in
resources for training and office and technological support resulted inpoor s k l l development and loss o f
productivity o f the civil service (see Chapter 5).

Historical Perspective

1.5      The present inadequate levels o fpublic spending on economic and social services as well as other
key public services such as law and order and the lag in providing these services are rooted inhistorical
developments in Pakistan. Pakistan's public sector became seriously overextended in the 1970s and
1980s. Several aspects o f this overextension should be distinguished. First, at the most basic level the
growth in expenditures far outstripped the growth in revenues, thus leading to large fiscal deficit and
unsustainable levels of public debt. Second, the growth was especially rapid in nondevelopment
expenditures, including defense, and as a result the general government savings turned negative after the
1980s, a pattern that has only slowly begun to change. Third, the state intervention in commercial and
industrial activities served as a strong disincentive for the private sector until at least the early 1990s.
Many activities in which the private sector could provide an efficient service without draining public
resources are in the hands o f government departments or public corporations. Fourth, in the process the
government's ability, both financial and institutional, to deliver basic public services such as basic health
and education was greatly undermined.

1.6      The level, financing, pattern, and effectiveness o fpublic sector expenditures greatly influence the
level o f economic activity, employment, and the distribution o f income in any country. The public
sector's role i s especially crucial in the early stages o f development in developing countries inproviding
not only strategic direction.and macroeconomic stability, but also key social services and infrastructure. In
Pakistan the level, pattern, and mode o f financing o f public expenditures initially influenced development
outcomes--that is, growth trends, progress toward poverty reduction, and social progress--to an
extraordinary extent at least untilthe 1980s.




                                                        2

1.7       There i s a great deal o f nostalgia in                    Table 1.2: Investmentas a Percentageof GDP (%)
Palustan about the relatively high average                                     Public          Private        Total GrossFixed
growth rate o f nearly 6 percent per year during                             Investment     Investment           Investment
the thirty-year period from 1960 to 1990.                   1949-50              1.6             2.5                 4.1
There i s not, however, sufficient memory that              1954-55              3.9             3.2                 7.1
the high and sustained growth during this                   1959-60              7.3             4.2                 11.5

period was made possible largely by the                     1964-65              11.6            9.2                 20.8

dramatic increases in public investment that                1969-70              7.3             7.0                 14.3

started in the 1950s and continued well into                  1970's             10.3            5.6                 15.9
                                                              1980's             9.2             7.8                 17.0
the 1980s(see Table 1.2).                                     1990's             7.5             9.1                 16.6

1.8                                                         1999100              6.0             8.4                 14.4
          Public investment (including that o f             2000101              5.5             8.4                 13.9
public corporations) as a share o f GDP rose                2001102              4.8             8.4                 13.1
steadily from the low initial level o f 1.6                 2002103              4.5             8.6                 13.1
percent in 1949-50 to the peak o f 11.6 percent             fource: Economic Survey,StateBank of Pakistan and Hasan (1998).
in 1964-65. Following the 1965 war with
India, bothprivate andpublic investment fell drastically. Public investment recovered from the setback in
the second half o f 1960s and remained on average at the high level o f nearly 10 percent of GDP during
the 1970s and 1980s. These levels were, however, unsustainable. Public investment showed little real
growth inthe 1990sand hadby 2002-03 come down to 4.5 percent o f GDP.

1.9       Although public investment was quite large until the 1980s, not all public investment was well
directed. Inthe 1960s and the first half o f the 1970s the government made huge investments inthe water
sector (including storage) equal to 4-5 percent o f GDP. No doubt, these investments, related inpart to the
Indus Basin Treaty, helped to sustain agricultural output growth and electricity production well into the
1 9 8 0 ~Buteven inthe water sector, large programs ofpublic tube wells proved costly and have remained
          . ~
a drain on the operation and maintenance budgets o f the provincial irrigation departments. The
investments in water storage and irrigation system expansion have not been matched by adequate
maintenance o f existing assets or by better water management systems to improve the efficiency o f
irrigation water use. Moreover, little progress has been made toward solving the major drainage problems
inthe IndusBasin andbuildingup an institutionalinfrastructure to guide andimplementthe plans for use
of scarce water resources.

1.10      The major power utilities in the public sector, the Water and Power Development Authority
(WAPDA) and Karachi Electricity Supply Company (KESC), have become a major drag on public
finances. Indeed, as discussed later, the financial losses of public enterprises in the power sector are at
present the principal source of the government's contingency payments. The historical roots of WAPDA's
financial problems lie not only inthe inefficiencies o f its operations and theft o f electricity, but also inthe
government's reluctance in the 1970s and early 1980s to allow timely adjustment o f electricity tariffs.
Following the sharp rise inthe intemationalprice o f oil in 1973, it took the government nearly a decade to
complete the necessary energy price adjustments. Consequently, the demand for electricity remainedhigh
while WAPDA's ability to generate funds for investment was badly damaged. This lack o f financial
capacity in the face of huge load shedding after the mid-1980s led to Palustan's heavy reliance on
independentpower producers (IPPs), leadingto a new set o f problem^.^


* Accordingto  Hasan(1998: 154), total water andpower investments (including the IndusBasinReplacementWorks) duringthe
1960s exceededUS$2.5 billion andaccountedfor more than 50 percent of total public sector development spending.
  The costs of power from IPPs are high, especially in the mediumterm, and are fixed in foreign exchange.Inthe rushto involve
the private sector in power investments, the government did not carefully scrutinize the demand for power or sufficiently take
into account the impact of needed increasesin real electricity tariffs. Also, the slowdown of the economy was to some extent
unexpected,andthus lower electricity demandhas necessitatedbigpaymentsto IPPsfor underutilizedcapacity.

1.11    Public sector involvement also extended heavily into industry, banlung, state trading, and
education inthe 1970s. Between 1969-70 and 1976-77 public investment in industry(including the steel
millproject) alone increased tenfold inreal terms while private industrialinvestment actually declined by
50 percent. Even though the private investment climate improved inthe 1980s, the public sector presence
inindustry,trade, andbankingremainedheavy. Byandlargethepublic investments inindustry,trade, and
banks tumed out to be uneconomic. Losses inthe public enterprises have been a major factor responsible
fo he pushtoward privatization and a rethinkingo f the role o f the state indevelopment (see Box 1.1).


                                               Box 1.1: Privatization

     In 1991Pakistan launched a broad-based program o f privatization with several objectives: reducing the
     drain on government resources caused by losses o f state-owned enterprises, improving the efficiency of
     resource use, and providing greater incentives and opportunities for private sector investments because the
     high public debt burden had made public sector resources for development very scarce. Though the
     urgency of these objectives has grown over time, the actual progress in transferring assets and control to
     the private sector during the 1990s was slow. Although the sale o f shares (12 percent o f total equity,
     mostly to foreigners) o f Pakistan Telecommunication Corporation Limited (PTCL) when netted more than
     US$SOO million in foreign exchange and provided about 50 percent o fprivatization receipts, management
     and control remained inthe public sector. Then the government delayed and lost the prime market so that
     now very little interest or demand exists for telecom. The only significant progress in transferring both
     ownership and control to the private sector was the sale o ftwo public banks, one power station, and a large
     number o f cement plants.

     The preparation and pace o f privatization has picked up greatly as part o f the reform program o f the
     Musharraf government. A major public bank has been sold, and a considerable number o f large
     transactions inbanking and finance, oil and gas, and power are ready to be brought to the market inthe
     next twelve months during 2003-04. At the moment, however, a large part o f the banking system, as well
     as almost the entire oil and gas, telecommunications, and power sectors are in the public sector, and as
     discussed earlier the past andpresent losses o f power and the banking sectors remain a particular source of
     concern.    If current plans to privatize significant parts of the power and banking sectors are not
     successfilly executed, public corporations will remain a serious drain o n public finances. This situation
     will also limit the role the government can and should play in providing services that the private sector
     cannot easily provide, such as law and order, justice, basic education and health, infrastructure, irrigation,
     roads, and urban and rural water supply.


1.12    An important consequence of the heavy emphasis on physical infrastructure development and
public sector industry until relatively recently was the low priority accorded to the social sectors. The
neglect o f education spending, especially in the 1980s, in face o f a rapidly growing population, i s a root
cause o f the social lag with which Palustan i s now confronted. The low level of human and skill
development i s in turn constraining productivity growth. It i s also interesting to note that the
nationalization o f private educational institutions in March 1972, undertaken ostensibly to accelerate
educational development, had perverse effects on the growth in basic education and the quality of
education. Though primary education was not nationalized, spending on primary education was squeezed
because national resources allocated to education did not increase, the management capacity o f the
government was overextended, and the urban bias ineducation was intensified by the takeover o f all high
schools and colleges."

1.13    Another long-term consequence o f public expenditure management during the high-growth
decades was the steep rise inpublic debt. The high overall level o f public spending, at close to 25 percent
o f GDP inthe 1970sand the 1980s, was unsustainable.As revenue growth lagged, fiscal deficits remained


loSee Hasan(1998: 224-25).


                                                         4

large, averaging 7.5 percent o f GDP for the period 1980-2000 and showing no strong downward trend.
Consequently, public debt grew from 66.3 percent o f GDP in 1980 to 93.7 percent in 1990 and to more
than 100 percent in FY2001. As a percentage o f government revenues, public debt grew from about 400
to 670 percent over the period. This level i s extraordinarily high by international standards and also
substantially higher than commonly judged sustainable levels (60 percent o f GDP and 300 percent o f
fiscal revenue, respectively)."         Initially, a large part o fthe public sector deficit was financed with external
borrowing at relatively concessionary terms. Over time, however, the government's reliance on domestic
borrowing grew, with the result that domestic debt and borrowing costs became increasingly more
important. The problem was compounded by a somewhat premature liberalization o f interest rates in
early 1989 and moving all government borrowing to market-based instruments when the fiscal deficits
remained high. Nominal interest rate payments were the main driving force behind the continued high
deficits during the 1990s, as the primary deficit first declined and then disappeared inthe late 1990s. But
since a large part o f the deficits in the first half o f the 1990s was financed through money creation, the
resulting inflation offset the gains innominal interest rates. The real interest rate on domestic public debt
was actually slightly negative during 1990-96.12 Paradoxical though it may seem, the liberalization o f
interest rates did not, in the ultimate analysis, lead to a deepening o f the public debt burden. It did,
however, leadto much larger fiscal deficits and muchhigher rates o f inflation.

1.14     The high and rising interest payments have made fiscal adjustment difficult. In the past three
fiscal years (2001-03) there was an average primary surplus (excluding grants) o f 0.8 percent o f GDP, in
contrast to a primary deficit o fnearly 4 percent o f GDP inthe 1980s. Still the fiscal deficit remained high
(see Table 1.3). Realizing the severity o f the problem, the government has recently set up a Debt Policy
Coordination Office (DPCO) inMinistry o f Finance, with the main purpose o f establishing an improved
debt management system.                The DPCO i s formalized through the Fiscal Responsibility and Debt
Reduction Ordinance (2003), which has been approved by the Federal Cabinet and i s likely to be tabled
before the Parliament in near future.                 This Ordinance aims at enforcing and preserving the fiscal
discipline through time bound restrictions onrevenue deficit and pubic debt.

                                         Table 1.3: Fiscal Indicators as a %of GDP
                              Total            Total             Fiscal             Interest            Primary
                             Revenue        Expenditure      Dedcit/Surplus      Expenditure         Deficit'Surplus
       1981-85                 16.5             23.6              -7.1                 3.1                -4.1
       1985-90                 17.9             26.3              -8.4                 4.7                -3.7
       1990-95                 17.3             24.9              -1.7                 5.4                -2.3
       1995-00                 16.3             23.1              -6.8                 7.1                 0.4
       1999-00                 16.3             22.5              -6.6                 7.8                 1.2
       2000-01                 16.2             21.0              -5.2                 6.8                 1.6
       2001-02                 16.6             22.0              -5.2                 6.8                 1.5
       Source: EconomicSurvey, GovemmentAuthorities, JMF and Staffestimates.
       Notes: 1)Fiscaldeficit excludesforeigngrants
               2) Total Revenues donot include surplusof autonomousbodies and grants.
               3) IMF total expendituredata hasbeenused from 1999-00 onwards. Statistical Discrepancyhasbeentaken into
                    account to compute fiscal balance.
              4) Total Expenditurefor 2001/02 excludesone-timeexpendituresof Rs.52 billion on KESCrecapitalizationand
                    CBRbonds.Ifincluded, fiscal deficit would be 6.6%andprimary surplus0.1%.



1.15     Overall, public spending adjusted for inflation increased by 19 percent over 1991-2001. All o f
this increase, however, was accounted for by the rise in interest payments; excluding interest, real



   See the Debt CommitteeReport, page 12.
  Debt CommitteeReportTable 17, page 53.


                                                                  5

budgetary spending declined slightly. Both defense and non-defense spending (excluding interest)
changed little over 1991-2001, though both fell sharply as a percentage o f GDP (see Figures 1.1 and 1.2).


                                           Figure 1.1: Trends in SelectedExpendituresin 1999/00Prices



                   ..............................................................................................................................




         1 +Total    Expenditure +Total               Non-InterestExpenditure +Total                  Non-Interest Non-DefenseExpenditure           I

     I                                   Figure1.2: Trends in Selected Expendituresas a Percentageof GDP

      1       30  I                                                                                                                                 I
              25

       s
                      .........%.....




     I    1 +Total   Expenditure +Total               Expenditureexcl. Interest +Total                 Expenditureexcl. Interest & Defence         I

1.16        The burden o f stagnant public spending (excluding interest) has fallen especially heavily on
development spending. Development spending in real terms dropped 40 percent over 1991-2001, and its
share intotal public spending dropped fi-om 25 percent to 12.8 percent over the period.13 During2001-02
development spending recovered by 34 percent as public resource availability improved thanks to large
external grants. Still, public investment by general government (excluding investment by public
enterprises) was only over 2 percent o f GDP in gross terms.14 If one allows for depreciation o f
infrastructure like roads, irrigationworks, and buildings, the public capital stock has probably grown little
inrecent years.

1.17        The consequences of growing interest payments, stagnation inrevenues, and relatively inflexible
defense spending over the past decade are brought out inTables 1.4 and 1.5, which also show the trends
in the level and pattern of major categories of non-interest, non-defense public expenditures. These
figures, which include both current expenditures and development spending, give a better idea o f the real




l3 This decline is somewhat exaggeratedbecause o f adefinitional change indevelopment spending.
l4The Public Sector Development Program, which rose from 2.6 in FYOl to an average o f 3.3 percent of GDP in FYs 02-03,
  includes alot of current spending coveringthe new categoryofpro-poor expenditures.


                                                                             6

trends in spending and the sectoral shifts than the development spending figures alone. (See Figures 1.3-
1.14).

                Table 1.4: ConsolidatedFederaland ProvincialTotalExpenditurein 1999/00 Prices

                                                 1991192 1992193 1993194 1994195 1995196 1996197 1997198 1998199 1999100 2000101
Total Budgetary Expenditure                      645,448 644,014 625,132 636,203 717,987 654,854 684,455 660,132 709,118 703,740
Interest                                         125,287 145,565 153,853 138,529 173,684 184,466 213,426 219,079 245,078 221,781
Defense                                          152,091 161,554 153,763 162,772 144,225 142,443 129,911 126,725 150,390 148,418
GeneralAdministration                             38,963  40,963   40,548  50,958  65,318   54,786  42,002    46,606 46,584    55,758
Law & order                                       26,091  25,360   25,551  28,429  27,510   24,348  25,230    25,018 27,624    28,790
Education                                         50,420  50,365   51,673  54,883  55,944   49,820  53,380    50,755 54,002    53,965
Healthand Population                              15,450  15,487   16,526 18,714    19,242  16,966  17,390    17,519 20,026    18,373
Agriculture                                       13,244  11,939   11,446  11,194   10,997   8,717   8,676     8,608   8,520    8,528
Irrigation                                        21,833  24,529   23,458  27,764  28,441   18,247 19,339     17,320  17,611   16,481
Fuel andpower                                     38,874  26,037   30,950  30,480  25,363   21,153  22,775    15,226  15,875   18,631
Transportand Communication                        24,951  35,283   32,670  25,524  22,700   18,863  19,829    21,849 23,912    20,316
Others                                           138,243 106,933   84,694  86,958 144,562 115,046 132,498 111,427 99,497      112,699


Memorandum items:
Total Non-InterestExpenditure                    520,161 498,450 471,279 491,674 544,303 470,389 471,029 441,052 464,040 481,958
Total Non-interestNon-defenseExpenditure         368,069 336,896 317,516 334,902 400,078 327,945 341,117 314,328 313,650 333,540
bfwhich: Development                             183,418 140,745 116,915 118,317 124,940    99,990 113,205 103,242 90,681      87,4941
Sources: IMF, "SelectedIssues and Statistical Appendix" November 9,2002; PlanningCommission and FinanceAccounts of the federaland
provincial govemments.
Note: Expenditures for Irrigation, Fuel & Power and Transport & Communication sectors include expenditures financed by the corporations
(WAPDA, OGDC, PTCLandNHA) from the budgetaryloans of the federal govemment.




                                                            (Percent of GDP)
                                           1991192   1992193 1993194 1994195 1995196 1996197 1997198 1998199 1999100         2000101
 Total Budgetary Expenditure                26.7      26.2     24.5     23.7     25.5    23.1     23.5    21.9      22.5      21.8
 Interest                                    5.2      5.9       6.0      5.2     6.2      6.5     7.3      1.3      7.8       6.9
 Defense                                     6.3      6.6       6.0      6.1     5.1      5.0     4.5      4.2      4.8       4.6
 GeneralAdministration                        1.6      1.7      1.6      1.9     2.3      1.9     1.4      1.5      1.5       1.7
 Law & order                                  1.1      1.o      1.0      1.1     1.0      0.9     0.9      0.8      0.9       0.9
 Education                                   2.1      2.0       2.0      2.0     2.0      1.8     1.8      1.7      1.7       1.7
 Healthand Population                        0.6      0.6       0.6      0.7     0.7      0.6     0.6      0.6      0.6       0.6
 Agriculture                                 0.5      0.5       0.4      0.4     0.4      0.3     0.3      0.3      0.3       0.3
 Irrigation                                  0.9       1.o      0.9      1.0     1.0      0.6     0.7      0.6      0.6       0.5
 Fuel and power                               1.6      1.1      1.2      1.1     0.9      0.7     0.8      0.5      0.5       0.6
 Transport and Communication                  1.o      1.4      1.3      1.0     0.8      0.7     0.7      0.7      0.8       0.6
 Others                                      5.7      4.3       3.3      3.2     5.1      4.1     4.6      3.7      3.2       3.5


 memo items:
  TotalNon-InterestExpenditure              21.5      20.2     18.5      18.6    19.4     16.6    16.2    14.6      14.7      14.9
  Total Non-interestNon-defense
 Expenditure                                15.2      13.7     12.4      12.5    14.2     11.5    11.7    10.4      10.0      10.3
    ofwhich: Development                     7.6       5.7      4.6      4.4     4.4      3.5     3.9      3.4      2.9       2.7
 Sources: IMF, "SelectedIssues and StatisticalAppendix" November 9,2002; PlanningCommission and FinanceAccounts of the federal and




1.18       Though the overall level of non-interest, non-defense spending was essentially stagnant over
much of the past decade, the pattern of spending shifted somewhat. Public sector power spending


                                                                    7

declined as financing fiom WAE'DA's own resources was excluded and more reliance was placed on the
IPPs for thermal power generation. In contrast, there was real growth in spending on primary education
under the impetus o f the Social Action Program. But even inprimary and secondary education spending
did not increase as a percentage of GDP in the 199Os, and their shares remained at the relatively low
levels o f 0.8 and 0.5 percent, respectively, in 2000/01. Spending on college, university, and professional
education declinedby nearly 20 percent inreal terms and even more sharply as a percentage o f GDP: only
0.3 percent o f GDP was being spent on higher education, including colleges, in 2000/01 compared with
0.5 percent in 1990/91.

                                                                                                                ~~




               Figure1.3: GeneralAdministration                             Figure1.4:Law& Order Expendituresin
                                                                                        1999/00Prices
                                                                         31
        65                                                         I
                                                                         29
        60                                                          ~




      --55
      050
      z45                                                        'i;:
                                                                   I     27


      240                                                        12      21
        35
  '                                                                      19
        30
  II                                                                1    17
      225 4
          0                                                  I




             Figure 1.5: EducationExpendituresin 1999/00       1               Figure1.6:HealthandPopulation
                                Prices                                           Expendituresin1999/00Prices          I
    I     60                                                  I
            1-                                       ~     1




    lI o  50
       =
     1 5 4 5

          40




          14               1999/00Prices                                                    Prices
                                                                         31
          13
          12                                                             26
      -.-210
       g 11                                                   I      --,021
                                                                     f                                                I
                                                                     .-
                                                                     a
      $ 9                                                     I      216                                              ~
                                                                                                                      I
          8                                                              11
          7
                                                                                                                       ~



          6                                                               6                                            I




 iource: Economic Survey 2001/02 and Finance Accounts of thefederal andprovincial governments.

                                                                8

                                                                   I                                                        I
         Figure 1.9: GeneralAdministrationExpenditures                      Figure 1.10: Law & Order Expenditures as a
                         as a YOof GDP                                                       Yoof GDP
       2.4                                                               1.2
                                                                    ~


       2.2

       2.0

       1.8
  IS   1.6                                                           S 0 9

       1.4                                                               0 8


  I    1.0 J                                                 11          0'7
                                                                         0.6 i

          Figure 1.11:EducationExpendituresas a YOof                            Figure 1.12: Health and Population
                                                              ~

                              GDP                                                  Expenditures as a YOof GDP
      2.2  f                              I--                           0 75

                                                                        0 70


                                                                     s. 0 65

                                                                        0 60


      1.6                                                               0 55

      1.5  1
                                                                                          -      ~ p $ ; ; ~
  ~                                                                     0 50
                                                                              m
                                                                              2E 2ox   ~    q   q   b   m   w    ~  m   m     o -

                                                                                          g    ~    g    ~     2    2 m o t
                                                                                                                        ~     2  z : ~



          Figure 1.13:Agriculture Expenditures as a                                                                          '
                                                   O ho f                  Figure 1.14:Irrigation Expenditures as a % of
                              GDP                                                              GDP                           I
      0.60 ,                                                i          1.3
      0.55                                                             1.1 -       A
      0.50
      0.45                                                             0.9 -
     S0.40                                                                                               \
                                                                     S 0.7

                                                                           1                        1                    1
      025
      030 4'                                                           0.5

                                                                       0.3




 ource: Economic Survey 2001/02 andFinance Accounts of thefederal andprovincial governments.


1.19      Agriculture and irrigation spending went down sharply inreal terms, especially inthe second half
o f the 1990s. In 2000-01, agriculture and irrigation spending was 30 percent lower than in 1990-91 and
constituted only 0.8 percent o f GDP, compared with 1.6 percent a decade earlier. Spending on transport,
communication, highways, and buildings dropped by 40 percent or 3 percent o f GDP over the period. A
large part o f spending on the Lahore-Islamabad Motonvay was outside the budget as it was financed
directly from the supplier's credit. Though the Highway Authority was inno position to service any debt,



                                                               9

its borrowing was outside the budget and its spending and projects were excluded from normal business
processes. Nevertheless, the Motorway-related debt didincrease the total public debt.

1.20     Spending on general administration and law and order maintenance fared better inthe 1990s than
public expenditures on social and economic services. Indeed, the real expenditure on general
administration nearly doubled between 1990-91 and 1995-96, far outstripping growth in either GDP or
revenues. Clearly, financial discipline was lax in these years and low-priority spending increased. With
the financial constraints deepening, the expenditure on general administration was reduced sharply inlate
1990s in real terms. This was achieved mainly by delaying the civil service salary adjustments, which
were finally      made in 2001 and were fully reflected in the 2001-02 budget. The net result was that
spending on law and order and general administration grew broadly in line with GDP, and the relative
reduction in the non-interest, non-defense spending o f nearly 5 percentage points o f GDP (from 14.5
percent to 9.5 percent over 1991-2001) fell entirely on economic and social spending.

1.21     The negative impact o f the squeeze in social and development spending during the 1990s on
growth and poverty reduction was greatly aggravated by declining effectiveness in the use o f public
resources, both borrowed and domestic. First, many low-priority projects were funded; the most egregious
example was the large uneconomic investment in the Lahore- Islamabad Motorway to the neglect o f the
 iational, provincial, and rural network. The Saindak copper project and the Tamir-i-Watan program are
 3ther examples o f misallocated resources. Second, many large and important projects and programs,
 iotably the Left Bank Outfall Drain (LBOD), the National Drainage Program (NDP), the Chashma Right
 Bank Canal Project, and the Social Action Program, did not have the desired impact because o f poor
 lesign, poor feasibility, poor coordination, delays inimplementation, or leakages inthe system.

                                                Figure 1.15: Spending onEducation"

                                                              Secondary                                           Tertiary

     90                                           270                                           f; 7,000 7-                                     I

                                                                                                   6,000      ................................
     go                                                                                         & 5,000       ..............................
                                                                                              6 e 4,000       .............................
                                                                                                I3,000        ............................
                                                                                                            ~




          ...................................                                                 .-   2,000      ................................

          ...................................   v
                                                        ....................................  x    1,000      ................................
                                                2n150                                           g      o
                                                                                                             ~




  "Annualspendingperfacility


1.22     The attempt to improve the delivery o f social services under the much-heralded Social Action
Program was a particular disappointment both because there was no sustained increase in the level of
social spending and because poor governance led to tremendous leakage and waste o f resources. The
problems o f governance, related to personnel and resource management, were the critical constraint to
achieving the government's development goals. These problems include absenteeism, hiring of staff on
grounds other than merit, frequent transfers, and mal-distribution o f staff to the disadvantage o f rural
areas.

1.23     Inaddition, some perennialpublic expenditure issues were to some extent intensifiedinthe 1990s
because o f stagnant or declining overall levels o f economic and social spending. Because the




                                                                     10

development budget cuts were particularly severe, there was inadequate spending in relation to the
optimum phasingo fprojects, andresources were spread thinly, delaying the benefits.

1.24      Notwithstanding the slowdown of development spending, the imbalance between maintenance o f
existing assets and creation of new ones intensified. As recurrent expenditure budgets stagnated or
actually declined, a greater portion of the budget was preempted by establishment costs. The nonsalary
expenditures crucial for running education and health facilities and maintaining physical infrastructure
were badly squeezed.

1.25      For example, official data show that during the 1990s the real per facility non-development
expenditures declined for each level o f education, most drastically for the tertiary level. Accounting for the
periodic increases inthe salaries o f government employees duringthis period, it i s possible to infer an even
steeper decline in the non-salary component o f recurrent expenditures. This decline in quality-related
operation and maintenance expenditures has led to a marked deterioration in school infrastructure, and a
large number o f schools across the country are in a state o f disrepair and lack basic amenities. A survey o f
rural facilities in2001 showed that only one-third o f rural schools had an adequate building,only about half
had h i t u r e , less than one-quarter hadtextbooks, 52 percent had dnnkmgwater, 48 percenthadtoilets, less
than one-third had electricity, and only one-sixth had a separate toilet for girls.

1.26      Service delivery and sectoral reforms have also suffered owing to a lack o f continuity at key
leadership positions. This i s especially true for important service delivery departments, like education
and health, in the provinces. Lack o f control o f line departments over appointments and transfers can
result in frequent transfers o f civil servants. Anecdotal evidence from Sindh suggests that department
heads have an average tenure o fjust over a year-barely enough time to understand the issues o f the sector
and far too little to develop a policy to tackle these issues.

RestructuringandRefocusinganExtendedPublic Sector

1.27      The preceding discussion brings out two key points. Palustan's public sector became seriously
overextended in the 1970s and 1980s. Until the 1960s the massive expansion o f public sector activities
did not seriously hinder the development o f the private sector. In addition to providing normal public
services such as law and order, justice, health, and education, the public sector concentrated on mobilizing
large foreign aid and loan resources and expanding investments in physical infrastructure, especially
water and power. The state role in industry was confined to a few strategic projects. This situation
changed in the 1970s, when very large-scale expansion o f public investments in industry and trade
actually led to a. sharp decline in private investment in manufacturing. Even though private investment
recovered after the late 1970s, the inexorable growth o f public expenditures, notably defense, continued
in the 1980s: the level of public expenditures as proportion of GDP rose further from 23.5 percent in
1979-80 to a peak o f 26.7 percent in 1987-88. The efforts in the 1990s to restructure and refocus the
public sector toward its traditional functions-maintaining macroeconomic stability, provision o f law and
order, basic social services, and essential infrastructure--were not very successful owing to a number o f
factors, including a lack o f strong political will resulting in part from political instability, limited
institutionalcapacity, poor governance, and limited fiscal space.

1.28      The current economic and social reformprogram attempts to deal with the various constraints on
progress toward a viable and vibrant public sector in a comprehensive fashion. Visible progress has been
made in a number o f areas, such as reducing macroeconomic imbalances and containing the debt burden,
accelerating the efforts toward privatization, and reorienting public expenditures toward social services
and poverty groups. Butin other areas, notably governance and strengthening o f institutional capacity, the
reform measures have yet to yield concrete improvements inthe delivery o f public services. Much hinges
on the success o f the ambitious programo f devolution o f authority to elected representatives at the district



                                                        11

level. The following chapters of this report discuss the various challenges the government faces and some
hard choices it will have to make in dealing with difficult issues. Here it i s necessary only to highlight
four broad areas, which will be critical for success in refocusing the public sector and increasing its
effectiveness:
          ....Creation o f fiscal space for spending excluding interest, defense and public enterprise losses.
              The successful completion o f the present privatization program, including early progress
              especially on the relatively slow-moving power privatization plans.
              Improved delivery o f social services through devolution andprivate-public partnerships.
              Improved govemance and more effective delivery o f law and order andjustice.

1.29     The creation o f fiscal space for public spending in support o f growth revival and poverty will b"e
central to the better management o f the public sector. As discussed in the next chapter, if broad-based
reform efforts continue, prospects are good for expanding real spending in the critical public service
areas.

1.30     The     preparation     and    pace    o f       Table 1.6: PlannedPrivatizationof StateEnterprises
privatizationhas picked up greatly as a crucial                          November2002-June 2004
part o f the Reform program o f the Musharraf                                                     Sale Proceeds
Government. A considerable number o f large                                                          (Rs Million)
transactions in banlung and finance, oil and           November2002 -June 2003                           21.813
gas, and power are ready to be brought to the          NBP Shares Lot 2                                      767
market by the end o f FY04 (see Table 1.6).            Pak Saudi Fertilizers(lOYoShares)                     815
Still at the moment, large part o f the banlung        GOP Working Interest in6 Oil Fields                   540

system,    almost the entire oil and gas,              UBL                                                11,749
                                                       Ban1AI-Falah                                          620
telecommunications and power sectors are in            MCB Zares throughCDC                                  661
the public sector and as discussed earlier the         PakistanOil Fieldsshares throughCDC                4,699
past and present losses in the power sector            Attock Oil Refineryshares throughCDC                1,011
remain a particular source o f concern.                MutualRight-ICP Lot A Fund                             175
                                                       MutualRight -ICP Lot BFund                             303

1.31     The political will to narrow Pakistan's       AI-HaroonBuilding                                       83
                                                       FalatiesHotel                                         450
social gap appears strong. Nonetheless, the            Pak Arab Fertilizers                               6,000
sheer magnitude of the task means that large           July 2003 -June 2004                              93,992
financial      resources      and     exceptional      PSO (51% shares)                                  20,300
institutional arrangements will be needed to           PTCL (26% of GOP shares)                           34,800
achieve even such basic and longstanding               OGDCL (5% shares throughIPO)                       4,640
goals as universal elementary education.               NPCC (100% ofGOP shares)                               667
                                                       FESCO(56% ofGOP shares)                             2,900
                                                       JPC (1% of GOP shares)                              2,030
1.32     Finally, improved govemance remains           HBL                                                23,200
a prerequisite not only for reviving private           National ConstructionCompany Limited                   100
investment,    but also for improving the              LasbellaTextile Mills                                  140
effectiveness      of    public     expenditures,      Lyllpur Chemicals                                      150
especially in the delivery o f services to the         BolanTextile Mills                                     140
                                                       Pak AmericanFertilizer
poor.                                                                                                      4,500
                                                       RepublicMotors                                          45
                                                       AC Rohri                                               280
                                                       Others                                                 100
                                                       TOTAL                                             121,865




                                                       12

Key Gaps inPublic Goods

1.33     A brief review o f the present position with respect to the social sectors and key physical
infi-astructure i s helpful in explaining the magnitude o f the challenge the public sector faces inproviding
more and better public services.

1.34     It is well known that Pakistan underperforms other countries with similar per capita incomes in
delivery o f basic social goods. What i s perhaps less commonly realized i s that the gaps inpublic goods
delivery have been widening during the past decade, especially delivery o f services to the poor.
Furthermore, the inadequacy o f public investment in key areas i s taking a heavy toll in efficiency losses
for the economy.

Health

1.35     Palustan's population o f about 140 million i s growing by 2.4 percent per year. The high fertility
rate o f 4.8 children per woman puts an enormous burden on the health o f mothers, resulting in a high
maternal mortality rate o f 200 per 100,000 live births.With 80 percent o f women still delivering without
assistance from trained personnel, a low contraceptive rate o f 28 percent, poor nutritional status, and only
48 percent o f tetanus toxoid immunization o f pregnant women, one inabout three hundred women die in
childbirth.

1.36     No change in malnutrition                  Table 1.7: Pakistanin comparisonwith RegionalCountries
levels in children has taken place in                           in Health and PopulationOutcomes
the 1990s:        the    percentage   o f                               Infant        Under 5     Annual       Total
children     under     five who      are    Country      Expectancy
                                                            Life       Mortality     Mortality  Population    Fertility
                                                                          Rate          rate      Growth        rate
underweight ranges fi-om 33 to 45          Pakistan         63              85           110        2.4         4.8
percent; micronutrient deficiencies         Bangladesh      61              60           82          1.7        3.1
o f vitamin A and iron persist; and         Bhutan          62              62          n.a.        2.9         5.4
the immunization coverage rate i s         India            62              69           87          1.8        3.1

only 52 percent, with a large urban-       Nepal            58              73           104        2.4         4.3
                                            Sri Lanka       73              15           17          1.6        2.1
rural differential and a smaller Source:Pakistan CountryUpdate; World Bank, 2003.
"
gender differential. Moreover, the
infant mortality rate o f 85 per 1,000 live births i s 10 percentage points higher than the average for low-
income countries and 16percent higher than the average for SouthAsia (see Table 1.7).

1.37     Public       dissatisfaction     with       the    Table 1.8: Present Statusof Health& PopulationOutcomes
government's        outpatient   health   services    i s                     and MDGTargetsfor Pakistan
reflected in the large numbers o f people seelung                                                         Millennium
the help o f private practitioners. Eighty percent o f       Indicators                        Latest    Development
outpatient contacts are with private practitioners.                                                        Targets

Govemment regulation o f private practitioners i s                                              110           47
virtually nonexistent. Poor access to basic services        Infant Mortality Rate               85            40
such as safe water and sanitation, along with poor          Proportion o f fully immunized      50            80
nutritional status, play a significant role in the             children 12-23 months
                                                            Matemal Mortalitv Ratio             300           120
                                                                                                20            100

low compared with those o f other countries in the
region and amounted to 0.6 percent o f GDP in 2000-01. The low level o f resources contributes
significantly to the poor quality o f health care, as do various inefficiencies in the allocation o f resources
and waste attributable to weak governance. The performance o f the health sector must be improved
drastically if Palustani s to reachthe MillenniumDevelopment Goals (MDGs) by 2015 (see Table 1.8).


                                                         13

Education

1.38      Pakistan's education sector has performed poorly in the past, with enrollment levels among the
lowest inthe developing world, and gender disparities andrural-urban gaps among the widest.

1.39      Nationwide, net enrollment rates at the primary level in 2000-01 were about 53 percent, masking
considerable differences between rural and urban populations (67 percent versus 45 percent) and males
and females (57 percent versus 44 percent)." Although the gender gap narrowed slightly inthe 1990s, this
improvement was inpart due to the decline in male enrollment levels. Quality problems are widespread
and often arise from institutional deficiencies such as nonfunctional schools. Were it not for the expansion
o f private education, the current situation would be much worse. From 1981 to 1998, private school
enrollment grew by 937 percent, compared with population growth o f 57 percent.

1.40      According to the Higher Education Task Force report, Pakistan's university and college system i s
in an unreformed state. The country has a population o f 18 million inthe 17- to 23-year age bracket, of
which only 2.6 percent are enrolled in higher education institutions (1996 data), compared with 6.2
percent in India (1990 data) and 12.7 percent in Iran (1994 data). Public universities, which enroll about
85 percent o f university students, are not only resource deficient, but also plagued by inefficient
governance and politicization.

1.41      Literacy        levels,     reflecting       historical           Table 1.9: Public Expenditure on Education
investments ineducation, are similarly poor (see Table                            for SelectedCountries (% ofGDP)
1.9). In2000-01 literacy was estimated at 48 percent,
with male literacy at 62 percent and female literacy at                                                      Expenditureon Education

34 percent. Although adult literacy rates are similar
across SouthAsian countries today (with the exception                                                                   2.3
                                                                                                                         3.7
o f Sri Lanka, where they are considerably higher),                   srihnka                                            3.1
Palustan's lower school enrollment rates imply that it                Nepal                                              2.9
will lag behind its neighbors in improving literacy in \Averagefor SouthAsia                             I               3.1              I
the future.                                                           Average for low-income
                                                                        countries                                        3.3

                                                                      Source: Pakistan Country Update, World Bank, 2003.
Irrigation

1.42      h g a t i o n and drainage are crucial to Pakistan's                       Table 1.10: CanalWater Diversion

agriculture sector, with about 80 percent o f arable lands and                                   Canal Head
90 percent o f agricultural output (25 percent o f GDP)                           Year         Diversions MAF             Remarks

entirely dependent on irrigation. Despite the country's                         1949-50              67.0          Just after independence

impressive irrigation infrastructure, overuse, aging, and poor                  1959-60              85.0            Before Indus Basin

maintenance have resulted in extremely low efficiency o f                                                                   Treaty
                                                                                1967-68              95.0               Post Mangla
water delivery. Average delivery efficiency i s about 35-40                                                             After Tarbela
percent from the canal head to the root zone (primarily                         1975-76              101.0            Commissioning
because o f seepage). Other than the Chashma Right Bank                         1977-82              105.0         PostTarbela (average)
Canal, expansion o f irrigation in the past fifteen years has
been limited to small isolated schemes outside the Indus                        1990-91              109.8           Justbefore Water
                                                                                                                            Accord
Basin and increased use o f tubewells (see Table 1.10). The                     1991-01              102.6          Post Accord Average
improvements made in the irrigation system through                              2000-01               81.1           Inflow 99.9 MAF
investments over the years have been offset by deterioration                                                               (drought)
                                                                           Source: Accelerated Development of Water Resources and
                                                                           Irrigated Agriculture, World Bank, 1993.

  The net enrollment rate is from the Pakistan Integrated Household Survey (PIHS). The govemment states net enrollment as 66
percent. In either case, Pakistan compares poorly with other low-income countries. The rural-urban and male-female differential
data pertain to 1998-99.


                                                               14

o f the system because o f inadequate operation and maintenance (O&M budgets fall short o f requirements
by more than 40 percent, whereas salaries o fthe overstaffed irrigation department account for more than
50 percent o f the total O&M budget) and poor performance o f the public institutions handling the
irrigation and drainage system.

1.43     Unlike the Indus Basin Irrigation System (IBIS), the drainage network i s not continuous and
integrated. L o w water charges, particularly compared with the productivity o f irrigated agriculture,
created incentives for farmers to increase water flow through illegal pumping o f water, resulting in
inequitable distribution o f water and inflicting physical damage on the canal system. Inadequate drainage
facilities have led to resource degradation, threatening the sustainability o f irrigated agriculture.
Pakistan's irrigation system, which was designed as run-of-the-river to maximize crop area, has created
serious problems o f waterlogging and salinity in the command area owing to lack o f drainage facilities.
Waterlogging problems claim about a third o f gross cultivable land, with another 16 percent o f cultivable
area suffering from salinity. Depending on the degree o f waterlogging and salinity, crop yields are
reducedby 25-60 percent from their potential levels.




Table 1.11). Rural water supply coverage i s particularly                  Water Supply Water Supply Water Supply
low inthe large provinces o f Punjab and Sindh.                Province    Coverage     Objective    Objective
                                                                             (2000)       (2011)      (2025)
                                                               Punjab          27           40           75
Transport                                                      Sindh           10           40           75
                                                               NWFP            45           65           75
1.45     Transportation currently accounts for about 10        Baluchistan     55           65           75
percent o f GDP and 17.3 percent o f gross capital             FATA            63           75           75
formation. Palustan's transport sector i s inefficient and     FANA            16           75           75
costly and imposes an annual cost of Rs. 180 billion, or 5     CDA             55          100           100
                                                               AJK             32           75            75




1.46     The road sector dominates the transport sector, accounting for 90 percent o f passenger traffic and
95 percent o f freight. The network i s comparable in density (0.3 M s q u a r e lan) to other developing
countries inthe region. The road network, however, faces severe limitations inthe form o f poor quality, a
serious shortfall in tertiary roads needed for basic access in poor areas, investments in economically
unviable megaprojects, a huge unfinished portfolio o f investment works that has increased budget
rigidities, and one o f the worst safety records.

1.47     The road sector has been the mainrecipient o fpublic sector funding.Total public expenditures on
roads are about Rs. 20 billion per year, o f which 65 percent are spent on national roads. Road
maintenance expenditures have been about 20 percent o f the required amounts for the national and
provincial networks. In 2001, whereas Rs. 21 billion were spent on roads, assets worth about Rs. 16
billion were lost as a result o f deferred maintenance.



                                                      15

1.48       The rapid growth o f the road sector has in part been at the expense o f railways. Currently,
Palustan Railways (PR) accounts for only 5 percent o f freight traffic, compared with 73 percent in 1955-
60. Its share o f the passenger market has dropped from 42 to 9 percent in the same period. This shift is
primarily due to mismanagement and resultingpoor service.

1.49       PR i s one o f the largest public money-losing operations in the country, with recent (1997-2000)
annual deficits regularly hittingthe Rs. 6-7 billion mark. Since 2000 the new management has succeeded
in slowing the loss-malung trend. The reduction inthe railway deficit is not sustainable, however, inthe
face of conflicting fare structures, decaying rolling stock and infrastructure, and huge fixed costs coupled
with low employee productivity.

Power

1.50       Palustan's power sector faces a number o f challenges in contributing more effectively to the
country's development. Only about 60 percent of the population, mainly in urban and periurban areas
close to the transmission corridors, currently has access to electricity, whereas the bulk o f the rural
populationrelies mainly on traditional fuels and kerosene for meeting its cooking and lightingneeds.

The Strategic Agenda for PublicExpenditureManagement

1.51       This report focuses principally on three key dimensions of better public expenditure management
inPalustan.

1.52       First, it is paramount to continue financial discipline and reduce the overall size o f the public
sector deficit, including the sizable losses o fpublic enterprises. The modest progress made inreducingthe
government's fiscal deficit duringthe past few years has been underminedby the persistence o f highlevel
o f losses o f public enterprises, especially WAPDA and KESC. To reduce the unsustainable burden o f
public debt, the fiscal deficit, which has averaged 5.5 percent o f GDP (excluding grants) and 3.4 percent
(including grants) duringthe past three years, must be brought down further. Provision needs to be made
for the large and continuing public enterprise losses and unfunded contingent liabilities of the public
sector.16 A strong and successful government revenue mobilization effort, which will gradually raise the
ratio o f revenues from 17 percent o f GDP (FY02) to say 20 percent over the next decade, remains central
to restoring Palustan's fiscal health. But as the experience o f the past few years shows, the structural
weakness in the taxation structure (relatively heavy dependence on trade taxes) and the institutional
weaknesses in the tax collection machinery (especially on the income tax side) will continue to dampen
revenue growth for some time. Thus it will be prudent to assume, at best, only moderate growth in the
ratio of government revenues to GDP over the next five years.

1.53       Even on the assumption of a steady increase in the ratio o f government revenue to GDP, the
growth in overall public spending in real terms will be modest over the next few years because o f the
need to reduce the deficit further and to fund public enterprise losses and contingent liabilities. Indeed, in
the medium term overall public spending as a proportion o f GDP i s unlikely to increase fiom the level o f
22 percent witnessed inrecent years, even ifgrant assistance remains at a relatively highlevel.

1.54       Yet the revival o f growth and the reduction inpoverty, the goals articulated inthe InterimPoverty
Reduction Strategy Paper (WRSP), will require substantially higher infrastructure investments as well as
higher social spending. The dilemma i s this. With the relative size o f overall public expenditures


l6These figures exclude the one-time expenditures o f Rs. 52 billion on KESC recapitalization and Central Board of Revenue
bonds in FY 2002. Ifthese are included, the average deficit for the past three years excluding grants i s 6.1 percent and including
grants i s 4.6 percent.


                                                               16

remaining stagnant, more adequate support for the government's economic and social goals can be
achieved only through a restructuring of expenditures away from interest payments, defense, and
nonessential expenditures toward social and development outlays, adequate maintenance o f infrastructure,
and support for improved governance. What are the possibilities o f restructuring public expenditures
away fi-om the inflexible burden o f defense and interest payments, as well as public enterprise losses?
What are the policy and institutional actions that can facilitate a better allocation o f public resources
withinthe major economic and social sectors? This then is the second major theme o f the analysis inthis
report.

1.55      While public development and social spending must sharply recover from the extremely
depressed levels o f the past decade and must play a significant part in reviving growth and reducing
poverty, the biggest economic and social gains must come from improved efficiency in the use of public
resources. Improving the effectiveness o f spending not only through a better allocation o f resources and a
more careful choice o f priorities, but also through better implementation, more efficient delivery o f
services, and pluggingof leakages and waste i s the third strategic issue discussed inthis report. This goal
requires, above all, restoring the capability and the authority o f public institutions, especially those
concerned with economic and financial management and the delivery o f social services.

1.56      Public institutions in Pakistan have suffered a long-term decline because o f poor pay for civil
service, increasing corruption, and widespread political interference in administrative matter~.'~The
institutional issues have a special urgency because o f the major devolution o f responsibilities to the
district level. The turnaround in efficiency o f education and health spending that i s needed now hinges on
the success o f the new district governments. But the adequacy o f planning, budgeting, expenditure
control, and monitoring and evaluation processes at all levels o f government need to be examined in
relation to both the distribution o f responsibilities among various levels o f government and the capability
o f the civil service to respond and contribute to the desired productivity gains. In looking at the selected
sector allocation issues, the analysis focuses more on the institutional strengthening needed to determine
priorities and improve implementationand less on the evaluation o f specific government plans that, inany
case, become quickly out o f date. The effectiveness of expenditure analysis gives attention to both the
process issues and the institutional capacity problems.




  "Pakistan: A Framework for Civil Service Reform in Pakistan," December 1998, World Bank; "Pakistan Development Policy
Review: A New Dawn?" April 2002, World Bank; and "Pakistan Country Financial Accountability Assessment" (Phase I),
World Bank, 2001.




                                                           17


   CHAPTER ENSURING2:                  FINANCIAL          DISCIPLINEAND CREATINGFISCAL                           SPACE

Introduction

2.1.     This chapter examines the key elements o f macroeconomic and fiscal policies that will ensure
steady and simultaneous progress toward the goals o f debt burden reduction, acceleration in economic
growth, and poverty alleviation with the help o f a medium-term macroeconomic and fiscal fkamework
that integrates growth, debt reduction, and poverty alleviation goals.

2.2.     The focus, however, is on public expenditures in the context of a consolidated medium-term
budgetary framework18 and on the critical issue o f how to create fiscal space for growth-enhancing and
poverty-reducing expenditures while maintaining financial discipline and malung progress in reducing
public debt burdento sustainable levels

2.3.     This focus involves the examination o f the scope of expanding social and development spending,
while maintaining the current progress on public debt burdenreduction (and keeping aggregate spending
at about the same level in% o f GDP). The illustrative baseline scenario explores the scope for mobilizing
more government revenues, cutting public enterprise losses, reducing o f interest payments, containing
defense spending and mobilizing privatization proceeds.                  This scenario representing macroeconomic
stability and continued structural reforms resulting inproductivity improvements i s contrasted with a low
case in which the structural reform sputters; there i s no significant progress in tax revenue mobilization;
public enterprise losses and defense spending increases inrelative terms; there i s little further progress in
reducing the fiscal deficit; and consequently there i s less external support in the form o f grants and
concessionary loans. Quite obviously in this low scenario there will be only limitedprogress towards the
goals of debt reduction, revival o f growth andpoverty alleviation.

Turn Around inDebtPosition

2.4.     The debt problems, which have been a major source o f the sharp slowdown in Palustan's
economic growth, and the consequent stagnation in poverty incidence, over the last decade have been
brought under control. While the debt burden, especially the public debt burden, i s still at an
unsustainable level and needs to be brought down steadily, the risk that economic and other shocks will
lead to a worsening o f the public debt burden inthe medium term appears small.

2.5.     For nearly two decades, the rising interest payments limited fiscal space and squeezed public
spending on development and social sectors. At their peak in 1999-2000, interest payments accounted for
nearly half the govemment revenue. The high interest payments reduced pro-growth public spending
directly while indirectly they discouraged private investment because the large government borrowing
kept domestic real interest rates high and tended to crowd out the private sector. Private investment also
suffered due to the shortages o f complimentary infrastructure, limited access to international capital
markets, and declining confidence inthe domestic currency, which contributed to capital flight.




  The medium-termbudgetaryframework beingpresentedhere consolidatesthe budgetary accounts of both federal govemment
and the provinces. The purpose is merely to illustratethe public expenditure level and pattem for the country as a whole and to
highlight future expenditure options. Since Pakistan is a federation, the operationalbudgetary frameworks must be prepared
separately for the central govemment and for the provinceso f Punjab, Sindh, NorthWest Frontier, and Balochistan.In addition,
AJK andtheNorthemAreas have independentbudgets.

2.6.       Insharp contrast to the almost inexorable rise inthe debt burdeninthe 1980s and 199Os,there
was a significant decline o f 21 percent inthe present value o f public debt from 600 percent to 472 percent
o f government revenues, between mid 1999 and mid 2003 (Table 2.1). The ratio o f interest payments to
government revenues dropped to about 30 percent during FY 2003. The drop in the external public debt
burden in present value terms and as a proportion o f exports o f goods and non-factor services has been
even more pronounced              -- 39 percent between mid 1999 and mid 2003. The dramatic increase in foreign
exchange reserves, to $9.5 billion by mid 2003 from $0.9 billion in mid 2000, has transformed the
liquidity position. The improvement in debt indicators is attributable to the impact o f policy reforms as
well as a number o f external factors that have resulted in a narrowing o f fiscal and external current
account deficits (excluding private and public transfers), reduction in cost o f government borrowing,
 sharp rise in private remittances, and increase in concessional and grant external assistance. The very
favorable terms o f rescheduling o f Paris Club debt have also contributed significantly to the reduction in
the burden o f debt. Still the external debt burden i s too high. By the criteria applied under HIPC, the
present value o f public and publicly guaranteed external debt in relation to exports o f goods and non-
factor services should not exceed 150percent. InPalustan, this ratio i s currently about 200 percent.

                                            Table 2.1: ExternalandPublic DebtBurden

                                                              1989/90       1995/96 1998/99       1999/00    2000/01 2001/02 2002/031'

                                                                                          (US$Billion)
Total ExtemalDebt (excl.foreignexchangeliabilities)              19.4         31.0      33.5         32.4       32.2       33.5        33.4
Total ExtemalDebt (excl. foreignexchangeliabilities)
   as a% of GDP                                                  48.5         48.7      58.8         59.5       54.8       56.5        48.7

Total External Debt (excl. foreignexchangeliabilities)
   as a % of Foreign ExchangeEamings'                           234.7        330.6      341.0       324.9      327.0      287.3       287.2


Total External Debt (incl. foreignexchangeliabilities)           22.0         31.3      34.7         38.1       37.2       36.6        35.5
Total External Debt (incl. foreignexchangeliabilities)
    as a % of GDP                                                55.0         49.1      61.0         69.9       63.4       61.8        51.8

Total External Debt (incl. foreignexchangeliabilities)
   as a% of ForeignExchangeEamings                              266.2         333.3     353.9       381.7       377.9     314.2       305.4

                                                                                          (Rs. Billion)
Total PublicDebt                                                 802.1       1,891.0   3,003.7      3,175.8    3,707.5    3,633.7    3,689.9
Total Public Debt as a %of GDP                                   93.7         89.2      102.2       100.9       108.3     100.1        91.8
Total Public Debt as a %of Revenues                              505.1        510.3     632.5       619.6       670.5     588.9       512.0

                                                                                          (US$Billion)
ExtemalPublic & Publicly Guaranteed Debt                         19.1         27.9      30.2         29.6       29.8       31.2        31.3

Memorandum items:
PresentValue of Total Public Debt as a% of Revenues"                                    599.8                             528.1       472.3
PresentValue of External Publicand Publicly Guaranteed
Debt as a % of Exportsof GoodsandNon-Factor                                             320.1                             248.7        196.1
+mice9 3141
I-...---
Source: "Debt BurdenReductionand ManagementStrategy", GoP, 2001; GovernmentAuthorities; and Staff Estimates.
"Datafor2002/03 isprovisional.
" Foreignexchangeeamingsincludegoods, services, income, andworkers'remittances(WorldBankdefinitionasusedintheGlobalDevelopmentFinancereports).
"Estimatesassumea10%grantelementin1998199,increasingtoanexpected20%in2002103.
" Estimatesassumea10%grantelementin1998/99,increasingtoanexpected20%in2002/03.


 2.7.       The government's basic approach to debt management and reduction i s sound. The government
 attaches high importance to steady reduction in the burden o f public debt but has at the same time opted
 for a ten-year time frame for attaining debt sustainability, defined as 60 percent o f GDP, which will not
require harsh cuts inthe already squeezed levels of public spending. The government recognizes that debt



                                                                          19

reduction, vital for long-term sustainable growth, cannot be the only economic goal even inthe short term
and must be balanced with the need to accelerate growth, create jobs, and reduce poverty in the short
term. We endorse this target and the basic strategy. As long as the public debt i s moving on a favorable
downward trajectory, most o f the negative effects o f highburden o f debt on growth and poverty, already
noted, can be minimized.

2.8.      The Government has approved a Draft Fiscal Sustainability and Debt Limitation Law. It i s a good
step towards operationalizing the government's debt reduction strategy. It specifies fiscal rules to
constrain discretion on fiscal policies and enforce fiscal discipline consistent with the target o f reducing
gross public debt to 60 percent of GDP by FY11/12. It limits annual issuance o f explicit guarantees to less
than 2 percent o f GDP, including rolling over existing guarantees and payments o f called guarantees. The
Draft Law provides for establishing a Debt Policy Coordination Office (DPCO) to serve as a secretariat
for the Fiscal Sustainability and Debt Limitation Law. It also creates an enforcement mechanism through
public scrutiny by requiring fiscal transparency and open fiscal reporting. Notwithstanding this progress it
would be useful to further refine the Debt Strategy with the following improvements: expand the
coverage o f guarantees subject to the limit in the Draft Law to other explicit and implicit contingent
liabilities; apply the reporting requirements to provincial governments as well; and supplement the
requirement for a mid-year economic report with a requirement for corrective fiscal measures when there
are deviations from mid-year targets. The Government could also increase the overall debt strategy's
credibility and its operational relevance with the additional suggested refinements presentedinBox 2.1.

2.9.      The ten-year goal should be to reduce public debt as a share o f revenues to 300 percent (and no
higher than 275 percent inpresent-value terms). For the medium term, a goal o freducingpublic debt as a
share of revenues from the 512 percent in mid 2003 to 400 percent by mid 2007 appears desirable. This
step will reduce the public debt burdeninterms o f GDP to 70-75 percent by the end o f FY07, which will
be on track to reachthe 60 percent goal by FY12.

2.10.     What would the target o f public debt reduction to 400 percent o f government revenues by FY07
mean for the level o f fiscal deficits during FYs 2003-07? More important, how can the medium-term
economic and fiscal policies ensure achievement o f the debt reduction goal consistent with progress in
reaching goals o f growth acceleration and poverty alleviation? Box 2.2 presents an analytical framework
for examining these issues. Broadly speakmg, as the analytical framework suggests, three sets o f
economic policies will be key to the success o f the government's strategy o f growth revival and debt
reduction:

    Policies that slow down the future rate of government borrowing by keeping the primary balance
    positive (normal fiscal balance before interest payments) and reduce public enterprise losses and
    payments o f these and other contingent liabilities. Privatization can help limit public enterprise losses
    and generate revenues, which reduce the need for new public borrowing. It can also facilitate growth
    through productivity improvements and new investments. Needless to add, the mobilization o f
    extemal grants can reduce the overall needfor borrowing.
    Reductions in the overall cost of government borrowing, both domestic and external, through
    better debt management.
     Structural economic reforms that promote growth, exports, and government revenues through
    increases ininvestment and productivity as well as governance improvements.




                                                     20

                                          Box 2.1: Refiningthe Debt Strategy

The government could increase the credibility and operational relevance o f its debt strategy with the following
refinements:

1.       Because the overall time frame for working out the excessive debt burden is long, Pakistan should define
medium-term debt burden goals that canbecome an integral part o f a budgetary framework.

2.       Public debt burden goals should be defined in terms o f both GDP and govemment revenues. In a sense
revenues are a muchbetter indicator of government's ability to service debt than GDP. Inany case it i s the public
debt-to-revenue ratio that i s exceptionally highinPakistan and needs to be brought down sharply.

3.       The government should set explicit medium-term targets for levels o f external debt in relation to
exports/foreign exchange earnings so that it can achieve the HIPC initiative criterion o f reducing the present value
o f external debt to exports o f goods and nonfactor services to 150 percent (from the present level of about 200
percent, assuming a grant element in external debt o f about 20 percent). This step will strengthen the case for
continued large concessionary assistance and external grants.

4.       Related to the previous recommendation, the external borrowing strategy for the next few years needs to
be made explicit and linked to the fiscal deficit financing requirements so that a clear domestic borrowing strategy
can also emerge. Optimal mobilization o f concessionary assistance will greatly alleviate the burden o f domestic
financing   and, combined with financial discipline, could lead to declining real levels o f outstanding public
domestic debt inthe not-too-distant future.

5.       The DPCO should give a high priority to developing estimates o f the present value of both external and
public debt. The calculation o f present value for public debt is not common. However, because more than 50
percent o f Pakistan's public debt i s external and a large part o f it i s concessionary, present-value calculations
assume special importance.        Estimates o f present value will help Pakistan set more appropriate debt burden
reduction goals andbetter enable it to monitor progresstoward them.

6.        The government should survey the exposure to exchange-rate and interest-rate risk o n its external loan
portfolio and, if necessary, restructure its currency composition and possibly use market-based hedging
instruments to manage these risks. The currency composition o f the external debt should reflect as far as possible
the currency composition o f external trade and official international reserves. In addition, market-based risk
management such as entering into currency swaps and interest-rate swaps at the time o f borrowing can hedge
against exchange and interest-rate fluctuations at a relatively l o w cost. Furthermore, the government could
minimize the risks o f commodity price fluctuations by employing commodity swaps, put options, or commodity-
indexed loans.

7.        Each year the DPCO should reconcile the final fiscal deficit numbers with the public debt stock changes
inorder to ensure accuracy andtransparency.' Inthe past, debt growthhasbeenconsistently larger thanreported
fiscal deficits.


'/ For instance, whereas the cumulativerecordeddomestic borrowing by the consolidated govemment to finance fiscal deficits
from FY99 to FY02 was Rs. 827 billion, the increase in the govemment's domestic debt (excluding SBP) in the same period
was Rs. 1,054 billion. The difference arises from the govemment's taking on liabilities of other public sector entities or other
operationsoutsidethe budget.




                                                           21

I                          Box 2.2: Determinants of the Rate of Public Debt Accumulation

 The analytical framework derives the following conclusions about the evolution o f the public debt burden. It shows
 (equation (2)) that the public debt burden (measured as a percentage of GDP) will decline if:

 1 The borrowing need is lower. The lower the expression in the first bracket (the sum o f the primary fiscal deficit
   and contingent liabilities paid by the government minus the fiiancing from privatization receipts and foreign
   grants), the lower is the borrowing need and the accumulation o f new debt. In particular, when the sum o f the
   primary fiscal deficit and payments o f contingent liabilities are lower than privatization receipts and foreign grants,
   the debt burden (debt/GDP ratio) declines, other things being equal.
 1 Domestic and foreign real interest rates on public debt are lower.
 1 Pakistan's real GDP growth rate i s higher.
 1 The Pakistani rupee appreciates in real terms. When the nominal exchange rate measured inlocal currency units
   per foreign currency unit appreciates, the local currency value o f foreign currency debt and the local currency cost
   of servicing this debt declines. Inthe case o f nominal depreciation, the exchange losses on the external debt can be
   substantial, since the local currency value o fthe whole stock o f foreign debt i s revalued. Ifthe nominal depreciation
   goes along with a real depreciation, the rise in the GDP deflator does not fully offset the valuation loss from the
   nominal depreciation and overall the debt/GDP ratio rises. Ina stable real exchange rate environment thls last factor
   inthe equationis zero. (Ofcourse, the exchange rate should only appreciate inrealterms when this isthe long-run
   equilibrium real exchange rate matched by domestic productivity gains. If not, the loss o f competitiveness from an
   overvalued exchange rate would leadto M e r accumulation o f extemal debt.)

 The time path for the evolution o f public debt is given by the relationship:

 Dt+l     =DDOMt+i DFOR$t+IERt+i
           +DDOMt
             DDOMt(P+nD)+DF0R$,ERt(rF +nF)+DFOR$tERte'rt+l +DFOR$tERte'rt+l(rF +nF)
                      ++DFOR$tERt +(PDt+i -PRIVZNtii +CL+i- GRANTSt+i)
                                                                                                                (1)

 where
 Dt=totalpublic debt at the endofperiodt,
 DDOM=domestic public debt,
 DFOR$=foreign public debt measuredinunits o f foreign currency,
 ER=the nominal exchange rate, local currency per unito f foreign currency,
 PD= primary fiscal deficit
 PRIVzN= fiscal revenue fromprivatization,
 CL= contingent liabilities being assumedbythe government,
 GRANTS = foreign grants,
 rD=domestic realinterestrate, rF:foreignrealinterestrate,
 nD=domestic inflationrate, nF:foreign inflationrate,
 e*rt= percent change innominal exchange rate from one period to next (+ shows depreciation),
 re'rt percent change inreal exchange rate fiom one period to next (+ shows depreciation)
      =


 Dividingthroughout by GDPt+l the following relationship is derived for the change inthe debt/GDP ratio (lower case
 shows all variables inpercent o f GDP):

 ddt      =(pd,+l -pri~m,,~+clt+l   -          - ddonq(g+nDY(l+g+nD) - dfort(g+nDY(l+g+n? + ddonq(P+n%l+g+nD) +
          dfort(?+nFy(l+g+nD) + dfor,e*r,,{( l+g+nD)+ dfor,e'rt+l(?+nF y(l+g+n?

          =(pd,,,
            + dfort)[- - nD+?+nF+e'r, + e'rt+l(?+nF)]
                  -privznt+l+c1,+] -grantqtl) +ddom,[- g-nD+P+ny/(l+g+n?
                      g                                 /(l+g+nD)

          =(pd,+, -privzn,+l + cl,+]-          +ddom,(P -g)/(l+g+nD)
            +dfor,[(?- g) + (nF- nD+ e*rttl)+ e'rt+l(?+nF)]/(l+g+n?

          =(pdt+l
             +dfor,[(p-
                   -privzn,+l+       - grants,+l) +ddom,(? -g)/(l+gd)
                         g) + e*rt+l(pd) + re*r,+l ]/(l+g-he)                (2)

 *This result (2) is described indetail inthe bullet points above.

                                                                22

     Baseline Scenario

     2.11.    The government's reform program stresses all three policy elements mentioned. The question i s
     whether these policies will be implemented in a timely fashion to create fiscal space that will allow the
     public sector to play its role in pursuit of the goals o f growth revival, poverty alleviation, and social
     development. Tables 2.2 and 2.3 summarize the assumptions and findings o f a baseline scenario that
     explores this question within the context of a medium-term macroeconomic framework that integrates
     real sector, fiscal accounts, and balance of payments projections. This section provides a brief overview
     ofthe key points. A more detailed description and technical material are providedinAnnex B.

               Table 2.2: Summary of key macroeconomicvariables in baseline scenario, 1999/00-2006/07
                                                                                      Est.Prov.                        Proj.
                                                            1999/00 2000/01      2001/02 2002/03 2003/04           2004105 2005/06 2006/07

Real GDP at market prices                                        4.2        2.6       2.8       5.8           5.3      5.5     5.7     5.9
Consumer prices (p.a.)                                           3.6       4.4        2.7       3.1           4.0      4.0    4.0      4.0

Savingsandinvestment
  Grossnationalsavings                                          14.1      13.6       17.4      21.3          16.7     18.0    18.3    18.5
  Gross fixed capital formation                                 14.4      13.9       13.1      13.1          14.5     15.5    16.1    16.7
  Changein stocks (nongovt)                                      1.6        1.6       1.6       2.4           2.0       1.6    1.6     1.6

Publicfinances
  Revenue (including grants)                                    17.3      17.3       19.3      20.8           18.5    19.3    19.6    19.8
  Grants                                                         1.1        1.2       2.3       2.8            1.1      1.7    1.6     1.5
  Revenue                                                       16.3      16.2       17.0      17.9           17.3     17.6   18.0    18.3
  Expenditure                                                   22.8      21.4       23.6      22.3          21.8     22.4    22.4    22.6
      Interest payments                                          7.8        6.8       6.8        5.2           4.8     4.6     4.2     3.9
      Defense                                                    4.8        3.8       4.1        4.0           3.6     4.0     4.0     4.0
      Paymentof contingent liabilities                           0.5        0.6       1.9        1.2           1.9      1.3    0.8     0.5
      Non-interest,non-defense,non-SOE spending                  9.8      10.2       10.9      12.0           11.6     12.6   13.4    14.3
Overallbalance(including grants)                                -5.5       -4.1      -4.4       -1.6          -3.4     -3.1   -2.8    -2.8
Primarybalance(including grants)                                 2.3        2.8       3.8        3.6           1.4      1.5    1.4      1.o
Grosspublic debt                                               100.9     108.3      100.1      91.8          83.3     80.2    74.2    70.9
Grosspublic debt in% of revenue                                619.6    670.5       588.9     512.0        480.6     456.3  411.40   387.3

Extemalsector
  Current account incl. off. transfers                          -2.0       -1.9       2.7        5.9           0.2      0.9    0.6      0.2
  Ext.public andpublicly-guar.debt                              48.6      55.7       52.7      45.8           38.6    35.8    32.8    30.4
   Grossreserves inmonths ofnext year importsof
      goods and services                                         3.8        1.7       3.7        7.0           6.3      6.0    6.2      6.2



                                   Table 2.3: Base Case Scenario: Principal Budget Aggregates

                                                 ChangeFY2002 FY2007 (Percent of GDP)
                                                                  -
                      Resources                                      Expenditures
                       Govemmentrevenues                   + 1.3     Defense expenditure                          -0.1

                       Fiscaldeficits(including grants) - 1.5
                       Extemalgrants                       -0.8      Interestpayments                             -2.9
                                                                     Public enterpriselosses 1/                   -1.4
                                                                        Non-interestnon-defensenon-SOE            + 3.4
                                                                            expenditures
                                                                        o/w Developmentspending                    1.5

                       Total resources                     -1.O      Total expenditures                           -1.0
                       1/ Public enterpriselossesare approximatedby subsidiesandpaymentof contingent liabilities.



                                                                      23

2.12.    The analysis suggests that to achieve the desired public debt reduction goal by FY07, the
government should bring down the overall fiscal deficit (including external grants, provision for public
enterprise losses, and other contingency payments) to an average o f 3 percent o f GDP duringFYs 2004-
07, compared with the average o f 4.3 percent duringFYs 2001 - 02 (see Table 2.2). Inlight o f the strong
fiscal performance in FY03 when the fiscal deficit (excluding grants) declined by 2 percent o f GDP the
fiscal adjustment required over the medium term i s not large. But the scenario suggests no room for
expanding the overall level o f public expenditures, which i s projectedat 22.3 percent o f GDP on average,
the same level as in FY03. The projected modest growth inrevenues i s partly offset by the reduction in
external grants (which were at an exceptionally high level in FYs 02 and 03) and partly absorbed by the
need to reduce the deficit.

2.13.    The projected level o f external grants assumed for FYs 2004-07 (1.5 percent o f GDP) i s lower
than in the last two years but somewhat higher than in earlier years. Provided reform efforts continue,
there would be a strong case for the donor community to give a high level o f budgetary grant support to
Pakistan because o f its heavy public debt burden and less favorable social indicators." But donors will
also expect a restructuring o f expenditure toward economic and social sectors and generally greater
effectiveness inpublic spending.

2.14.    Our projections show that if defense spending can be contained and losses o f public enterprises
and the interest payments can be brought down, non-interest, non-defense and non-SOE spending will rise
to 14.3 percent in FY07, compared with 10.9 percent in FY02 (and an estimated 12.0 percent in FY03;
see Table 2.2). The policy issues and options involved inbringingabout a shift inpublic spending toward
development and social spending are discussed at some lengthbelow.

Slowing theRate of Borrowing

2.15.    The first step in slowing down government borrowing i s to continue to run significant primary
surpluses. This means that government's own resources and not fresh borrowing should cover at least a
part of the interest payments. Untilrelatively recently, Palustan actually ran large primary deficits, which
meant that all o f the borrowing costs (and off-budget borrowing) were financed from fresh borrowing.
During the past few years Pakistan has made clear progress; the primary surplus (excluding grants)
averaged over 1percent o f GDP during 1998/99 - 2002/03. Under the Debt Committee projections and
the PRGFprogramwith the IMF,the primary surplus was expectedto rise to well over 2 percent o f GDP
by FY 2004. This level no longer appears either realistic or necessary because our projected fiscal path
shows sufficient progress toward debt sustainability with a somewhat less restrictive fiscal stance.
Pressures for expanding non-interest public spending from the very low level reached in FYOO have
grown following the finalization o f the I-PRSP, the continued sluggishprivate sector, and hence slow job
creation.

2.16.    Nonetheless, the government will need to weigh some choices. Clearly, it will be desirable to
maintain a significant primary surplus (including grants) in FYs 2004-07 in order to underpin continued
progress toward debt reduction. Yet genuine political and economic pressures for development and social
spending are mounting. Further growth in tax revenues as a percentage o f GDP, reduction in public
enterprise losses, and acceleration o f privatization receipts can help to resolve the government's dilemmas
and minimize the short-term conflict among key economic and social goals.




   The external grants projections also reflect the recent announcement by the US government that it will provide support
amounting to US$3 billion over the next 5 years.


                                                             24

Increasing Revenues

2.17.     Pakistan's revenue mobilization has been low, both by international standards and compared with
its potential. In FYs 2000 and 2001 total revenue collection was the equivalent o f 16.2 percent o f GDP;
tax revenue was only 13 percent o f GDP and had actually stagnated for several years.20However, revenue
collections during the last two years increased sharply to 17.9percent o f GDP in2002-03 reflecting both
the success o f tax collection efforts and the revival o f the economy. The structure o f taxation has also
improved considerably in recent years, and the heavy dependence on foreign trade taxes has been
reduced. Still, revenue mobilization falls short of that o f many developing countries at comparable
income levels in Afi-ica, Asia, and Latin America.*l In light o f the weaknesses in the current revenue
mobilization system, there i s considerable scope for enhancing collection and broadening the income tax
base. For instance, while there are only 1-1.5 million income tax filers, there are 3 million cell phone
subscribers, over 10million electricity consumers, 3-4 million gas consumers and 2-3 million car owners.

2.18.     The main factors behind the low revenue mobilization are governance problems in tax
administration, a narrow and inflexible tax base, a large informal economy that escapes the tax net, and
pervasive smuggling with associated revenue losses. Governance problems have been reflected in
widespread collusion (corruption) between taxpayers and tax officials, a situation that leads to tax evasion
and lack of tax compliance. These problems stem from the discretionary powers o f tax officials, complex
tax rules, and weak supervision o f staff.

2.19.     To address these problems, the government took a number o f actions beginninginFY02.

          (i)theareaoftaxadministration,thegovernmenthasbegunafundamentalrestructuringofthe
             In
             CentralBoard o fRevenue (CBR). This restructuringincludes a complete reorganization over a
             three-year period (FYs 02-04) and fundamental changes in human resource management. A
             Large Taxpayers Unit in Karachi has started operations, and a pilot income tax office for
             medium-sized taxpayers was recently opened in Lahore. Both units operate with greater
             functional specialization in tax administration. Universal self-assessment for income tax and
             audits to reduce contact with tax officials has also been introduced. The Government and the
             World Bank are preparing a major technical assistance program for tax administration reforms.

         (ii) policieshavealsobeenstrengthened: the coverageofthe general salestaxhasbeen
              Tax
             expanded to include agricultural inputs and now requires compliance o f small traders. A new
             income tax policy came into effect in FY 03 with a number o f reform measures (minimal tax
             exemptions, simplified and more equitable tax rates). The provinces have started collections o f
             the agricultural income tax for high-income farmers that were instituted in2000.

2.20.     Inour view, it should notbe difficult, inlight of current efforts, to increase revenuemobilization
further to 18.3 percent of GDP by FY07, the assumption in the base case. But the suggested level o f
revenue mobilization will merelyregain the level o f the early 1990s.

2.21.     To consolidate recent revenue gains and to underpin future growth o f the economy, determined
efforts are necessary to accelerate institutional and policy reform, especially the fundamental restructuring
of CBR in the next few years. Pervasive smugglingmust be addressed resolutely. It i s a major cause o f
loss o f revenue and foreign exchange. In the area o f tax policy, reforms should aim at continuing to


~~




2oNote that Pakistan has undertaken considerable trade liberalization inrecent years with an initial import tariff revenue loss.
*'Countries at comparable income levels with higher revenue/GDP ratios include Bhutan, Botswana, Kenya, Maldives, Namibia,
Nicaragua, Sri Lanka, Vietnam, and Zambia. All o f these countries had revenue mobilization in excess o f 20 percent o f GDP,
with the exception of Sri Lanka and Zambia.


                                                               25

reduce the number of tax exemptions (such as for customs duties and income and withholding tax) and
make sales taxation more effective. Reforms are also needed for tax on income from financial and land
assets. Inparticular the current practice that allows a single person to hold assets under different names
should be changed, since it weakens tax collection on income from such assets. On the nontax revenue
side it will be necessary to seek greater cost recovery in higher education, health (subject to
considerations about access to basic services for the very poor), and water usage for irrigation.

2.22.    Inaddition, the government should undertake administrative and governance reforms to ensure
better documentation o f economic transactions, ownership o f assets, and sources o f income, by, among
other things, expanding efforts to ensure auditing o f more businesses. Better documentation will enable
the government to implement existing tax laws more effectively as well as to broaden the tax base inthe
future.

2.23.    Sustained increase in fiscal revenue mobilization nationwide requires stronger revenue effort not
only by the federal government but also by provincial governments. Provincial resource mobilization has
dropped steadily in recent years and i s currently only 0.5 percent o f GDP. Provincial own revenue
collection would need to increase in line with the expansion in sub-national spending foreseen in this
report (in particular on education, health and certain types o f infrastructure) to reach the Government's
goals for poverty reduction and to be consistent with the ongoing devolution process. However,
provincial (and local) governments only have tax authority over taxes which for the most part are
characterized by narrow, inelastic bases and low buoyancy. Inaddition there has been poor utilization o f
provincial and local taxes and user charges, reflecting both lack o f political will to collect key provincial
taxes and deficient revenue administrationmarredby governance problems.

2.24.    Increased provincial revenue mobilization hence requires reforms both in revenue policy and
revenue administration. Revenue policy reforms are needed to expand the provincial revenue base. This
would require a review o f tax authority powers o f the provincial as well as local governments. Reforms
could possibly include allowing the provinces to share selected federal taxes through levying "add-ons"
on the federal taxes with federal collection; determinedly expand and enforce collection o f the agricultural
income tax; and increase reliance on user charges wherever appropriate and feasible. Furthermore,
incentives for sub-national level resource mobilization need to be strengthened as part o f larger
[discretionary] resource transfers to the provincial and local governments that appear necessary. Such
transfers can be linked to provincial revenue efforts and demonstration o f capacity to effectively utilize
resources. Significant revenue gains could also be realized by strengthening the provincial tax and non
tax administrations and plugleakages of revenue; by mergingthe two revenue collecting departments into
one; by better liaising with the federal Central Board o f Revenue; through improved automation and
computerization in key areas of provincial taxes (e.g., land and property records); and through enhanced
training o f and incentives for tax officials.

Public Enterprise Losses and Other Contingent Liabilities

2.25.    There i s not a sufficiently widespread appreciation by the public and Pakistani policymakers o f
the extent to which very large public enterprise losses have contributed directly or indirectly to the
buildup o f public debt andor contingent liabilities. According to our estimates, cumulative public
enterprise losses and the cost of government guarantees through the end of FY03 amounted to close to Rs.
400 billion, or about 10 percent o f GDP22(see Table 2.4 and Box 2.3). These losses initially remained
partly unfunded and took the form of arrears buildup or borrowing from state-owned banks (that cannot
be serviced), but the losses must eventually be reflected inthe fiscal budget and hence add to public debt.


22
WAPDA and KESC in FYs 01 - 03.
  This estimate i s based on the data contained in the Debt Committee Reportup to FYOO and actual losses of PIA in FYOl and




                                                              26

Insomecases, suchastheequityinjectionsbythe governmentintoWAPDAandKESC, thelossesdonot
enter the budget directly. Nonetheless, they affect the level o f debt by either reducing govemment receipts
or requiring special off-budget government borrowing, such as policy assistance from ADB to facilitate
clearance o fKESC losses.

2.26.    Though Palustan has made progress during the past three years (FYs 01 - 03) in eliminating the
operating losses of a number of public enterprises, notably Palustan Railways, Palustan International
Airlines, and most state-owned banks, the losses in the power sector continue at a high level. Indeed,
WAPDA and KESC losses in FY03 (accrual balance) were close to 1percent o f GDP (Rs 33 billion)
whereas total budgetary support to these companies was Rs 73 billion or 1.8 percent of GDP. The
FY04 budget foresees budgetary support to WAPDA and KESC of 1.3 percent o f GDP. The continued
large losses o f public-ownedpower entities i s the most troubling element o f Pakistan's fiscal position and
the only one that has not shown improvement inrecent years.

2.27.    Many factors account for the continued large losses o f WAPDA and KESC: operational
inefficiencies, weak bill collection, very high levels o f transmission and distribution losses (about 25
percent o f units generated or purchased inthe case o f WAPDA and 40 percent inthe case o f KESC) due
to theft and leakage, inadequate tariff adjustments to increased fuel prices, the relatively high cost o f
purchases from IPPs, including payments for large unused generation capacity, and the politically
mandated practice o f subsidizing household and farm consumption at the expense o f industry and
commerce. Corruption, concerns about law and order, and lack o f development funds have also
contributed to the decline in KESC's financial performance. In addition, WAPDA's finances have been
suffering from the Federally Administered Tribal Area's (FATA) failure to pay utilitybills amounting to
Rs. 12billion annually, part o fwhich may be overbillingbyWAPDA. The coming on stream o fthe large
Ghazi Baroda Hydel Project (1,450 MW) and the substantial decline inpayments to PPs after FY04 will,
however, help to reduce the average generation cost. A pick-up ineconomic activity by increasing power
demand will also reduce the burden o f large, relatively fixed payments to PPs.




                                                     27

                                                     Box 2.3: Contingent Liabilities

  The federal and provincial governmentscould face serious fiscal costs as a result of their contingent liabilities. These are
  fiscal obligations contingent on the occurrence o f particularevents and are not adequatelybudgetedand accountedfor in
  the fiscal budget.

  Pakistan has made some progress in recent years on contingent explicit liabilities, at least on monitoring. These
  liabilities are specific obligations, created by law or contract that the government must settle if particular events occur.
  They include government guarantees to autonomousbodies, public-sector banks, andthe private sector. The govemmentis
  trackingthe issuance and stock of explicit govemment guarantees, andthe Draft Fiscal SustainabilityandDebt Limitation
  Law limits the annual issuance of such guaranteesto 2 percent of GDP. Over the past decade, however, annual issuance of
  such guarantees averaged only 0.6-0.7 percent of GDP. In FY 2002 the impact of contingent explicit liabilities on the
  budget was Rs. 24.4 billion, or 0.7 percent of GDP, andinFY03 it was Rs. 16.2 billion, or 0.4 percent of GDP.

  Contingent implicit liabilities have remained substantial. These are moral obligations or burdens that, although not
  legally binding, are likely to be bome by the govemment because of public expectations or political pressure. In recent
  years public enterprises have run substantial operating losses. The impact o f contingent implicit liabilities on the federal
  budget (and on public debt) stemming from public enterprises was Rs. 126.5 billion in FY02 (3.4 percent of GDP,
  including a major equity injection to KESC) and Rs. 84.5 billion in FY03 (2.1 percent of GDP). Whereas other public
  enterpriseshave improvedfinanciallyinrecent years, the power sector enterprises-WAPDA and KESC-remain major loss
  makers. In FY03 the drain on the budget from these two enterprises amounted to 1.8 percent of GDP'. Without
  fundamental reform they will continue to absorb massive fiscal resources. Vigilance is also needed on other contingent
  liabilities, however, includingnonperfoming loansofthe public-sectorbanks.

  Direct implicit liabilitiesare expected to increase over the medium term. These include the unfundedpart of public
  pensionliabilities, which are expectedto rise considerably. Pensionliabilities must be estimated and made transparent so
  that proper measures can be taken in time to avoid a pension time bomb. The treatment of interest costs on NSS
  instruments leadsto the buildupofadirect implicit liability to bepaidwhen NSS instrumentsare redeemed. Inthe medium
  term a certain lumpiness in redemptionand interest payments i s expected, materializingas a sudden large claim on the
  fiscal budget.

  The governmentcan limit the impact of contingent and direct implicitliabilities by reducingthe potentialsource of
  the problem.Privatizationand restructuringof the remainingpublic enterprises to reduce losses could greatlyreduce the
  pressure onthe govemmentto assume contingent liabilities. Inaddition, guarantees mustbegiven muchmorerestrictively,
  subject to awell-fimctioning tracking, monitoring, and valuation system.
  'Toput    this inperspective, this is the same as the total educationbudget (also 1.8 percent of GDP) and more than double the total health
  budget (0.7 percentof GDP).



2.28.      The impact o f the financial problems in the power sector companies extends beyond the fiscal
budget. These companies have under invested in power generation capacity for some time, and the
country's investment needs in the power sector cannot be met without private sector participation. In
addition to the direct consequences for federal and provincial govemment finances, the power sector
problems have a major impact on private sector competitiveness and hence economic growth, exports,
and employment (See the forthcoming World Bank study "Investment Climate Assessment inPakistan").
Infact, the costs associated with the unreliable power supply are one ofthe manufacturingsector's top
competitiveness concerns.

2.29.     The govemment has adopted a strategy to restructure public enterprises to improve their
operational efficiency and privatize most o f them. In the power sector-the                                      source o f the largest
contingent liabilities and enterprise losses-the                  government has started implementing a strategy that aims
at improving the efficiency and the financial viability o f the enterprises through competition,
accountability, and managerial autonomy, while also establishing a multiyear tariff framework and
improving the regulatory framework. Efficiency gains will be achieved by unbundlingand corporatizing
WAPDA's power wing into independent distribution and thermal generation companies as well as a



                                                                       28

national transmission company. The goal i s to privatize the distribution and generation companies and
keep the transmission company ingovernment ownership.

2.30.    As will be discussed later, although the needto privatizeKESC andparts ofWAPDA has become
more urgent because o f continued losses and inadequate resources for investment as well as the costs to
other industries, the progress toward privatization i s the slowest in the power sector. In part, this slow
progress i s due to factors beyond the government's control, such as the lack o f foreign interest inKESC.

2.31, With regardto the pricing framework, NEPRA has recently followed-albeit             with some delay-
the policy o f adjusting tariffs for changes in fuel prices to achieve financial viability. On occasion
WAPDA has also delayed applying for and implementingtariff increases.

2.32.    It is now clear that inFY03 public enterprise losses were higher than inFY02because ofthe lack
o f financial improvement by the power companies and will remain so in FY04. These higher losses
compared to previous projections will limit the projected expansion in fiscal space over the next two
years. Beyond FY 2004, it shouldbe possible to reduce public enterprise losses through a combination o f
reforms and structural improvementsinthe power sector, lower IPPpayments and increasing share o f low
cost hydro- power. In our view, with reasonable strong policy actions, it should be possible to reduce
overall public enterprise losses to 0.5 percent o f GDP by FY07 from the estimated peak of 1.9 percent o f
GDP inFY 2004.

2.33.    Looking ahead at the remaining reform agenda, WAPDA needs to decisively reduce line losses,
enforce bill collection, if necessary by cutting off nonpayers more aggressively, and ensure timely
implementation o f tariff adjustments determined by NEPRA. The government should help WAPDA in
resolving the issue o f FATA's nonpayment of its utility bills since this problem seems to require a
political solution. The transmission company, remaining in government ownership, i s likely to runmodest
operating losses even at optimum technical/operational efficiency, because o f the govemment's pursuito f
noncommercial objectives (for instance, rural electrification, which has a development and social
objective). The losses should be minimized and limited to the cost o f pursuingnoncommercial objectives
as imposedby government mandate. Over the medium term the losses should be fully financed by grants
from the government to cover the cost o f pursuing the social objectives, and the cost should be made
explicit inthe fiscal budget.

2.34.    With privatization, gains in operating efficiency, reduction in theft, and expected lower
generation costs because of reduced payments to IPPs, the large Ghazi Barota project coming on-stream
and hence an increase in the share o f hydro power, it should realistically be possible to reduce the
budgetary support to WAPDA and KESC from about Rs. 73 billion, or 1.8 percent o f GDP in FY03, to
Rs. 13 billion over the medium term (by FY07). The government needs to ensure, however, that other
public enterprises such as Pakistan Railways, Palustan Steel, PIA, and state-owned banks do not fall back
into deficit and that guarantees provided by government do not have a net cost to the budget (see Box
2.3). Meeting this goal will require the government to continue privatizing most commercial and
industrial public enterprises, to restructure remaining public enterprises and subject them to hard budget
constraints, and to restrain tendencies to set prices administratively in the public sector to protect the
consumer.




                                                     29

                                         Table 2.4: Losses of State Owned Enterprises
                                                                (Rs Million)
                                                                                                                        Cumulative
                                                        Forwardupto
                                                        Losses Carry                  Losses During                     Losses upto
                                                          1998199           1999100         2000/01      2001/02         2001/02

      KESC                                                24,399            12,787          16,203        17,159         70,548
      WAPDA 11                                             36,400            8,527          15,086        33,128         93,141
      Pakistan Steel Mills                                 8,123             1,403           1,065        1,280           11,871
      PakistanRailways                                    25,721             2,197            727         -1,819         26,826
      PIA                                                  7,046             5,242           1,982        -2,156          12,114
      Banks 11                                             30,700            8,000             0          7.900          46,600
       Others 21                                           60,439             NA              N A           NA           60,439
       Total SOEs loss (Debt Report)                      192,828           38,156          35,063        54,492         321,539

       1/ Carry forward losses are approximated by GOP equity injection into the enterprise andmay therefore overstate the accumulated
      losses o f SOEs.
      21 Include: Pakistan National Shipping Corporation (PNSC), Karachi Steel and Engineering Works (KS&EW), Utility Stores
      Corporation (USC), Heavy Mechanical Complex (HMC), State Engineering Corporation (SEC), Heavy Electrical Complex (HEC),
      Pakistan Broadcasting Corporation (PBC) and Pakistan Tourism Development Corporation (PTDC). Also included are the losses (of
       about Rs 50 billion) of Pakistan (TCP), Ghee Corporation o f Pakistan (GCP), Cotton Export Corporation (CEC), Rice Export
       Corporation of Pakistan (RECP) and the Saindak Copper and Gold Mining Unit, which were covered through issuance o f govemment
       guaranteedbonds (as almost all of these entities have been closed down, this debt is serviced by the federal govemment).
      Sources: The figures for carry forward losses of KESC, Pakistan Steel Mills, PIA and Others SOEs are taken from the Govemment of
      Pakistan's report A Debt Burden Reduction and Management Strategy, Finance Division, March 2001. Losses o f SOEs (other than
      Banks) during 1999/2000-2001/2002 are takedcompiled from the IMF's report Selected Issues and Statistical Appendix, October 21,
      2002.


2.35.     According to our estimates, even with continued forceful reform o f the public enterprise sector,
the losses o f public enterprises (mainly power entities) are likely to total Rs.150 billion during FYs 04-
07, higher than the estimate in the Debt Committee Report (Rs. 100 billion during FYs 2000-04). The
Debt Committee also assumed that the enterprise losses and contingent liabilities would be offset by
privatization revenues and thus that there would be no government borrowing beyond the normal budget
deficit. This assumption appears optimistic; we foresee net privatization revenues to the fiscal budget o f
only Rs. 54 billion over the four years, FYs 04-07. Even after taking into account privatization revenues,
enterprise losses are likely to be a drain on the budget. This projection further underlines the need to
vigorously pursue the overlapping agendas o fprivatizationand public enterprise reform.

2.36.     With regard to contingent liabilities, the government should assess the full impact on public
finances beyond the budget and fiscal debt to include these extra-budgetary items. It should calculate the
potential budget cost in expected value terms23 and make the potential fiscal cost o f off-budget items
visible beforehand to ensure a proper cost-benefit assessment o f government support. It should also limit
moral hazard by announcing the limit o f its financial support and then sticking to that limit.

Defense Spending

2.37.     Historically, the high level o f defense spending has contributed to the buildup o f public debt in
two ways. It has preempted public resources for development, and it has been a major factor in the
government's inability to cover current expenditures from current revenues. Government savings (defined
narrowly as the excess o f government revenues over current expenditure) deteriorated from a positive 0.9
percent o f GDP in 1980-85 to a negative 3.1 percent o f GDP in 1996-99. As the Debt Committee has


*'The financial cost of the event ifit occursweightedby itsprobability.


                                                                     30

stressed, there was "a serious upsettingof the balance between defense and development" after the early
1 9 8 0 ~ .Inthe 1990s, however, real spending on defense increased little, and as a percentage o f GDP, it
           * ~
declined steadily from the peak of 6.9 percent inthe second halfo f the 1980s to less than 5 percent during
1998-2000. Despite the spurt in spending inFY02 due to regionaltensions, defense spending remained at
4.7 percent o f GDP (4.0 percent excluding defense pensions, which have been shown separately since
FYO1).The steady relative reductionindefense spending inrecent years has helped to contain the fiscal
deficit, reduce the negative government saving (to less than 2 percent o f GDP inFY02), and to a degree
correct the imbalance between defense and development.

2.38.      The government feels that in current circumstances assuming further reduction in defense
spending as a percentage o f GDP would not be realistic. While easing o f tensions on the Easternborder i s
very likely, the situation on the Western border as well as combating terrorism in the country will
continue to put pressure on the defense expenditures The report, therefore, assumes that in the base case
scenario defense spending (excluding pensions) in FY04 would be as in the federal budget approved by
National Assembly and in coming years would remain at 4 per cent o f GDP. T o the extent that the
effectiveness o f defense spending can be improved without compromising national security, the
government's policy options will be expanded and the level o f defense expenditures will not be entirely
hostage to regional geopolitical realities.

The Cost of Borrowing

2.39.     As the analytical framework inBox 2.2 indicates, the real cost o fborrowing is a critical variable
in determining the growth of the debt burden. In Pakistan, because the primary balance (including
payment o f contingent liabilities) has been in surplus for some time, the growth o f public debt in recent
years has been driven almost entirely by the cost o f borrowing. Unfortunately, the real cost o f public debt
rose to 4.5 percent annually during 1996-99, compared with the historical level and intemational norm o f
2.5 percent. Surprising though it may seem, real borrowing costs during 2000-02 were even higher, at 5.2
percent per year (see Table 2.5). The highreal cost o f public debt since the mid-1990s reflects ajump in
the cost o f both external and domestic borrowing.25 On the domestic side, the sharp deceleration in the
rate o f inflation from more than 11percent per year inthe mid-1990s to less than 4 percent during the past
three years, combined with the locked-in high interest rates on National Saving Schemes (NSS)
instruments, kept the real annual cost of domestic debt at 7.4 percent (11.5 percent in nominal terms)
during 1999/2000-2001/02. Onthe extemal side, the nominal interestrate on external public debt i s about
2%-3 percent. The real net depreciation inthe value o f the rupee over FYs 1999-2001 however, added to
the real cost o f external debt, raising it to 7.6 percent per year. Including the recent period o f real
appreciation inFY 2002, the real cost o f extemal borrowing averaged 3.0 duringFYs 1999-2002 and was
substantially below the 4.7 percent annual figure for 1996-99.26

2.40.     A significant reduction in the cost o f public debt must be a part o f the strategy to reduce the
burden o f debt. The objective o f policy should be to sharply reduce the overall real cost o f public
borrowing over the medium term. Unless the cost o f public debt i s brought down, there will be limited
scope for sharply reducing the share of interest payments in total government spending, at present 30
percent.




24Debt Committee Report, pp. 14-19.
25The real cost o f domestic borrowing is the average nominal interest rate minus the rate o f inflation. In estimating the real cost
of foreign debt, the capital loss on foreign exchange because o f the depreciation o f the rupee inrelation to the U.S.dollar must be
added to the nominal interest payments, and then this nominal cost must be adjusted for the rate o f domestic inflation. For details,
see Debt Committee Report, pp. 18-19.
26The estimate reflects the high rate o f real depreciation of the rupee during the period.


                                                                   31

                   Table 2.5. Pakistan:Realborrowingcost onpublicdebt, FY1996 FY2002
                                                                                   -

                                                       Realborrowingcost (%) "

                                                       FY96-99          FY00-02

                              Domesticpublicdebt          4.3             7.4
                              Externalpublic debt         4.7             3.0
                              Weightedaverage             4.5             5.2
                              1/ Compoundrates

Domestic Debt Costs

2.41.     The most immediate problem i s the high real interest rate on domestic debt. The average real
interest rate peaked in FYOO at 10.8 percent and declined to 5.6 percent in FY03. In the past, the high
level o f domestic interest rates reflected the upward pressure placed on the country's narrow financial
markets by the considerable borrowing needs of the government. N o w that the net government
borrowings are projected to be modest (see Table A.2 in Annex B), the government has much greater
degrees o f freedom inthe management o f its domestic debt.

2.42.     A central problem i s the still highcost of borrowing under the National Saving Schemes (NSS),
notwithstandingrecent rate reductions. Interest rates on the NSS instruments such as the ten-year Defence
Savings Certificates (DSCs), the five-year Regular Income Certificates (RICs), and the three-year Special
Savings Certificates (SSCs) are fixed administratively. Holders o f NSS instruments can redeem them at
any time, encashing both the face value o f the bond and the accumulated interest earnings. On the initial
assumption that NSS serves a population o f small, relatively unsophisticated savers (such as widows,
orphans, and pensioners), who are unlikely to place their funds with banks or other financial institutions,
rates o f return on these instruments tended to be generous and tax free. Inthe 1980s, as budget deficits
grew, reliance on NSS instruments increased, and these certificates were made extremely attractive
through highly positive tax-free real returns. Naturally this attracted major institutional investors. During
the period from late 1996 to mid-1999, compound rates on DSCs of ten years' maturity peaked at 18.04
percent when held to maturity. The yields on all NSS instruments were tax free except that Zakat was
payable. Since the interest is paid only on encashment, the large accumulated interest liability i s not
reflected in the reported stock o f NSS debt. This situation implies two things: an understatement o f
government debt obligations and the prospect that a sizable part o f the govemment's interest bill will
continue to be driven by the high-cost certificates issued during 1997-99.

2.43.      The government has taken several strong measures in recent years to address some o f the
weaknesses o f NSS instruments. It eliminated the access o f institutional investors to the schemes in early
2000. Recently it took commercial banks out because of abuses o f the system. It has made a number of
successive reductions in the administratively set interest rates on new issues since M a y 1999 and has
linked the NSS rates to the market benchmark Pakistan Investment Bonds (PIBs), which have been
revived as an important instrument o f debt management. Furthermore, the income from NSS on deposits
larger than Rs. 150,000 was made taxable effective July 1, 2001, though this change applies only to new
issues. In the most recent rate reduction in July 2003, the rate on DSCs with ten-year maturity was
reduced to 8.5 percent per year. The rates on 3- and 5-year instruments remain at 7.7 percent.

2.44.     However, two fundamental problems remain. First, despite the actions taken inrecent years, NSS
i s still a source o f high-cost borrowing for the government. This i s because the sharp drop in the rate o f
inflation has greatly offset the impact o f nominal interest rate reductions for NSS on the real rates.
Second, the on-tap nature o f the schemes limits the govemment's options to explore optimal mix o f



                                                      32

borrowing. Also, the way the NSS instruments are administered makes it difficult for the government to
accurately plan the interest payments inthe coming budget years.

2.45.    Despite the sharp reductions made innominal rates on NSS instruments during FYs 2000-03 real
long-term rates o f retum on DSCs o f 4 ?4percent (assuming an annual inflation rate o f 4 percent) and on
the recently introduced 10-year certificates for widows and pensioners o f 6 percent on what are risk-free
assets are still highby international standards. The spread between the yield on PIBs and comparable NSS
instruments should be eliminated. This spread i s currently 2.9 - 4.5 percent in the case o f 10-year
instruments, 3.5 percent for 5-year instruments and 4.6 percent for 3-year instruments. Also, the rates on
SSCs and RICs were not adjusted downward as much as the DSC rates during the past two years, and
thus there was internal shift away from DSCs. The continued attractiveness of NSS instruments can be
judged from the fact that their share in total domestic debt continues to rise, increasing from 37 to 45
percent from FYOl to FY03. Indeed, over the past two years the expansion in the stock o f NSS
instruments was more than double the increase inoverall public domestic debt.

                                Box 2.4: Beneficiaries of National Saving Schemes

      The after-tax return onNSS instruments has persistently been generous compared with market rates, and
      hence this instrument has been an expensive source o f funding of the government deficit. The stated
      rationale for paying high interest rates on these instruments has been that they serve a population of
      small, relatively unsophisticated savers (such as widows, orphans, and pensioners), who are unlikely to
      place their funds with banks or other financial institutions and who have low incomes from other
      sources.

      Certainly, a large number o f pensioners, from both the government and the private sector, widows, and
      people belonging to low-income groups depend o n profits from these schemes. The bulk o f the funds
      invested in NSS instruments, however, come from a relatively smaller group with a relatively large
      average holding o f NSS instruments. Holdings o f this size are not congruent with the mentioned target
      group; in other words, most o f the holdings of NSS instruments are with relatively well-off groups in
      society, and these groups are the ones that receive the largest share o f the indirect income support
      (subsidy)that the government provides through the NSS.

      Byproviding this subsidy through the NSS instruments, the government is diverting scarce public h d s
      from high-priority programs with a much more substantial development impact, such as poverty
      reductionprograms, basic health or education services, or improved infrastructure.

   I

2.46.     Ideally it would be desirable to phase out the use o f NSS instruments and rather find government
deficits domestically by selling PIBs in sufficiently small holdings to households through the existing
network o f NSS outlets. However, phasing out the NSS instruments would meet with substantial public
opposition. Also, the government feels strongly that it needs to balance its social obligations with
economic imperatives. As mentioned above, some o f the recently launched schemes act as safety net to
pensioners, widows and the small savers. Still, the Governmentmust continue with progress on reforming
the NSS while educating the public on the true opportunity cost and problems of the instruments. A key
objective s i s to provide the public with a safe alternative savings instrument with a good yield. In that
regard the Government could make the PIBs more accessible to individuals. The Government i s also
considering CDNS entering into mutual funds to facilitate developinga secondary market for such assets.

2.47.     The successful launch o f the PIBs has not only set a market-based long-term benchmark yield,
but also helpedthe government to finance its budgetary requirements with stable long-term debt.27 Inany
event, the government should make further use o f PIBs, end the on-tap feature o f the NSS instruments


27State Bank ofPakistan Annual Report 2001-2002, p. 136.


                                                         33

and reduce the margins over PIB yields. The yield curve inPalustan has become very steep. The spread
between the yields o f three-month treasury bills and ten-year bonds was 3.9 percent in July 2003
(currently the most recent observation of the ten-year bonds yield). The over reliance on NSS clearly
limits the government's options for relying on the cheaper short end o f the market. Inlight o f the current
low yields on treasury bills (less than 2 percent per year), the govemment could drive down the average
borrowing cost by issuing treasury bills and retiring expensive foreign and-to the extent possible-
domestic debt.

2.48.     The government must do a better job o f educating the public on the high cost o f NSS and its
implications for the national debt problem. It can be made clear that the bulk o f public debt interest
payments go to well-to-do Pakistanis and that unless these payments can be reduced, the domestic debt
problem cannot be solved and poverty programs will suffer. Direct income support programs shown
explicitly in the fiscal budget can address the policy objective o f providing income support to widows,
orphans, andpensioners, ifit i s to be pursued.

2.49.     The discussion here suggests that significant reductions in the real costs o f government domestic
borrowing will be possible only with a lag. The cost o f domestic debt i s currently underestimated in the
fiscal budget. The true cost i s the accrued interest liability. Since the stock o f NSS instruments has been
rising steadily and the rates were also increasing until 1997, the cash-based figure for interest cost as
reported in the fiscal accounts i s substantially lower than the accruals. In other words, every year the
government has been buildingup an unfunded liability that will burden the budget in the future over and
above the accrual cost. This situation reinforces the need to limit fi-esh government borrowing and to
carry out further fiscal adjustment.

2.50.     We also recommend that the DPCO estimate the accrual cost o f the interest liability for the NSS
instruments with a view to reflecting the true cost inthe fiscal accounts. The accrued but not paid interest
could be added to debt obligations. The calculations will also improve the estimates o f future interest
payments likely to be actually paid.

The Cost of External Borrowing

2.5 1.    From a fiscal point o f view, the total real cost of foreign borrowing i s the sum o f the nominal
interest payments in domestic currency plus the capital loss or gain inthe domestic value o f external debt
stemming fi-om depreciation or appreciation o f the currency due to nominal exchange rate changes--all
adjusted for the changes inthe domestic price                          As already mentioned, the total real cost o f foreign
borrowing rose during FYs 1996-99 to 4.7 percent annually, reflecting, among other things, substantial
capital loss due to real depreciation o f the Palustani rupee inthis period. The real depreciation continued
duringFYs 2000 and 2001 but has been sharply reversed since then because o f a remarkable tumaround
inthe current accountbalance ofpayments and improved externalresource availability, which isreflected
in a strong buildup in foreign exchange reserves. For the period FY 1999-2002 as a whole, the average
real cost o f external borrowing for the budget came down to 3.9 percent, reflecting real appreciation inthe
last year and the effect o f debt rescheduling. InFY 2002 and the first half o f FY 2003, the total real cost
o f foreign debt was actually a negative 5.7 percent per year. This shift reflects an average annual real
appreciation against the U.S.dollar o f 7.9 percent and the effect o f debt relief.

2.52.     What assumptions should be made about public external debt costs inthe medium term? H o w i s
the exchange rate likely to behave in the near future? What should be Palustan's external borrowing


*'Iftherei s a real depreciation the valuation element i s a capital loss, and if there is a real appreciation the valuation element i s a
capital gain.


                                                                  34

strategy, keeping in view the high public debt burden on the one hand and substantial strength in the
balance o fpayments on the other hand?

2.53.     It i s difficult to predict the course o f exchange rate movements, but the sharp real appreciation in
FY03 i s not something one can expect to continue over the medium and long term." Still, in view o f the
very strong foreign exchange reserve position, it seems unlikely that the rupee will depreciate in real
terms inthe near future. More precisely, we assume that the nominal changes in the exchange rate inthis
(FY04) and subsequent fiscal years will only reflect the differences between the domestic and
international rates o f inflation. This situation suggests that the real cost o f external borrowing can be kept
down to about 1 percent per year if the mix o f borrowing remains heavily tilted toward concessionary
assistance, as it has been inthe past few years.

2.54.     The case for a continued very soft blend o f external assistance i s linked to the high level o f
external indebtedness. As mentioned, despite the sharp slowdown in the growth o f public and publicly
guaranteed debt and the very favorable terms o f Paris Club re~cheduling,~'the present value o fpublic and
publicly guaranteed external debt to exports i s still well above the desirable cutoff suggested for H P C
countries. Pakistan's case for reducing relative dependence on market loans from multilateral banks such
as the World Bank and the Asian Development Bank and expanding its access to their concessionary
facilities-the     International Development Association (IDA) and the Asian Development Fund (ADF)-
remains strong. In fact, the World Bank has increased its IDA exposure in recent years, from US$3.9
billion in June 1999 to US$5.6 billion in June 2003. However, soft financing outstanding through the
ADF as of December 2002 remainedat the June 1999 levelo fUS$4.2 billion.

2.55.     The prospective stability o f the real exchange rate and the wide gap that is emerging betweenthe
real cost o f public domestic debt (projected at about 5 percent per year during2003-07) and extemal debt
also suggest the desirability o f greater reliance on external financing (as opposed to the most expensive
domestic debt instruments), provided the mix o f external assistance remains heavily concessionary. If
extemal borrowing on appropriate terms can be arranged, it may be feasible to totally eliminate the
growth indomestic debt inreal terms ina few years and thus lower costs dramatically.

2.56.     If the government continues its recent economic policy reforms and maintains a strong track
record o f implementation, which gives it more credibility in the eyes o f its development partners, the
country will have considerable opportunities for concessionary assistance and adjustment lending to
restructure the present expensive debt. Continued strong reform policies will also facilitate substantial
policy and adjustment lending from multilateral institutions. Palustan's challenge will be to develop
strong sector investment and poverty-oriented programs that donors can support through policy-based
quick-disbursing assistance, even though balance o f payments support may be disappearing as a rationale
for adjustment lending. Meeting this challenge would also require the government to increase the general
effectiveness o f public spending and hence increase the emphasis on soundpublic debt dynamics.

2.57.     Related to the cost o f foreign borrowing are the issues o f choice o f currency denomination and
the degree o f interest rate variability on foreign debt. At present, the government has only limited
information on the currency composition o f its external debt and therefore does not manage this risk in
any meaningful way. W e recommend that the DPCO give priority to improving the information base in
this regard and also develop a strategy to reduce the currency and interest rate risks (variability) which
could include use o fmarket-based hedging instruments.




29BetweenSeptember 2001 and December 2002 the real effectiveexchange rate appreciatedby 4.5 percent andthe bilateralUS.
dollarhpee real exchangerateby 12percent.
30See World Bank, Development Policy Review, April 2002, pp. 24-25.


                                                          35

Structural Reforms and Growth Revival

2.58.    The revival o f economic growth i s critical not only for alleviating poverty, but also for
strengthening the debt-canying capacity o f the economy. The debt burden i s determined not only by the
absolute level o f debt, but also by the size o f the economy, the level o f government revenues, and exports.
But as discussed, the highburdeno f debt itself contributed to the slowdowninthe average annual growth
rate o f real GDP from more than 6 percent to less than 4 percent over the last four years. The
government's strong economic reform program aims at exiting the debt trap by pursuing simultaneous
progress on macroeconomic stability and structural reforms. The key elements o f structural reform aim at
improving governance, the health and efficiency o f the financial system, and the climate for private
investment. However, the progress made inpursuingthe reform agenda has only with a considerable lag
helpedrevive the economy: the average growth rate duringFYs 00-02 was only 3.2 percent but picked up
to 5.1 percent inFY03.

2.59.    Several factors help explain the sluggish growth and the pick-up in growth in FY03. The
agricultural growth rate, which averaged more than 4 percent in the 1990s, dropped to 1.6 percent during
1999-2002 mainly as a result o f drought but rose to 4.2 percent in FY03. Gross fixed capital formation
has continued its downward drift; the rate o f fixed investment reached a low point o f 13.1percent o f GDP
in FY02 and FY03. A number of factors continue to hamper industrial growth including structural
problems in industryand exports, continued heavy reliance on cotton textiles, especially cotton yam, and
a lack of diversification to growth areas in international trade in manufactured goods. Decades o f low
investment in human capital are also takmg their toll in terms o f slow growth inproductivity. In spite o f
these challenges that still remain, manufacturing and exports experienced rapid growth in FY03. Key
factors were better access to the EU and U S markets and improved competitiveness o f the domestic
textile ind~stry.~'

2.60.    Some o f the structural problems, such as the low levels o f education and slulls and insufficient
diversification o f manufactured exports, will take time to resolve. But the remarkable improvement in
Pakistan's external resource position in the past three years has fundamentally transformed the prospects
of higher investment without an excessive reliance on external flows. H o w will the strength o f the
external position translate into accelerated growth? The answer lies in the dramatic turnaround in the
national savings rate. During 1996-99the national savings rate averaged little over 12 percent o f GDP as
government current expenditures substantially exceededrevenues and capital flight continuedbecause o f
fears of continued depreciation and mounting public debt. The strong improvement inprivate remittances,
the sharp reduction inthe government revenue deficit facilitated inpart by external official grants, and the
reversal of capital flight pushedthe national savings rate to a record highlevel o f more than 21 percent o f
GDP inFY 2003. Ifconfidence inthe currency can be maintained through a further orderly reduction in
both public and external debt burdens and an adequate level o f foreign exchange reserves, there i s every
reason to hope that the national savings rate can remain inthe 17-18 percent of GDP range. On the fiscal
side, achieving further increases in government savings and total elimination o f the reliance on foreign
grants to finance   current expenditures are goals that appear within reach ina couple o f years.

2.61.    During FYs 2001-03, a significant part o f the increase in national savings took the form o f
addition to foreign reserves---partly involuntary because o f the sluggishness o f private sector investment.
Inthe future the need for increases inreserves will be relatively modest and thus most of the national
savings should be available for gross capital formation. Palustan can also reasonably hope to supplement
its national savings with net foreign capital. The position inFYs 2002 and 2003, when Palustan has had a


31Elimination of the incentivefor under-invoicingof exports after the disappearanceof the premiumon exchange rateinthe kerb
market mayalso partly explainthe stronggrowthinrecordedexports.



                                                             36

sizable current account balance o f payments surplus both with and without official transfers, cannot be
considered the equilibrium. With the further likely improvement in the extemal debt position, Pahstan
can safely runcurrent account balance o fpayments deficits o f 1-2 percent o f GDP, financed with a mix of
grants, net borrowing, and foreign private investment (see Annex B, Table A.3). Thus, at least from the
point o f view o f resource availability, the stage i s set for sustained and strong increases in the level o f
investment. An increase in the fixed investment rate from the low o f 13.1 percent o f GDP in FYs 2002
and 2003 to around 17 percent over the next four years need not be constrained by financial availability.
While public investment should also increase, more o f the increase must come from the private sector
(see Box 2.5).

2.62.   It i s a puzzle why private investment has remained stagnant inthe past few years o f reforms even
though both foreign exchange availability and the liquidity position o f the banks have improved greatly.
Possible explanations are the real interest rate level (until very recently), perceived uncertainty about the
outlook for policy reforms in the run-upto elections in October 2002, and the new political environment.
Although funding costs (nominal interest rates) have come down, real lending rates have remained high
partly because o f very highintermediation costs and lower inflation inthe last few years. Acceleration o f
banking sector reform and further progress on privatization o f state-owned banks i s clearly indicated.
Currently, perceived uncertainty about the continuation o f policy reforms in the new political
environment may have led investors to continue with a wait-and-see position. Hence, it i s important that
the government reaffirm its commitment to reforms that improve the investment climate.




                                                       37

I           Box 2.5: GrowthRevival and PovertyReduction: PrivateInvestmentand Public Spending                           I
 The focus o f this report i s on Public Sector Management issues inPakistan. However, the report has underlined at
 various places the critical importance o f reviving private investment. The baseline scenario assumes that 70
 percent o f the additional investment over 2002-07 would be in the private sector. This assumes that with the
 continued privatization efforts, investment bypublic enterprises would grow little.

                             Pakistan:Investmentinpercentof GDP, baseline scenario "
                                                                    2001/02        2006/07  Difference

                     Total investment                                      13.1        16.7     3.6
                          Public                                            4.8         5.7     0.9
                             Government                                     3.5         5.0     1.5
                              Public enterprises                            1.3         0.7    -0.6
                          Private                                           8.3        11.0     2.7

                     Public investment in% o f total                      36.6         34.1    -2.5
                     Private investment in% o f total                      63.4        65.9     2.5

                     Public investment, Rsbillion                        174.2        333.2

                      I/Grossfixedcapitalformation
                     21Assumes continued privatization of selected public enterprises.

 Nevertheless, our projections provide for a substantial growth indevelopment budget as well as non-development
  spending (excluding interest, defense and SOEs). Development budget i s expected to grow from Rs. 126 billion in
  2001-2 to Rs. 292 billion in current prices, rising from 3.5 to 5 percent of GDP. The growth innon-development
  spending would be slower but still this spending would still account for about 60 percent o f the fiscal space created
  for non- interest non-defense spending.

  H o w would this potential expansion of public spending help the revival o f growth and poverty reduction? H o w
  will this public sector expansion support the critical increase needed in private investment? The institutional,
  allocation, and governance options that will best ensure positive outcomes from public spending and would
  provide necessary support for the private sector are discussed at length inChapters 2 to 5. Only two points should
  be noted here. First, without substantial increase in capacity at all levels o f government and without significant
  improvements in budgetary and planning processes and fiduciary controls, a large expansion o f public spending
  would not be justified. Secondly, the government spending should strictly avoid substituting for private sector
  activities. This can be ensured if the privatization efforts continue and determined efforts are made to promote
  public - private partnerships inthe delivery o f services especially social services.

  With the above caveats, there appears to be substantial need of and room for larger public spending especially in
  human resource development, maintenance of existing infrastructure, and longer term new investments especially
  inwater. The reorientation ofpublic sector prioritiesto facilitate growth andmake the growth more pro-poor are
  discussed inthe next two chapters.



 2.63.     The government needs to continue its efforts to improve governance and strengthen institutions,
 including those in the public sector. This entails reforming government institutions and behavior to limit
 bureaucratic harassment and red tape (especially in the administration o f regulations and taxes). Further
 institutional strengthening and reform i s also needed in the financial sector, where the government in
 recent years has initiated improvements in the regulatory environment and enforced more stringent
 supervision. More broadly across the economy, the government also needs to more strongly assert the rule
 of law, including law enforcement, and reduce corruption and crime. It should also improve the provision


                                                                  38

o f public services such as the quality and availability o f infrastructure (such as power, water, transport,
and telecommunications) and-through          improved education-raise      the quality o f the labor force.
Facilitatingthe availability o f trained and slulled manpower could involve reviewing and improving labor
laws and policies. The current anti export bias in some industries could be reduced by lowering duties on
intermediate products.

2.64.    The challenge for the new government is, on the one hand, to maintain the confidence o f
international financial institutions and private investors, both domestic and foreign, and, on the other
hand, to use appropriate incentives, policy guidance, and institutional support to ensure that development
resources are well directed. There i s a substantial unfinished economic and social agenda, and major
problems remain. But for the first time in many years there i s no threat o f financial crisis hanging over
Palustan's head.

2.65.    The government i s now aiming to revive economic growth to 6 percent per year inthe next three
or four years. Achieving this rate will require, above all, a strong recovery inthe agricultural growth rate
and a sharp expansion o f manufactured exports. Some o f the policy and institutional elements required for
improved performance inagriculture and exports are listed.

2.66.    Inagriculturethere is aneed'for improvedcanal irrigationandmaintenance; better incentives for
water conservation, specifically sharply higher water rates and better management o f irrigation systems;
reduction in adulteration o f seeds, pesticides, and fertilizers through forceful use o f regulatory authority;
and shifting o f economic resources, including scarce water, away from production o f sugarcane to other
crops through suitable price support adjustments.

2.67.    Increased exports can be achieved by undertalung the following measures to improve Palustan's
external competitiveness: raise productivity by buildinghuman capital, upgrading labor skills, and hence
more easily absorbing new technology; facilitate the availability o f trained and skilled manpower by
improving labor laws and policies; improve product quality and the reliability o f delivery (important
contributions would include more reliable power supply and transport systems); reduce the antiexport bias
by lowering duties on intermediate products and possibly establishing virtual free-trade zones; attract
foreign direct investors for export-oriented production, drawing on their established international
marketing channels; focus policy on higher value-added exports rather than volume o f exports and the
emphasis on cotton yam; and prepare for the new international regime in garments and textiles trade in
2005.

Alternative Scenarios

2.68.    To sum up, the room for maneuver inthe public sector i s increasing. It will be possible to expand
public development and social spending at a pace faster than GDP growth, consistent with debt burden
reduction, but only if public enterprise losses are reduced, defense spending i s contained, government
revenues show healthy growth, and donors are willing to provide substantial grants and concessionary
loans. In the medium term the revival o f well-directed private investment will be critical for reviving
growth, revenues, and exports. More rapid increases in social spending and public investment than
envisaged inthe baseline scenario can certainly bejustified on need grounds but will require a much more
forceful resolution o f public enterprise issues, especially inpower, and reduction in domestic borrowing
costs than appears likely at the moment. Given the constraints on public spending, the institutional
implications o f resource allocation and the effectiveness o f spending assume special importance. These
are discussed in subsequent chapters. The improvements in the effectiveness and targeting o f spending
will contribute to a higher growth rate but, as discussed later, will be particularly significant for attaining
poverty alleviation and social development goals. The performance o f social sectors has been poor in the
past, as much because of inadequate spending as because o f poor governance and de facto poor targeting


                                                      39

o f service delivery (the benefits o f the public services have not accrued to the poor but rather to middle-
and higher-income groups).

2.69.    The government should certainly explore the policy implications o f doing better than the
cautiously optimistic baseline scenario presented here. But it should be stressed that the base case
scenario does assume a continued strong reform effort. If reform falters in just a few areas--revenue
mobilization, public enterprise restructuring and privatization, and domestic and external debt
management--it will quickly reverse the trend toward a higher growth rate and compound the problem o f
poverty alleviation. This i s illustratedinthe low case scenario discussed below.

Low CaseScenario

2.70.    The low case scenario assumes slippage in maintaining recent proactive and reform-oriented
policies, inparticular with regard to the fiscal budget, public enterprises, and other structural reforms that
have a bearing on the investment climate. The main differences inoutcomes over the medium term inthis
scenario are worsened fiscal balances, less fiscal space, and slower economic growth compared with the
baseline scenario (see Table 2.6). A more detailed description i s provided inAnnex B.

The main assumptions are
     .   Fiscal revenue collection gradually declines as a percentage o f GDP over the medium term (FYs
         2004-07) from the relatively highlevel in FY 2003. The average revenue/GDP ratio i s projected
     ..  at 16.8 percent o f GDP, substantially lower than in the high-growth scenario (average 17.8
         percent o f GDP).
         Public enterprise-related costs average 1.6 percent o f GDP during FYs 2004-07 as against 1.0
         percent inthe base scenario.
         Domestic interest payments average 4.0 percent o f GDP during FYs 2004-07 compared to 3.6
     .   percent in the base case. This reflects both that domestic debt i s higher in the low case and
         interest rates decline more slowly.
         The Government does not manage to contain defense spending as a share of GDP. It rises
         throughout the projection period, reaching 5 percent of GDP by FY07 and averaging 4.6 percent
         of GDP, compared with 3.9 percent inthe base case.




                                                       40

                  Table 2.6: Summary of key macroeconomic variables in low case, 1999/00-2006/07


                                                                   Est.Prov. Est.                 Proj.
                                                  1999100 2000101 2001102 2002103 2003104 2004105 2005106 2006107
                                                                      (annual changes in percent)
 Real GDP at market prices                            4.2     2.6     2.8       5.8       5.0     4.9   4.9   4.9
 Consumer prices (p.a.)                               3.6     4.4     2.7       3.1       4.1     4.1   4.2   4.2

                                                                          (in percent of GDP)
 Savings and investment
    Gross national savings                           14.1    13.6    17.4      21.3      15.1    15.3  14.7  14.3
    Gross fixed capital formation                    14.4    13.9    13.1      13.1      13.7    14.1  14.3  14.4
    Change instocks (nongovt)                         1.6     1.6     1.6       2.4      2.0      1.6   1.6   1.6

 Public finances
    Revenue (including grants)                       17.3    17.3    19.3      20.8      18.1    18.3  18.0  17.7
     Grants                                           1.1     1.2     2.3       2.8       1.0     1.4   1.4   1.3
     Revenue                                         16.3    16.2    17.0      17.9      17.1    16.9  16.7  16.5
     Expenditure                                     22.8    21.4    23.6      22.3      22.2    22.4  22.4  22.4
       Interest payments                              7.8     6.8     6.8       5.2       4.8    4.8    4.6   4.5
       Defense                                        4.8     3.8     4.1       4.0       4.1     4.4   4.7   5.0
       SOE-related spending "                         0.5     0.6     1.9        1.2      2.0     1.8   1.7   1.5
      Non-interest, non-defense, non-
            SOE spending                              9.8    10.2    10.9      12.0      11.4    11.5  11.4  11.3
     Overall balance (including grants)              -5.5    -4.1    -4.4      -1.6      -4.1    -4.1  -4.4  -4.6
     Primarybalance (including grants)                2.3     2.8     3.8       3.6       0.7     0.7   0.3  -0.1
     Grosspublic debt   ''                          100.9   108.3 100.1        91.8      83.6    81.5  78.9  77.9
     Grosspublic debt in% o frevenue                619.6 670.5 588.9         512.0 488.8 482.5 473.0 473.2
  Extemalsector
     Current account incl. off. transfers            -2.0    -1.9     2.7        5.9     -0.4    -0.4  -1.2   1.7
    Ext.public andpublicly-guar.debt                 48.6    55.7    52.7      45.8      38.1    35.0  32.5  30.4
     Gross reserves inmonths o f next
      YearimPofisofgoodsandservices                   3.8     1.7     3.7        5.4      4.4     4.4   3.5   3.5
  1/ Subsidies andpayment of contingentliabilities.
  21Excludes SBP.


2.71.     Consequently, the fiscal deficit grows to an average o f 4.3 percent o f GDP during FYs 2004-07
instead o f declining to 3.0 percent as in the base case. The average fiscal space for economic and social
spending increases only by 0.6 percent o f GDP, compared with 1.9 percent inthe base case.

2.72.     The policy slippage leads to less strengthening o f confidence. Hence, the reduction in interest
rates on domestic public debt i s slower than in the base case. There i s also less foreign financing-both
grants and concessional loans--reflecting less confidence by donors and IFIs in the government's
commitment to pursue reforms and address the social and public investment needs. Progress on
privatizing public enterprises i s slower, and the projected sales receipts are lower than inthe base case.

2.73.     Total public debt i s higher and with less concessional terms on average. By FY 2007 public debt
as a share of annual fiscal revenue declines only to about 480 percent, whereas inthe base case it falls to
about 390 percent o f revenue. In other words, progress toward reaching public debt sustainability i s
insufficient inthis scenario.


                                                              41

2.74.   Furthermore, economic growth, job creation, poverty reduction, and improvements in social
indicators are slower than in the base case. GDP growth averages 4.9 percent per year during 2004-07
compared with 5.6 percent in the base case. The slower growth reflects lower private investment and
lower growth-related public spending on health, education, and infrastructure-both            operations and
maintenance and new investments-and             hence lower productivity o f public capital stock and human
capital. It also reflects slower improvements in the effectiveness o f govemment spending, govemance,
and implementation capacity.

2.75.   The differences described inthe fiscal outcomes between the scenarios are even more accentuated
when one looks at the momentum at the end o f the projectionperiod. By FY 2007 the fiscal deficit grows
to 5.9 percent o f GDP (from 4.5 percent in FY 2003) instead o f declining to 4.3 percent as in the base
case. Inspite o f the larger deficits, the fiscal space for economic and social spending increases only by 0.4
percent o f GDP by FY07 compared to 3.4 percent in the base case. Over the longer term the base case
would continue to show improved debt reduction, growth, and social indicators, and the gap between the
two cases would continue to widen. The low case would basically show stagnation.

                      Table 2.7 Pakistan: Low case scenario,PrincipalBudgetAggregates
                                    Change FY2002-FY2007 (Percent of GDP)


            Resources                               Expenditures

            Government revenues            -0.5     Defense expenditure                       +0.9
            External grants                -1.O     Interestpayments                          -2.2
            Fiscal deficits                +0.3     Public enterprise losses 1/               -0.4
            (including grants)
                                                    Non-interestnon-defensespending           +0.4
                                                        olw Development spending               -0.7
            Total resources                  -1.3   Total expenditures                         -1.3
           l/ enterpriselossesareapproximatedbysubsidiesandpaymentofcontingentliabilities.
              Public

High-Growth Scenario

2.76.   A scenario with more rapid economic growth and more fiscal space than in the base case is
certainly possible. Achieving this outcome will, however, require three things: First, the government
must make a determined effort to lower the real cost o f domestic borrowing. Lowering the real interest
rate from 6 percent to 4 percent by FY07 will save interest payments equivalent to 1 percent o f GDP.
Second, it must virtually eliminate public enterprise losses and privatize these enterprises more
successfully, especially in the power sector. Third, it must improve the effectiveness o f govemment
spending and institutional capacity to better utilize the additional fiscal space on poverty- and growth-
related spending. Ifmarked improvements in institutional capacity do materialize, spending on health and
education as well as on infrastructure-both O&M and new investments-should increase. Over the
medium and long term, this spending could contribute to higher productivity o f the public capital stock
and human capital (and improved social indicators), further boosting economic growth.

2.77.   In the event, however, that the effectiveness of govemment spending, govemance, and
implementation capacity does not materially improve even with higher economic growth and revenue
collection, the govemment should use the additionalresources available to repay those debt items with the
highest cost attached (currently long-term domestic debt instruments) to at least capture the gains o f
saving future debt-servicing costs.




                                                         42

Conclusions

2.78.      Pakistan i s at an important economic turning point. The debt problems that have been a major
cause o f the sharp slowdown in Pakistan's economic growth over the past decade, and the consequent
stagnation in poverty reduction, have been brought under control. Although debt, especially public debt,
i s still at an unsustainable level and needs to be brought down steadily, the risk that economic and other
shocks will lead to a worsening o f the public debt burden in the medium term appear minimal (see the
sensitivity analysis in Annex B).N o doubt, there i s a substantial unfinished economic and social agenda,
and major problems remain. But for the first time in many years, there i s no threat o f a financial crisis
hanging over Palustan's head.

2.79.      Looking ahead, the issue i s not so much whether macroeconomic stability can be maintainedand
orderly reduction indebt burden achieved, but whether public expenditure management and restructuring
can play a more positive role in the economy than in the past decade. A financial equilibrium at a level
and pattern of public expenditures that does not support provision of essential public goods and does not
further social goals will hardly serve Palustan's long-term interests.

2.80.      With continued reform aimed at increased mobilization o f government revenues, a sharp
reduction inthe losses o f public enterprises, a reduction in the costs o f domestic government borrowing,
and defense spending not exceeding 4 percent o f GDP for the near future, it should be possible to create
substantial fiscal space to further the government's objectives o f reviving growth and reducing poverty
while continuing to make progress on reducingthe debt burden.

2.81.      This chapter illustrates essentially two economic and fiscal scenarios and discusses, at some
length, the policy issues and options involved in bringing about a shift in public spending toward
development and social spending. Table 2.8 presents a summary overview of the main differences
between scenarios in the size o f the fiscal space available for non-interest, non-defense and non-SOE
spending and the elements that contribute to different outcomes.

                                Table 2.8: Summary of FiscalSpaceAlternatives over the mediumterm
                                                                                              Fiscal space altemativesl'


                                                                   Reference:
                                                                2001102 estimate        LOW   case             Base case

 Fiscal space in % o f GDP inFY07                                       10.9               11.3                  14.3
 Fiscal space in% o f GDP - average FY03-FY07                           10.9               11.5                  12.8
 Fiscal space - average annual real growth fromFY03 to FY07                                 6.3                  11.7
 Real fiscal space inFY07, in constant FY02 prices, Rs billion         395.1              531.4                 684.0
 l/Fiscalspaceisdefinedasspendingexcl.interest,defenseandpublicenterpriselosses. Thelatterisapproximatedbythesumofsubsidiesand
 payment of contingent liabilities.




2.82.       Inthe baseline scenario real non-interest, non-defense and non-SOE spending expands robustly
over the medium term, and this expansion, together with stringent measures to ensure improved
effectiveness and targeting o f the spending, should enable the government to facilitate economic growth
and make a substantial dent in poverty, markedly improve health and education indicators, and hence
narrow Palustan's substantial social and gender gap. Non-interest, non-defense spending under the
baseline scenario grows on average by 11.7 percent per year inreal terms, be on average 27 percent larger
duringFYs 2004-07. Inthe low scenario this growth is only 6.3 percent, and thus there is little room for
reversing the relative decline in economic and social spending witnessed in the 1990s. In the baseline



                                                              43

scenario non-interest, non-defense and non-SOE spending to 14.3 percent o f GDP, compared with only
10.9 percent inFY02, and recovers to levels seen inthe 1980sand early 1990s.

2.83.   Inthe low scenario, government revenues decline gradually as a share of GDP over FYs2004-07,
domestic interest payments remain at the FY03 level, public enterprise losses remain high, and defense
spending rises steadily relative to GDP. The interesting point i s that even with a higher fiscal deficit,
compared with the baseline scenario, the fiscal space under the low scenario i s only a fraction o f what
could materialize under the continued reform scenario.

2.84.   It i s crucial for successful implementation o f the baseline scenario that governance improves
sharply. This means that government should combat corruption, leakage, and deficient revenue collection
and control expenditures resolutely and assertively. The benefits will be seen in larger fiscal space to be
used for high-priority spending, a reduced debt burden, rising investor confidence, and hence increasing
private investment-both domestic and foreign-and growth. In addition, the government i s more likely to
get a higher share o f concessional foreign financing if strong measures to improve govemance are part o f
an overall reformpackage.

2.85.   A firm commitment by the government i s needed to stay the course with regard to economic and
governance reforms inthe face of political opposition, including from vested interests and inparticular in
the area o f public enterprise reform. If the government can obtain the projected gains for another 3-4
years, the fiscal and external dynamics could be entrenched on a pathtoward sustainability.

2.86.   Pakistan's development partners have a crucial role to play in providing sufficient foreign
financing in the form o f grants or loans on concessional terms. Without such financial assistance in the
magnitudes envisaged here, it will be very difficult for the Government o f Palustan to achieve the three
central goals (debt burdenreduction, acceleration in economic growth, and poverty alleviation) all on its
own. Hence, substantial additional foreign financing on concessional terms i s an integral part o f this
reform and development strategy.




                                                    44


                  CHAPTER REORIENTING PUBLIC SECTOR PRIORITIES
                                 3:

Introduction

3.1      The previous chapter outlined the fiscal space that can be created and enhanced by the
continuation o f reform efforts. This chapter and the next discuss the issues involved in making optimal
use o f the possibilities o f higher economic and social spending. The effectiveness o f public expenditures
has several dimensions:

    1    Strategic approaches to public interventions
         Choice o f sectoral and subsectoral priorities
    1    Soundness, transparency, and accountability o f budgetary and planningprocesses
    1    Careful screening o fprojects and speedy and effective implementation o fprograms
    1    The existence o f strong monitoring and evaluation systems

3.2      The presence o f hnctioning institutions and a framework o f defined distribution o f
responsibilities at various levels o f government are the cross-cutting issues that influence all aspects o f
effectiveness o f public spending. This situation requires not only the statues, laws, and rules that
determine the establishment and functioning o f various institutions, but also effective management and
leadership so that each institution has the flexibility to perform its statutory duty in collaboration and not
incompetitionwith others andto adapt itsroles according to changing needs.

3.3      This chapter focuses principally on the intersectoral and subsectoral priorities, though obviously
the issues o f project choice, budgetary and planning processes, and implementation cannot be entirely
separated from sectoral allocations and priorities.

3.4      It is clearly not feasible to undertake a comprehensive sector-wise review o f public expenditures
inthe context of this overview report. The attempt here is to examine, in some detail, a few key sector
allocation and implementation issues that will greatly influence the ability o f the government to achieve
its stated goals. A look at the selected sectors can also illustratethe constraints public spending has faced,
the less than optimal way this spending has been used to advance economic and social policy goals, and
the institutional weaknesses that have contributed to often disappointing and occasionally adverse
consequences o f public spending for the economy and the society. Even more important, looking to the
future, such an analysis can suggest ways for minimizing the public policy dilemmas (and increasing
policy choices) arising from the extremely constrained financial resource position on the one hand and the
desperate need on the other hand to step up spending on education, health, infrastructure, and direct
poverty alleviationprograms.

3.5      In reviewing the government's expenditure priorities and options in allocating the fiscal space,
this chapter examines the entire public expenditure inthe selected sectors and not just the development
spending. The notion of development spending, including as it does the investment expenditures,
technical assistance, and some initial recurrent expenditures related to new capital spending, may have
outlived its usefulness as a sectoral and nationalplanning tool. Indeed, to some extent, the impliednotion
that nondevelopment spending i s o f lower priority has led to an inefficient use o f economic resources, It
has contributed to inadequate allocations to the social sectors, woefully inadequate allocations for
operation and maintenance, and the creation o f new physical assets at the cost o f the effective use of
existing assets. It has also led planners to be preoccupied with the development budget while neglecting
sectoral analysis and the impact o f so-called non-development spending on the country's economic and
social goals. The introduction o f LTRSP and the Medium-Term Budget Framework (MTBF) provide

important opportunities for moving away from the excessive focus on the development budgets and
toward a more comprehensive approach to public expenditure management.

3.6       Using this comprehensive sector-wise approach, the discussion here focuses on allocation and
implementation issues inthree sectors that will greatly influence the ability o f the government to achieve
       ...
its stated goals:

           Irrigation water
           Power
           Education and health

3.7       Public spending on a host of other sectors, such as law and order maintenance and provision o f
justice, agriculture, roads, population welfare, rural development, and targeted transfers (food support
program, Zakat, etc.) no doubt also deserves priority. The government includes many o f these (inaddition
to education and health and a part o f irrigation spending) in the list o f high-priority spending to be
monitored under the I-PRSP. The costing o f PRSP-related expenditures and relationships between
expenditure inputsand output targets are as yet extremely tentative, but the government has rightly given
the protection o f I-PRSP spending the highest priority. Issues related to the adequacy o f funding o f I-
PRSP cannot be analyzed in a comprehensive fashion in this report because o f a lack o f costing data.
They are touched upon, however, in the next section and discussed at some length in the section on
education and health. Since the social sectors account for 60 percent of identified high-priority spending
for poverty reduction in the I-PRSP, there i s a considerable overlap between the latter and the sectors
selected here. Inany case, the issues o f the coverage and adequacy o fhigh-priority poverty spending must
be revisited in the context o f the finalization o f the PRSP. This spending should have the first claim on
the fiscal space that will be createdby the continuationofreform policies.

Fundingthe PovertyReductionStrategy

3.8       The Government o f Pakistan i s following a multiprong poverty reduction strategy, which i s
outlined in the IRRSP and will be further elaborated in the full PRSP now under preparation. The core
principles o f the strategy include engendering growth, implementing broad-based governance programs,
improving social sector outcomes, and reducing vulnerability to shocks. The main macroeconomic targets
inthe I-PRSP that is, growth rates of GDP, revenue, and inflation--are broadly inline with the baseline
scenario presented in Chapter 2.      Two questions related to public expenditures are relevant for the
implementationo fthe poverty reduction strategy:
          .Are the indicated YPRSP expenditures consistent with the achievement o f specified goals
           especially in the social sectors and with the internationally agreed-upon Millennium
          .Development Goals (MDGs)?
           Can the financing needed for the poverty reduction strategy in the medium term be
           accommodated within the fiscal space likely to be created?

3.9       The high-priority expenditures for poverty reduction identified in the VPRSP show a relatively
modest planned growth, from 3.4 to 4.0 percent o f GDP, over the period 2001-04. This allocation appears
to be totally inadequate inrelation to ambitious targets, especially in education. On the one hand, there i s
a needto set more realistic but still ambitious social targets consistent with MDGs. On the other hand, the
financing needs for meetingthese targets require more careful assessment.

3.10      As discussed later in this chapter, achieving goals in education and health will require
substantially larger funding than assumed at present. The full PRSP will probably extend the planning



                                                       46

horizon to FY 2006 and i s expected to provide a fuller costing o f social sector goals so that inadequate
fundingdoes not frustrate the essential goals o fhuman development, as often happenedinthe past.

3.11     It i s suggested that the high-priority PRSP spending also cover law and order maintenance and
provision o f justice and rural electrification. The governance problems resulting from inadequate
spending on police and the justice systemhave their greatest impact on the poor. Similarly, it i s the poor
inthe rural areas who most lack access to electricity; of the roughly 40 percent of Pakistani households
that do not have electric connections, most live inthe rural areas and are poor. Speedy rural electrification
must be an important part o f increasing access to education and health care, dissemination o f new
technologies, and general modernization. Rural electrification cannot be undertaken as a commercial
proposition and will require special public sector financial support. A system o f capital grants to rural
communities and electric cooperatives can be developed that gives priority to those who mobilize enough
own resources to ensure financial viability after the initial grant. It i s worth looking into Bangladesh's
experience. Donors will probably show substantial interest in providing grants or low-cost loans to a
program whose benefits go overwhelmingly to the poor. The development and expansion o f the rural
electrification network should not be limitedto the Water and Power Development Authority (WAPDA)
and should involve the private sector.

3.12     Using the broader definition o f poverty-related spending suggested, the actual priority poverty
spending in 2001-02 was close to 5 percent o f GDP. Our rough estimates suggest that this spending
should increase to about 7 percent o f GDP by 2006-07. Higher education and health spending would
probably take up half o f the increase in order to achieve the goals set; the rest would go largely to higher
allocations for law and order, provision o f justice, and rural and urban infrastructure, including rural
electrification. The baseline scenario, which projects that non-interest, nondefense spending could
increase from 10.9 to 14.3 percent o f GDP over the period 2002-07, can accommodate the rise in high-
priority poverty-related spending o f the order o f magnitude suggested here, provided the political
commitment to reorient the public spendingtoward social goals and better governance remains strong. It
i s very encouraging that the recently completed PRSP (December 2003) provides both for a broader
definition o f poverty spending suggested above and proposes to increase this spending steadily to 6.8
percent o f GDPby 2007-08.

Remedyingthe Neglect ofEducation

3.13     Palustan's I-PRSP recognizes the importance o f improving education outcomes. Learning from
the experience of the Social Action Program (SAP), the government strategy i s to significantly improve
the social service delivery system by addressing the underlying governance problems. It will then
gradually increase funding as allowed by the impact o f improvements in tax administration, stronger
economic growth, and assistance from donors. In education, the linchpin o f this effort i s the federal
government's action plan in support of education sector reform (ESR), whose implementation lies with
the provincial and district governments.

3.14     Educational attainment and health improvements are among the most valuable benefits o f human
development and are also--since they constitute vital investment in human capital--important
determinants o f growth and poverty reduction. Pakistan has seriously neglected investment in human
capital and has paid the price for it not only inpersistently highgrowth inpopulation for a long time, but
also inthe slowing o f growth and persistence o f poverty. Ineducation, the problems are the very low level
o f enrollments not only at the primary, but also at the middle, secondary, and higher education levels and
the low and declining quality o f public education.

3.15     The irony i s that an increased emphasis on elementary education and protection o f its relative
share in public spending in the 1990s did not achieve the desired (primary and middle) results, whereas


                                                      47

per-student spending at the secondary and higher education levels declined by as much as 30 and 50
percent respectively inreal terms, compounding the crisis o f quality. As it moves ahead, Pakistan needs to
make a major dent in the social gap at all o f levels o f education and accompany it with a demonstrable
reversal of the trend in declining quality. The approach to remedying the social gap needs to be
comprehensive, setting ambitious but attainable targets and backmg them up with adequate funding, clear
institutional responsibilities, and a system o f incentives and workable public participation arrangements
including private-public partnerships, citizen boards, and parent-teacher associations.

3.16     The basic strategy embodied in the ESR i s sound. It sets explicit targets for each level o f
education (Table 3.1) expected to be realized by provincial and district governments and aims to achieve
these targets through increased funding, improved resource mobilization and utilization, and the
introduction o finstitutional reforms at all levels o f education to improve governance.

                                 Table 3.1: ESR Targets by Sub-sector, 2001-2005
                                                                   2001 Actual      2005 Target
                                                                        ("A)            (%I
                             Literacy                                    49.0              61
                             Gross Primary Enrollment                    83.0            100
                             Net Primary Enrollment                      66.0              76
                             MiddleSchool Enrollment                     47.5              55
                             Secondary School Enrollment                 29.5              40
                             Higher Education Enrollment                  2.6               5
                           Source: Governmentof Pakistan,Ministry ofEducation's EducationSector
                           Refoms documentbasedonprovincial Governmentdata. PIHSdatahas differences.

3.17     As already noted, however, several parameters o f ESR, notably the specific targets, funding
requirements, and financing mechanisms, need to be revisited. More important, the education sector plans
needto be made more operational inthe context o f finalizing the PRSP, usingmore recent information on
actual enrollment levels and the decision to allow district governments more freedom in choosing
expenditure priorities.

3.18     The ESR targets are ambitious and will entail considerable commitments from the government-
for example, gross primary enrollments are targeted to increase by 16 percentage points fkom 2001 to
2005. Tahng into account the populationgrowth, meeting this target would require enrollment o froughly
an additional 5-6 million children at the primary level. Even this figure understates the problem. There i s
lack o f agreement on the starting level o f gross primary enrollments: P M S data suggest that current
primary gross enrollments may be as low as 72 percent compared with a figure o f 83 percent from
government administrative data (Table 3.1.) Similar discrepancies may exist at other levels o f education.

3.19     The ESR had projected a cost of Rs. 100 billion, with roughly 34 percent o f the total to be
allocated to primary education,32 over and above the present annual national expenditures on education. I-
PRSP, on the other hand, assumes a growth inpublic education spending in line with growth innominal
GDP: the percentage of expenditure on education remaining at 1.8 percent o f GDP. Figure 3.1 illustrates
the existing difference inprojected expenditures between I-PRSP and ESR.


32The original ESR package was prepared at a cost of Rs. 55.5 billion for the years 2001-04. The duration of this packagehas
been extended to 2001-05 to accommodate the President's Programs, the mainstreaming of Mmadaris, and the setting up of
polytechnicsat the Tehsil level. Therefore, the cost estimateshave increasedto Rs. 100billion. These estimates,however, do not
include recurrent expenditures of the ESR program. The districts and provincialiareagovernments are expected to review and
assess their own requirements and adjust the non-development budgets. ESR estimates also do not include the provincial
educationAnnual DevelopmentPrograms(ADPs).




                                                              48

3.20     The    bulk o f       the    ESR
expenditure at the primary level i s                  Figure3.1: EudgetaryExpenditureon Educaation
allocated to improving and expanding            1"
rundown physical infrastructure, with           ioqmo
additional support for institutional             sqmo
changes, such as the development o f
parent-teacher associations.         Other
features of the overall strategy include
teacher training,      an adult literacy
campaign,     mainstreaming       o f  the
madarassas, a pilot nutrition program,
introduction o f a technical stream in
secondary      schools,      reform     o f
examinations, and introduction o f a
national education assessment program. With the devolution o f authority to the district level for the bulk
o f education spending, the funding flow arrangements for supporting the objectives o f ESR are also not
clear.

3.21     As already mentioned, the finalization of PRSP during the next few months provides a good
opportunity to resolve some o f the issues and contradictions noted.

3.22     First o f all, the targets for elementary school enrollment need to be reviewed. The 100 percent
gross primary enrollment target by 2004 (I-PRSP) or 2005 (ESR) i s no longer realistic. On the one hand,
firmer information needs to be developed on the present position (at all levels o f enrollment), and on the
other hand a credible universal education attainment target (say by 2006 or 2007) need to be developed
and backed up by a policy package, funding, and institutional arrangements. The ESR middle and
secondary school and higher education targets will also need to be moved out in time by a couple o f
years. Although all education sector goals appear ambitious in terms o f the effort required to achieve
them, they remain modest by international norms. All countries in South Asia except Palustan have gross
enrollment ratios o f 100 percent or higher. The ESR, if successfil, will achieve a middle school
enrollment rate o f 55 percent by 2006 or 2007. In Korea three-fourths o f the relevant age group was
completing middle school in 1975. Pakistan hopes to raise its secondary enrollment rate to 40 percent
over the next four to five years. Malaysia had achieved this rate by 1974. Inhigher education Pakistan has
an enrollment rate (1996) o f 2.6 percent. This i s one o f the lowest proportions inthe world--for India the
figure i s 6.2 percent (1990); for Iran (1994) it i s 12.7 percent, and for several East Asian counties this
rate exceeds 30 percent. Ineducation Palustan appears to be at least thirty-five to forty years behind the
East Asian countries and ten to fifteen years behind the South Asian countries. Consequently, an
acceleration i s needed to enhance enrollment levels at all levels o f education. This goal will require the
government to increase resources allocated to education and to improve the institutional capacity to use
the additional resources effectively and efficiently.

3.23     In the very short term, highpriority must be given to primary education, including nonformal
education, because without rapid universal education, progress toward increasing literacy levels, now
estimated at 45-48 percent, will remain slow. Expanding public expenditure for primary education i s also
crucial because it i s more pro-poor and also helps the rural areas more than the urban areas (see Box 3.1).
Buthigh-qualityprimary (and secondary) educationrequires parallel development o fhigher education. It
is from this level that the country gets its teachers, scientists, and managers.

3.24     The lessons learnt from the experience o f the past decade, in which expansion o f elementary
education spending resulted in only moderate benefits, are already being reflected in approaches to
elementary education. The provincial governments o f Sindh and Punjab are no longer giving priority to


                                                       49

the opening o f new schools. As emphasized in the ESR, attention i s beinggiven to improving quality by
hiring new teachers, increasing accountability o f teachers, improving physical facilities in schools, and
raising nonsalary spending. InPunjab special salaries for teachers (and doctors) in rural areas are being
considered. The Human Development Foundation set up by the Government o f Palustan as an
autonomous corporation i s emphasizing enrollment processes, nonformal education, especially for recent
dropouts, and innovative approaches to capacity building. The Pakistan Poverty Alleviation Fund has
expanded its mandate to include education and social sectors. But it i s not at all clear that the programs,
new initiatives, institutions, and fundingcommitments are inplace on a scale necessary to achieve the key
targets in primary education: 35-40 percent growth inprimary enrollments over the next five years, the
narrowing o f gender and geographical differences, significant improvements in quality, reduction in
dropout rates, and expansion ofprograms o fnon-formal education.

                       Box 3.1: The Rich-Poor Gap andPublic Expenditureon Education

      According to Figures 3.2 and 3.3, income groups from the second decile to the fifthdecile receive the
      highest benefits from public expenditure o nprimary education. Inthis sense, public expenditure for
      primary education is pro-poor - an effect that is partly attributable to the fact that poor children are
      more likely to attendpublic primary schools, as opposed to higher income groups who have a higher
      private-public enrollment ratio. Onthe other hand, income groups from the seventh decile to the ninth
      decile receive the highest benefits from public secondary school education - an effect that arises
      mostly from the fact that poor children are far less likely to enroll insecondary school.


                                                                          Figure3.3: Distribution of Public
                           Incidenceof Primary and                           Expenditure for Primary and
                             SecondaryEducation                                     SecondaryEducation



                    zw                                                   14
                                                                         12
                    150                                                  10

                    1w                                                $  8


                     50

                      0 4 ,    ,    ,    ,    ,    ,  ,  ,    ,  1       0       ,    ,    ,  ,    ,    ,    ,    ,   ,
                        1   2    3   4     5    6   7  8   9   10             1     2    3  4  5    6    7    6     Q  10
                       Per Capita ConsumptionExpenditue                              Per Capita Conumption
                                          deciles                                      Expendituredeciles

                             / p r i m a r y ~  n   d a r Iy                        / p i m a w p      ~    a     r   y


       The Urban-Ruralgap andpublicexpenditureon education
      As Figure 3.4 shows, rural communities, on average, receive more benefits from public primary
       education than urban communities. It is, however, worth noting that poor urban families receive more
      benefits from public primary education than urban communities. As for secondary school education,
      urban communities receive more benefits from public secondary schools than rural communities

                       Figure 3.4: BenefitIncidenceof
                       publicprimary education (Urban                   Public SecondarySchool(Urban vs
                                    vs Rural)                                               Rural)
                                                                                                                          I
                   2%                                                2501

                   m0

                   1%                                                150

                   100                                               IW
                   Ea                                                 50 -
                    0                                                 0 1    ,     ,    ,   ,    ,    ,    ,    ,    ,  A
                       1   2    3   4     5    e   7  8  9  10            1    2     3    4   5    5    7    8   9  1  0
                      Per Capita ConsumptionExpenditure                 Per Capita Consumption Expenditure                  1
                                      deciles                                               deciles                       I

     L
                                                                  50

3.25      Inadditionto issues relatedto the devolutionprocess (discussed indetail inChapter 4), three sets
o f issues appear especially important. First, enrollment levels in Pakistan are among the lowest in the
developing world, and gender disparities and rural-urban gaps33are among the widest and point to
tremendous unrealized potential and missed opportunities. But there are several incentive programs that
have been used around the world to stimulate demand for education and encourage enrollments. Second,
the private sector i s the most dynamic segment of education.34It accounted for more than 40 percent o f
the growthinprimary school enrollment inthe past decade. The average teacher salary ina private school
i s actually lower than in public schools, but the quality o f teaching appears to be better because o f much
greater accountability. Effective public-private sector partnerships that tap into the strength o f the private
sector have yet to be developed. Last but not least, more detailed and realistic estimates o f public funding
requirements, not only for primary education but also for the entire education sector, are needed.

3.26      Clearly increasing public funding alone will not meet the urgent challenges in the education
sector. But higher public spending i s a necessary if not sufficient condition for improving educational
attainments. Without additional funding, devolution will not work, needed quality improvements,
especially at the primary level, will not materialize, the scope for incentive schemes will be limited, and
the neededrecovery inhigher and secondary education will not take place. This situationwill once again
frustrate the attempt to achieve the ambitious targets. Detailed work i s necessary to firm up the education
funding requirements, including for the secondary and higher levels. Our tentative estimates suggest the
need to gradually raise public education expenditures from 1.8 percent o f GDP in2001-02 to 2.5 percent
in 2006-07. This growth will translate into average spending of 2.2 percent of GDP over the next five
years, malung possible an increase o f about 45 percent in real terms over the 2001-02 level. It will also
mean that education spending will take up more than one-third o f the fiscal space generated under the
baseline scenario. A substantial part of the growth inrelative spending will be absorbed by primary and
secondary education, though the downward trend in allocations for higher education also needs to be
reversed.

3.27      Within the recommended overall increase ineducation expenditures, a more rational allocation o f
spending i s needed. More rational allocation would include increasing resources for nonsalary recurring
expenditures for primary and secondary school improvements; provision o f missing facilities in existing
infrastructure, especially at the primary and secondary level; provision o f quality services such as teacher
training; increasing resources for new infkastructure where there i s a need, that is, at the middle and
secondary levels o f education; girls' incentive programs; and demand-side interventions.




33  In 1999, for example, female enrollment rate was 23 percent lower than that for males. As in the case of gender, there are
significant disparities in educational attainment across regions. Most striking i s the urban-rural gap. Unlike the gender gap, the
urban-rural gap has widened over the years. In 1991 primary school enrollment in the rural areas was 22 percentage points lower
than the urban areas, by 1999 it had increased to 30 percentage points. Furthermore, gender gaps in school enrollments vary
substantially across regions and between urban and rural areas. In 1999 the primary schooling enrollment rate was more than 99
percent in urban males o f the Punjab but barely 28 percent for rural females o f Sindh. The gender gaps in these enrollment rates
ranged from less than 3 percent in urban areas o f the Punjab to about 34 percent in rural areas o f the North West Frontier
Province (NWFP).

34 According to a recent Census o f Private Schools (2000), Pakistan has as many as 40,000 o f such private schools, the result o f
significant growth inprivate schooling over the 1990s. The share o f the private sector intotal national primary school enrollment
increased from 14 percent in 1991 to about 28 percent at present. The private share o f secondary schooling increased from 8 to
17 percent over the same period. This phenomenon was not limitedto the higher-income groups. The private share in primary
school enrollment grew from 5 to 10 percent for the lowest-income group, while it grew from 35 to 60 percent for the highest-
income group. Research on private schooling, based on data from the aforementioned census, also shows that private schools
seem generally to offer better gender outcomes, partly because they are coeducational institutions, and also have been successful
inrecruiting female teachers.



                                                                 51

3.28     Incentive schemes can be used to promote enrollment in both government and private schools.
Voucher schemes, which directly support the students, can help create demand for schooling by meeting
some o f the direct and indirect costs o f attending school.         They can introduce school choice and
competition between schools (reducing unit costs and improving quality) and can be targeted to
subpopulations that are underserved or excluded (such as the poor or girls).               Other demand-side
enrollment incentives have been widely used around the world and take many forms-free               textbooks,
uniforms, transport, scholarships (Bangladesh has a very successful female secondary school scholarship
program), and in-kmdpayments to families. Palustan has some experience with these schemes, such as
free textbooks inBalochistan and Sindh and some form o f scholarships inalmost all provinces. The Food
for Schoolgirls program run by the World Food Program provides inducement in the form o f a tin o f
edible oil to each student if she attends school for twenty days in a month. Initial results are encouraging
and point to the wide potential that such schemes hold for improving enrollment and retention rates. The
government i s also financing a food support program (Taiwana Pakistan) that should be evaluated after a
year of implementationto assess its impact on enrollment and retention.

3.29     While raising the level o f public spending on the social sectors i s an important medium-term goal,
and two o f the four provinces (Sindh and NWFP) have economic reform programs underway that hold
promise for more spending for the social sectors, an additional concern i s achieving improvements in the
efficiency o f present-day spending. Getting budget allocations released in a timely fashion to front-line
spending agencies, as well as ensuring that allocations are appropriately balanced across salary,
nonsalary, and the development budget, i s a top priority. Adequate provision and timely release o f the
nonsalary recurrent budget i s critical to smooth operations at the point o f service provision. While
increased budget allocations are a necessary step, it should also be recalled that in the past actual
expenditures have often fallen below planned expenditures-for         example, in Punjab actual expenditures
have fallen short by as much as a quarter inrecent years. Institutional bottlenecks that have disrupted the
overall flow o f funds inthe past are likely to be even more critical with the devolution o f service delivery
to the districts, and the removal o f these bottlenecks i s clearly a high-priority area for reform. According
to anecdotal information, the flow o f funds, especially nonsalary funds, continues to be unpredictable
under the district governments. Predictability o f fundingi s crucial, whatever the absolute level o f funding.

3.30     As a part o f increasing public expenditure on education, the government needs to explore new
funding mechanisms to increase enrollment, raise literacy, and improve quality. These mechanisms
include the Education Foundations, the Pakistan Poverty Alleviation Fund, the Human Development
Commission, and education nongovernmental organizations (NGOs), such as Bunyad, which are active in
the nonformaleducation sector.

3.3 1    H o w can public-private partnerships ineducation be strengthened? One obvious way i s to provide
vouchers for students, especially girls, where there i s no public school nearby (for other possibilities o f
public financing o f private education (see Chapter 4). This seems quite practical because in a number o f
cases private school fees are not high.35Access to educational facilities i s an important determinant o f
school participation. Physical proximity to a school appears to be quite important and i s a factor in
explaining the gender gap in enrollments, since proximity to girls' schools i s far from universal in rural
areas. The regressions o f school parti~ipation~~show that the presence o f a school offering primary
education within one kilometer increases the probability that a girl in a rural area will attend school by 15
percent. Existing initiatives, such as private sector use o f public facilities in the second shift and
concessionary access to public land, should be strengthened. Finally, the government should resist the
temptation to strongly regulate the private sector, at least in the next few years when the system
desperately needs expansion. To some extent the institutions that will oversee the activities o f the private


35See "Poverty Assessment, World Bank, 2002.
36See Chapter 3, Pakistan:PovertyAssessment, The World Bank, 2002.


                                                        52

sector should be the same as those that oversee the activities o f the public sector, such as student
assessment and examinationbodies.

3.32      The fundamental problems that the government must resolve to get better outcomes in education
are institutional. Inthe absence o f fundamental reform, additional funding by itself will achieve little.
Many o f these problems apply to the entire public sector, but they manifest themselves most visibly and
with the greatest impact inthe education sector because o fthe size and multitude o f actors that constitute
the system. Palustan has more than 150,000 primary and middle schools and more than 450,000 primary
and middle school teachers.          This size speaks to the need for accountability, responsibility, service
orientation, and incentives that rewardpeople for doing theirjob well.

ImprovingHealth and PopulationOutcomes

3.33      Duringthe last decade, Pakistan has made a noteworthy progress in improving its health and
population indicators. Infant mortality rate declined from more than 100 per 1000 live births in early
1990's to 82 in 2001/02 and the total fertility rate has declined from 6.4 to 4.8 during the same period.
Yet because these improvements are from a very low base and a number o f problems in the health
services delivery system remain unresolved, therefore there i s hardly in room for being complacent.
Eighty percent o f women are still delivering without assistance from trained personnel; contraceptive
prevalence i s still very low (28 per~ent);~'and due to poor nutritional status and low tetanus toxoid
immunization coverage o f pregnant women (48 percent), about one in every three hundred women die in
childbirth. Moreover, even the improved aggregate indicators conceal vast gender and location-wise
disparities. Mortality among the girls ages one to four i s reported to be 1.7 times higher than inboys in
the same age cohort, and the mortality rate for women between the ages of twenty and twenty-nine is
about double that o f men. All these indicators being significantly worse in the rural areas. Poor
governance and low public sector spending on the health sector i s the main cause o f the persistence o f
poor health indicators.

3.34      Poor governance, weak health sector management and l o w levels o f public spending are the key
factors behind poor health outcomes. During the second half o f 1990s, public expenditures on health
amounted to only about 0.6% o f GDP, among the lowest in the region. The government health services
have suffered from insufficientbudget allocations to fund current expenses other than staff salaries. These
include expenses for drugs, diagnostics (laboratory tests, x-rays), repairs and maintenance o f facilities,
replacement o f equipment, utilities, transportation, in-service training expenses, and health education
materials. As a result o f its poor quality, public satisfaction with the public health services i s low and i s
reflected in low use o f government's heath facilities (about 80 percent o f outpatient contacts occur in
private facilities, and only slight less for hospitalization). Inaddition, access to other basic services such
as safe water and sanitation i s still poor inmany areas, which, together with poor nutritional status plays a
major part inthe vicious circle o f illness, low productivity and poverty.

Health Sector Reforms

3.35      The government i s well aware o f the magnitude and depth o f the problem and has recently tried
to address the broader issues o f governance in social sector delivery through major refoms, especially the
devolution initiative, while a new agenda for improving health and population outcomes has been defined
under the I-PRSP. The I-PRSP contains targets to 2003/04 for child mortality, the number o f Lady Health
Workers deployed, prevalence o f malnutrition among pre-school children, percentage o f low-birth-weight
babies, contraceptive prevalence rate, and total fertility rate, and it aims to introduce a mechanism to
monitor outcomes on a regular basis.


37This is based on PIHS data, the govemment estimates are somewhat higher .


                                                            53

3.36    Moreover, the Government's new health policy i s specifically designed to achieve the medium-
term sectoral goals specified in the I-PRSP. The policy has targeted ten specific areas for reform
including, reducing the widespread prevalence o f communicable diseases, addressing inadequacies in
primary/secondary health care services, correctingmale and urban biases, bridgingbasic nutritional gaps,
improved regulation o f private health sector, and capacity building for health policy monitoring. The
Government's Population policy i s designed to achieve population stabilization by 2020 through
mainstreaming the concept o f Family Welfare in national and development programs, expanding service
delivery and improving its quality, fostering public-private partnerships wherever appropriate, promoting
health education through advocacy with key influential leaders, training and capacity building o f service
providers, decentralizing and de-federalizing the family welfare program, and strengthening health sector
monitoring and research. The provinceshave preparedplans to implement the policy, which are at various
stages o f development.

The Way Forward

3.37     The policy directions envisaged in the I-PRSP and the current health and populationpolicies are
by and large appropriate to address the issues faced by the sector. Overall policy objectives are however
demanding in terms o f local capacity, and there i s as yet no clear vision on how to improve
implementation o f health and population programs to deliver good results on the ground. In addition,
there are no definite initiatives or plans and means outlined to improve the quality o f health care inboth
the public and private sectors. The biggest concem however i s that I-PRSP's projected financing for
health sector seems to be totally out of line with the targets laid down inI-PRSP and the sectoral MDGs.

3.38    To make the public system more efficient and provide better value for money, the government
needs to take a series o f strategic actions:

        The public health services should seek to achieve a sharper focus on service priorities, interms o f
        bothresources and management attention. These priorities need to be developed keeping inview
        the level and direction o fprivate sector involvement inthe sector and the appropriate role of the
        government. Optimally, the government health services should seek to "crowd in" rather than
        "crowd out" private health services.

    1   Direct provision o f public health services should generally be based on equity considerations and
        on existence o fpublic goods, externalities, information deficiencies, and insurance market failure.
        Control o f disease vectors, protection o f water and food safety, and health education, preventive
    .   health, women and child health care and rural heath are areas where public sector interventions
        appear welljustified.
        Establishing better inputpriorities, for example setting up new facilities on a well-defined criteria
        and only inunderserved areas, improving the balance between doctors and paramedical staff, and
    .   focusing new recruitments only to correct s h l l mix; should increase the efficiency o f scarce
        public resources
        Building on, and learning from, successful initiatives like LHW, preventive and population
        programs can help lead to better channeling o f public sector efforts and resources.            The
        effectiveness o f these programs may be enhanced by a phased devolution o f management,
        implementation and financing o f these programs to the district governments.
        Forging better public-private partnerships to make more effective use o f health facilities and
        slulls. Inthis regard, each provincial government should develop a framework for contracting out
    .   packages o f services for defined population to NGOs or other private parties individual
        entrepreneurs or doctors groups.
        Making devolution to work for achieving the health outcomes (see Chapter 4).



                                                      54

3.39      Most importantly, however, public sector spending in the health sector should increase
significantly.   At present budgetary allocations for health sector falls well short o f what is needed to
achieve the PRSP targets and MDGs. Although, private sector spending can help close this gap to some
extent, the government's redistributive role requires its increased involvement to ensure access o f quality
health services to the poorest and the excluded, through direct intervention or financing health initiatives
for the targeted groups through the private sector health delivery system. The magnitude o f the
government's spending effort required to achieve the I-PRSP goals, though not yet fully costed, would be
considerably larger than that indicated by the I-PRSP estimates. Rough calculations indicate that
adequately financing the high-priority programs (like the LHW, Maternal and Child Health, HIV/AIDS,
Malaria and TB Control, Health Education, Population Program, human resource development and
capacity building, etc.), along with additional finances for public-private partnership programs, enhancing
the non-salary budget, and a normal increase in salary would require, at an average, an additional Rs 10
billion per annum. This would increase sectoral financing from the present 0.6% to an average o f 0.8%
during2003-07.

Accelerating Water Resource Development

3.40      Agriculture i s a major part o f Pakistan's economy and one o f the key engines o f overall economic
growth. Agriculture today contributes about 25 percent o f GDP, employs nearly 50 percent o f the labor
force, and i s responsible directly or indirectly for 60-70 percent o f exports. Within the agnculture sector,
crop production (principally wheat, cotton, sugarcane, and rice) accounts for about 59 percent o f value
added and livestock about 38 percent. Increases in water resources available for irrigation and expansion
o f irrigated area have playeda central role inagriculture growth since the 1960s. Agriculture now uses an
estimated 95 percent o f available water resources, and irrigated land i s responsible for about 80 percent o f
agricultural production. Total irrigated area in Pakistan i s about 46 million acres; nearly 97 percent o f
this irrigatedarea is inthe IndusBasin.

3.41      As Table 3.2 shows, water availability at farm head doubled over the period 1960-90 but has
remained unchanged since then (hopefully the sharp drop in last two extreme drought years will prove
temporary). This vast expansion o f irrigation water supplies was made possible by the large investment in
expansion o f the irrigation system and the construction of two major storage siteshydroelectric projects at
Mangla and Tarbela inthe 1960s and 1970s under the Indus Basin Treaty, as well as by the exploitation
o f groundwater resources mainly by private tubewells. Inthe 1960s total water and power investments in
the public sector exceeded US1.5 billion, accounting for more than 50 percent o f total public pend ding.^'
Half o f this amount represented the Indus Basin Replacement works. The long-term investments made in
development o f water resources in the 1960s and 1970s sustained economic growth well into the 1980s.
Yet the long-term vision of the planners o f that era,39inwhich Palustan would further integrate water and
power development and engage in highlyproductive agriculture inwhich water and other inputsare used
very efficiently, has not been realized. Increases in water supplies have tended to extend and prolong
traditional low-productivity agriculture. Slow growth inproductivity in agriculture i s due to many factors,
including insufficient investment in human development and unequal asset ownership. But the low level
o f public investment in water including drainage and lack o f sufficient progress on institutional
arrangements that encourage farmers to pay for water and undertake on-farm development have also acted
as major constraints on the development o f modern agriculture. Public spending in agriculture and water
touched a low o f 0.8 percent of GDP in 2000-01. Other than the Chashma Right Bank Canal and the
Pehur High Level Canal, expansion o f irrigation in the past fifteen years has been limited to small,
isolated schemes outside the Indus Basin. More important, the original expectation that a second stage
storage on the Induswould be required and built by late 1980s or early 1990s has proved elusive because


38ParvezHasan, "Pakistan s Economy at the Crossroads- Past Policies and Present Imperatives," 1998.
39See Water and PowerResources of West Pakistan, a Study in SectorPlanning; John HopkinsUniversityPress, 1968.



                                                          55

of the political controversy surrounding the Kalabagh Dam and the lack o f timely development of
alternatives.

                               Table 3.2: Water Availability, 1960/61-1999/00

                   Year        Water Availability at FarmHeadMAF         Irrigated
                               CanalWater Groundwater         Total      Area MA    Acre-ftlAcre


                  1960-61          48.4           4.2          52.5         25.7        2.04
                  1967-68          56.8           11.7         68.5         30.9        2.22
                  1977-78          61.6          27.8          89.4         35.1        2.55
                  1985-87          61.4          40.9         102.3         39.8        2.57
                  1990-91          65.0          41.0         106.0        41.9         2.53
                 1999-2000         61.5          41.6         103.1        44.0         2.34

3.42     With the rapid increase inpopulation, the competition for water is becoming more intense and
disruptive among upper and lower riparianprovinces and among the agricultural, industrial, domestic, and
environmental sectors. Since 1999 water availability in the system has been inadequate to supply the
water allocations agreed-to inthe 1991Water Accord owing to drought and inadequate water availability
inthe major reservoirs as well as canal capacity. Water shortages on the order of 40-50 percent were
experienced during the 2002 rabi and early kharif crop seasons. These shortages created a strong sense o f
crisis--they mobilized sector authorities to a degree that other persistent problems, such as deteriorating
performance o f infrastructure, low productivity, or persistent poverty, didnot.

3.43     WAPDA was the first of the key sector organizations to prepare and present a plan--a water
resources development strategy document called "Vision 2025" (2001) that calls for huge investments in
multipurpose storage reservoirs and several new canals to expand irrigation during the period 2000-25.
The goal o f the plan i s to address the problems o f water scarcity, the flow o f flood water to the sea
unutilized, safe disposal of drainage effluent to the sea, tapping o f groundwater, andreduction o f seepage
loses inthe country's existingirrigation system. Partly because o fthe sense o furgency created by drought
and the political impasse over the KalabaghDam, the government approved Vision 2025 inprinciple, and
its main elements were, by and large, included in the Planning Commission's Ten-Year Perspective
Development Plan (TTPP--FY 2001-11). Some projects have actually entered the Public Sector
Development Program (PSDP) during the past two years. The cost o f federal water projects included in
the TTPP during2003-07 i s estimated at Rs.202 billion, or roughly Rs. 40 billionper year compared with
Rs.9 billionspent in2001-02.

3.44     Because o f great neglect of water investments in the recent past, a substantial acceleration in
overall public spending inthe water sector appears justified. WAPDA and the Planning Commission must
be commended for their initiatives inreviving water sector investments. The thrust o f the present program
has several problems, however, that require careful consideration and review by the Govemment o f
Pakistan. Many o f the issues involved, the tackling o f which can yield a more optimal and balanced
expansion o f irrigation water-related public spending, are discussed at length in Volume 2 o f this report.
These issues can be summarized as follows:

         The need for a clearer vision that integrates long-term agriculture, irrigation water, and
         hydroelectric development
       . An excessive emphasis on extending irrigation facilities, some o f whom may prove marginal and
         all o f which will yield economic benefits only after a number o fyears
         Inadequate attention to the urgent need to rehabilitate existing irrigation assets and on-farm
         development



                                                      56

        .Inadequate attention to the political and technical issues that may prevent early resolution o f
         much-needed additional storage on Indus
         The need to harmonize ambitious water sector plans with overall resource availability and to
         avoid preemption o furgent social investments
        .The need to balance institutional development needs, including federal-provincial coordination
         and greater investment inplanningand investigation, with higher water sector investment
         The need to identify follow-up work to firm up long-term plans and to seek necessary
         mobilization o f external assistance

3.45     WAPDA's Vision 2025 and the Medium-                                              Figure 3.6 Public Sector Expnditurer
Term Investment Plan (MTIP) o f the Ministry o f
Water and Power, completed in October 2002, a                        "0

year after the TYPP, do not clearly relate the role o f
water development either to future agricultural                  2    lSaa

needs and agncultural potential or to the limits o f             s
                                                                 0

irrigation water availability. Any long-term plan                2    loom


for agricultural and water development must come                       %om
to grips with the fact that groundwater development
i s already nearly at its limit, and surface water
resources development will approach its overall
limit--through increased storage o f flood flows in
the Indus River--in the next twenty-five years at the
latest.

3.46     Even making the optimistic assumption that
two large storage sites on the Indus can be built
over the next twenty-five years, the addition to the
total water supply in the Indus system would be
only about 15 percent, a sharp contrast to the
massive additions that were possible during 1960-
90.    The environmental, political, and economic
feasibility o f a second storage, however, remains an
open question.       One o f the concerns, now being
addressed      through      environmental      studies,    i s
whether there will be sufficient water below                    e                                                                                    \
K0t1-i.~' The wony is that additions to surface                            Figure3.8: Public SectorExpendituresas % of total PSDP

water supplies, without adequate drainage and                       35 0

evacuation o f salts, will aggravate the problem o f            & 300
salt balance in the Indus Basin.41The finalization              8
                                                                    250
o f a national drainage accord inthe near future and            fB  200
adequate investments in a national drainage                      0


program will help to mitigate environmental                      1
                                                                2   I 5 O
                                                                    100
concerns. But the basic fact i s that the past pattern
o f agricultural growth,         in which insufficient               5 0

                                                                     0 0

                                                                    1980/81198~831984/851986/871988/891990/911992/931994/951996/971998/99l000/012002i03




40 The Ministry o f Water and Power and the Indus River System Authority (IRSA) are jointly commissioning a study to
determinethe flow requiredbelowthe Kotribarragefor ecologicalienvironmentalneeds andfor preventingseawater intrusion.

41 Surface water supplies bring an estimated 33 million tons of salt each year into the irrigated areas of the Indusplain, and a
further 28 million tons are added from groundwater from tubewells. Only a small fraction o f the total salt is carriedout of the
basinby the IndusRiverbelowKotri Barrage or the Left Bank Outfall Drain(LBOD) inSindh.


                                                             57

attention was given to water conservation and effectiveness o f water use, i s not sustainable.

3.47    An optimal water strategy must emphasizeall thefollowing elements:

   m  Supply expansion: storage expansion (Mangla, Bhasha, andor Kalabagh),                      small dams,
      improvementsindiversion capacity and function, improvements inoverall system operation
   m  System expansion: external expansion (three canals), internal expansion by remodeling and
      extension o f existing Canal CommandAreas (CCAs).
   4  Management: on-farm water management interventions (watercourse lining and land leveling, for
      example), supply management (rehabilitation and modernization, flood protection, improvements in
      operations and maintenance versus new investments), improved water allocations and scheduling,
      improved equity at system and distribution levels, groundwater management, incentives,
      introduction o f new agriculture and irrigation technology, effective access to markets, improved
      knowledge and information, humanresource development and research
   4  Sustainability and environment: waterlogging and salinity, reducing drainable surplus, water
      quality, safeguarding ecology below the Kotri Barrage
   4  Productivity:croppingpatterns, water rights andmarkets, water pricing
   4  Governanceand institutionalreform: improvedpublic service and accountability, cost reduction,
      decentralization, empowerment o f users

3.48     The optimal water strategy for Pakistan's integrated Indus Basin System i s technically complex.
The trade-offs-among the set o f alternatives and complements (storage, conservation, reservoir scheduling
policy, estuary maintenance policy, sSystem expansion) are not simple because o f the varying degrees o f
capital lumpiness, the uncertainty o f costs and benefits, and the dependence o f benefits and costs on the
sequencing o f investments. In addition to the technical complexity, the integrated system includes a
cascade o f upper and lower riparians, which introduces political economy considerations inthe choice o f
the optimal strategy. To address this technical-cum-political complexity, it i s essential that provincial
governments (lower and upper riparians) as well as intraprovincial stakeholders (system rehabilitation
versus system expansion) engage in extensive dialogue to agree on the fkamework and the rules for
sharing the resource and the costs and benefits. This framework may be formalized as water and drainage
accords to permit investment choice and sequencing.

3.49     On storage, the decision to raise Mangla Dam appears sound because the increased storage
capacity will have a reasonable probability (71 percent) o f filling each year. Also, it i s the only option
available to the government inthe next five to seven years.

3.50     The decision to accelerate the feasibility o f Basha Dam i s also an important step in resolving the
political impasse on Kalabagh. Much more work needs to be done, however, duringthe next year to deal
with the political economy and technical questions before Pakistan is ready to propose a large storage
project to donors for financial support. Political consensus needs to be built on the sequencing o f storage
projects, assisted by good economic studies that will throw light on the options available for operation o f
new reservoirs. The technical choices between water releases for the agriculture sector and electricity
generation and the management options for releases in rabi versus early kharif must be debated and
settled as a part o f the decision to build additional storage. This dialogue will, on the one hand, clarify the
economic and political rationales for future storage and, on the other hand, help minimize future conflicts
among provinces and incanal commands within provinces.




                                                        58

                                                                           (2016)

                                      Onginal Storage Capacity


         I" -                       (Tarbela + Mangla + Chashma)              L


           I                                                                                      12 8
    G
    a
    .-b
    E                                                                                             10 4
                                                            Mangla Raising
                                                               (2008)
    8P
    m




        I
       1976    1980   1984   1988    1992      1996       2000      2004     2008  2012  2016  2020    2024
                                                          Years




3.5 1   The government has initiated a large system expansion program focused on three canals: Greater
Thal, Kachhi, and Rainee canals. These canals will enable diversion o f 4.6 MAF o f water and will cost
more than Rs. 90 billion. These projects are driven primarily by the desire o f the provinces to utilize their
respective shares under the Water Accord, though until more storage i s built they will receive perennial
supplies for only seventy-five to ninety days during the kharif season. While these projects respond to
strong regional interests, the final scale, phasing, and cost coverage o f these projects must be reexamined,
especially after the feasibility studies of Kachhi and Rainee canals have been completed. The trade-offs
among provinces and stakeholders within provinces must be revisited and agreed-on as part o f the new
drainage and water accords mentioned. The conflicting political economy considerations underlying these
choices cannot be minimized, but with the help o f technical feasibility studies, consensus should be
possible on the broader national interest.

3.52    The following further points need to be taken into consideration before fully committing the
funds to the program o fcanal expansion during2004-07:
    .   Since these canals would receive water for a period o f only seventy-five to ninety days in a year
    .   (during July to September), the non-availability o f water during April-June (the kharif sowing
        season) would affect croppingpatterns and yields.
        The command areas include depressions, sand dunes, and large areas with low population. Thus
    .   water losses would be high, and agricultural development would take place very slowly as
        colonizationand command area development take place.
        As currently proposed, the projects do not provide funds for the government's share incommand
        area development, and without this development project benefits would remain elusive. But
    .   adding these costs, whether they are borne by the government or the farmers, will sharply lower
        the economic rate o freturn.
        Detailed social and environmental assessments should be carried out to determine canal
        alignments that would explicitly target the poorest areas and small landholders.



                                                                59

     m    The projects would add a substantial burden to already strained operation and maintenance
          budgets. Full agreement o f the provinces that needed requirements will be funded must be
          obtained before committing large funds to these projects.

3.53      The Ministry of Water and Power i s considering a "modular approach" for implementing these
projects in a phasedmanner. This approach may move toward satisfying both the political commitments
and the TYPP objectives, as well as providing a highly visible demonstration o f the new future o f
irrigated agriculture in the Indus Basin. Without prejudging the completion o f feasibility studies and
detailed appraisals, we believe these projects also provide an exceptional opportunity to establish a
decentralized and farmer-managed system right from the start.            Instead o f expanding the existing
imgation departments into the new command areas with their traditional modes o f operation,
maintenance, and implementation, the government should make concerted efforts to establish Farmers'
Organizations (FOs) through a program of social mobilization and capacity building. This step should be
followed by the establishment o f area water boards. Regardless o f their ultimate benefits, the major canals
will yield economic benefits only after several years. Short-term benefits will remain elusive. These
long-term programs must be strongly supplemented with programs and projects that will yield immediate
benefits.

3.54      Various estimates put the potential savings from the current supply at about 4.7 MAF,just under
5 percent o f the current average availability o f surface water in the Indus Basin. These savings would
come primarily from lining watercourses and leveling land under the on-farm water management projects
o f the provinces. Water conservation and efficiency gains are achievable provided that farmers benefit
from the savings--not only because equity would improve and previously unserved farmers would get
water, but also because all farmers in the improved canal command would be able to decide how to use
the saved water to increase their production and income. A substantially higher allocation for operation
and maintenance and fundamental changes in the operation and maintenance regimes are also urgently
needed. Current estimates o f system rehabilitation (essentially accumulated deferred maintenance) and
remodeling costs exceed Rs. 200 billion.42Although the ongoing institutional reforms would ultimately
address issues o f overstaffing and lack of farmer participation, immediate action i s needed to stem the
deterioration o fthe irrigation and drainage system. A well-designed program would:
     ..   Establishrealistic finding requirements through an independent evaluation;
          Rationalize fund allocationto meet priority needs;
          Initiatebenchmarkingand operation and maintenance performance audits: and
          Rationalize abiana rates and improve assessment and recovery (empowering farmer organizations
          not only to operate and maintaintheir canals, but also to collect abiana and retaina sufficient
          portion o fthe recovered funds to meet their needs would be a key incentive for increased cost
          recovery).

3.55      Higher allocations for maintenance of existing assets and speedy on-farm development must be
underpinned by reforms in the system of governance. Based on the experience gained so far, a new
consensus on the way forward within the provinces, and between the provinces and the federal
government, i s needed that would essentially re-launch the water management program.              This new
consensus should focus on critical steps to be taken and on piloting and monitoring a variety o f
approaches within each pilot canal command over a definite time period. Business as usual would prove
costly in terms o f both forgone benefits of participatory irrigation management and further deterioration
o f the irrigation system. The lessons leamed from the ongoing reform efforts are that these fundamental
reforms require champions at the highest level o f federal and provincial governments and should not be
left to the discretion o fthe agencies being reformed.


42PER consultant's estimate.


                                                      60

3-56       A balanced program of institutional development, higher investments in water, and appropriate
incentives to conserve water and use it more efficiently i s well within Pakistan's reach. The substantially
larger additional funding for the water sector during 2004-07 can bejustified provided the program gives
due weight to all key components and strikes a balance among short-term, medium-term, and long-term
investment and institutional needs. An indicative program on these lines totaling Rs. 252 billion during
2003-07 is presented inVolume 2. The suggested program provides for ongoing programs (notably Right
Bank Outfall Drain I1(RBOD2) and National Drainage Program (NDP)), as well as a number o f new
projects (Greater Thal Canal, Gomal Zam and Mirani Dams, Mangla). It also allows for substantial
acceleration o f nondevelopment and development budgets o f the provinces. It makes provision for the
government share o f command area development for the new canals, new drainage initiatives including
institutionalreforms, and on-farm development.

3.57       The balanced program that i s beingproposed can help initiate the transformation o f the water and
agriculture sector even in the medium term. The more than doubling o f the real spending in water over
FYs 2002-07 will take about 25 percent o fthe additional fiscal space that is expected to be created under
the baseline scenario and will sharply reverse the declining trend in water sector spending. The
acceleration will be even faster than proposed in the education sector. Needless to say, such large
spending carries risks and could waste resources if it i s not accompanied by much tighter controls on
program and project priorities and stronger implementation and institutional arrangements than exist at
present. A fundamental strengthening of the sector's                  planning, research and development,
implementation, and monitoring mechanisms, including much greater interprovincial and federal-
provincial coordination, i s also necessary to facilitate decision malung on the more difficult longer-term
issues.

Transforming the Power Sector

3.58       On the surface it may appear that the power sector investment program will not have major
claims on public resources during 2003-07. The Planning Commission's TYF'P suggests an allocation in
the Public Sector DevelopmentProgramfor the power sector o f only Rs.90 billion for the next five years,
or less than 0.4 percent o f GDP-lower          than in the recent past. Comfort i s drawn from the fact that at
present there i s surplus power in the country, and with the commissioning o f the large Ghazi Baroda
hydroelectric project by the end o f 2004, there will be a substantial addition to capacity. The coming on
stream o f the Ghazi Baroda project will no doubt be a milestone inthe history o f power development in
Palustan. It will substantially improve the hydrothermal balance in electricity generation and help to
reduce overall costs. After 2003-04 its completion will provide room for other projects in the power
sector: PSDP allocations for Ghazi Baroda for FY 2002-04 alone are estimated at Rs.37 billion.

3.59       In reality, however, there are many unresolved issues in the power sector, which could both
complicate the creation o f fiscal space and seriously hamper larger allocations to high-priority sectors.
These issues center on the financial weaknesses in the two state-owned power entities, WAPDA and
KESC. These weaknesses stem from:
      .....Continued large losses in the power system, reflecting a combination o f factors (low level of
           efficiency, theft o f electricity, and inadequate investments inpower distribution).
           The highcost o f power from independent power producers (PPs).
           Increasing consumer resistance and political sensitivity to higher electricity tariffs.
           A highlevel o freceivables from public sector consumers.
           An implicit large subsidy to consumers inthe Federally AdministeredTribal Area (FATA).

3.60       More recently, the financial situation o f WAPDA and KESC (notably o f KESC) has deteriorated,
which has further eroded the entities' capacity to finance their investment programs. A number o f factors


                                                          61

account for this deterioration. Although tariffs have been raised, power purchase and fuel costs also
increased substantially, and these increases were not fully passed on to consumers. Targets for reduction
intechnical andnontechnical losses were not achieved. And the level ofreceivables--notably frompublic
sector consumers--remained high.

3.61     Recognizing these challenges, the government has established a reform agenda for the power
sector driven by the following key objectives:

        Mobilize private resources for the development o f the sector, without recourse to the government
        budget, through the March 1994 Independent Power Producer (IPP) Policy.
        Improve the utilities' performance and efficiency through corporatizatiodcommercialization (in
        preparation for their eventual privatization), enhanced managerial autonomy and accountability,
        competition, and profit incentives.
        Introduce a transparent regulatory fiamework (particularly in setting electricity tariffs) to provide
        comfort to potential investors, while enabling the government to maintain certain socially
        desirable activities such as "lifeline rates" for poor consumers and rural electrificationprograms.

3.62     Considerable progress was achieved in the mid-1990s in mobilizing private resources, through
IPPs, for new generation projects. The Kot Addupower plant was privatized in 1996, through the sale of
36 percent o f shares (along with management control) to a strategic investor. The TPP projects, however,
were solicited under fixed-price (or fixed-return), long-term contracts, and the government's main focus
was on the early commissioning of the projects to overcome the supply shortages being experienced at
that time. Inthe process, too little attention was given to the demand side issues. Moreover, the impact o f
these investments on the overall financial viability o f the sector, including the ability o f WAPDA and
KESC to pay for power purchases from IPPs, was not adequately analyzed. Although shortfalls in
generating capacity have been overcome, there i s actually some excess capacity in the system because
economic growth has been unexpectedly slow. Consequently, the power sector i s facing a financial crisis:
KESC does not cover even its operating costs with the existing tariffs, and WAPDA i s unable to mobilize
adequate funding, through internal cash generation, for its investment program.43 While WAPDA has met
the internal cash generation target in some recent years, it has done so mainly by deferring critical
investments, defaulting on debt service obligations to the government, or buildingup large arrears to fuel
and equipment suppliers and contractors.

3.63     Meanwhile, actions have been taken to unbundle WAPDA and to establish the regulatory
framework. WAPDA's Power Wing was reorganized into eight distribution companies (discos; a ninth
disco covering the tribal areas in NWFP was recently established); one transmission and dispatch
company (NTDC); and three thermal generation companies (gencos). WAPDA i s now responsible solely
for developing and managing the hydropower generation facilities. Assets and manpower have been
allocated to these entities, and most operational functions (including the management o f investment
programs) have been delegated.            NEPRA has notified its main rules (for tariff setting, license
applications, etc). It held hearings on tariff and license applications, which are open to the public, and
issued a number o f tariff determinations. A formula for passing the impact o f changes in fuel costs to
consumers, through quarterly tariff adjustments, has been approved and i s starting to be implemented.
Licenses were issued to all Discos and Gencos by the end o f June 2002 and to NTDC by the end o f
December 2002.




43The financial performance targets for WAPDA and KESC, which were agreed-to with the government and the entities under
ongoing(andpast)World Bank and Asian DevelopmentBankprojects, includedWAPDA's financingo f at least40 percentof its
investment programthrough intemalcash generation; the correspondingfigure for KESCwas 25 percent.


                                                           62

3.64     Although the government has taken a number o f steps to promote private investment inthe power
sector, the desired goal of ensuring the development o f the sector without recourse to the government
budget, except for hydroelectric development and possibly transmission, i s nowhere in sight. The
continuing financial distress in the KESC, notwithstanding a substantial injection o f public monies to
fundpast losses, and international conditions make it highly unlikely that KESC can be privatized inthe
near future. It also appears somewhat unlikely that the newly corporatized distribution and generating
companies under WAPDA will actually be privatized on a timetable to fund power sector investments
during2003-07.

3.65     Consequently, there i s a disconnect between large national power sector requirements and the
amounts that can realistically be expected from the development budget and own resources o f WAPDA.
Whereas the TYPP has notionally allocated Rs. 90 billion for WAPDA mainly from external loans,
WAPDA's proposed programfor 2003-07, includingrural electrification, is Rs.261 billion. So there will
be an additional requirement, after providing for WAPDA's own resources, o f Rs. 90 billion to be met
from the budget and/or borrowing in the market with a government guarantee. KESC investment needs
will be over and above this. The extra requirement for the power sector investment could wipe out one-
thirdof the fiscal space that is supposed to be available underthe baseline scenario.

3.66     What can be done? What should be done? Although the system o f automatic adjustments in
electricity tariffs to reflect costs must be further strengthened, there may not be much scope for increasing
real as distinct from nominal, electricity tariffs44. As Table 3.3 shows, the average residential tariff
increased by 50 percent over 1996-2002 and i s currently about 6 U.S. cents per Kwh. The industry and
commercial tariffs have been difficult to increase because they are already high and adversely affect
competitiveness.

3.67     Clearly a much more detailed plan for the power sector for the period 2004-08 urgently needs to
be developed and harmonized with the development o f rolling three-year plans and the MTBF. In the
absence o f a plan, the power sector could derail the government's economic and social priorities. Some
elements o f the power sector review should be the following:
    ..
    .....A realistic butproactivetimetable for privatizationofcorporate entities inthe sector
         Carefulreview o fthe present plans o f WAPDA, especially for generation
         Reinforcedefforts to reduce technical losses
         Solutions for managing load shedding inJanuary when low level o f hydroelectric generation will
         reduce supply
         Separation o f rural electrification          from normal WAPDA funding and inclusion o f rural
         electrificationas a part o fthe poverty reduction strategy
         The formal launching o f the new policy for power generation for the private sector
         Innovative approaches to raise capital for the power sector ahead o f privatization




44According to recent World Bank estimates, 10-15 percent increase in nominal tariffs over the next few years will be
inescapable even with aggressive cost reduction and speedy progress towards reduction o f theft.


                                                              63

                         Table 3.3: ElectricityTariffs (WAPDA System)-FY96-2003
                                                  /Paisas/kWh'h)

                                  FY96 FY97 FY98 FY99 FYOO FYOl FY02 FY03
           Domestic                 146      117    223    247   243     259     318       352
           Commercial               534      615    702    698   698     704     709       729
           Industry                 335      391    444    417   417     416     419       442
           Bulksupply               275      355    419    437   437     424     448       483
           Agriculture              161      192    206    344   281     258     294       345
           SYSTEMAverage          223.9     269.6   319    348   339     337     372.7    405.4


                        Electricity Tariffs inrealterms (WAPDASystem) FY96-2002
                                                                        --
                                     FY96       FY97      FY98   FY99       FYOO      FYOl     FY02
  Domestic                           146.0      103.1      182.6  191.1     182.9      184.5   216.5
  Commercial                         534.0      542.1     574.8   539.9     525.3     501.6    482.7
  Industry                           335.0      344.7     363.5   322.6     313.8     296.4    285.3
  Bulk supply                        275.0      312.9     343.1   338.0     328.9     302.1    305.0
  Agriculture                        161.0      169.2      168.7  266.1     211.5      183.8   200.2
  SYSTEMAverage                      223.9      237.7     261.2   269.2     255.1     240.1    253.8

  GDP Deflator (1995-96=100)         100.0      113.4      122.1  129.3     132.9      140.4   146.9



                              ElectricityTariffs (WAPDA System) FY96-2003
                                                                  --
                                                  (in Usckwh)

                           FY96       FY97      FY98       FY99   FYOO      FYOl       FY02     FY03
Domestic                    4.35       3.OO      5.16       5.28   4.69      4.59       5.44    6.02
Commercial                 15.91      15.77      16.25     14.92   13.48     12.48     12.13    12.47
Industry                    9.98      10.03      10.28      8.91   8.05      7.37       7.17    7.56
Bulksupply                  8.19       9.10      9.70      9.34    8.44      7.51       7.66     8.26
Agriculture                 4.80      4.92       4.77      7.35    5.43      4.57       5.03     5.90
SYSTEMAverage               6.67       6.91      7.38      7.44    6.55      5.97       6.38     6.94

Exchange Rate (Rs/US$) 33.568        38.994     43.196     46.79  51.771     56.43     58.457   58.457




Conclusion

3.68     The financing o f the poverty reduction strategy shouldbe the first charge on the additional fiscal
resources that are likely to become available to the Government o f Palustan. High-priority poverty
spending should be defined more broadly to include law and order maintenance, access to justice, and
rural electrification. A large part o f higher spending under the PRSP would also be on education and
health. Without adequate funding (and stronger delivery systems), social sector goals will again be
frustrated. There i s also a definite need to accelerate spending on water resource development. But the
present government program appears lopsided and has not been carefully thought through. A balanced
program o f increase in storage, system expansion, rehabilitation o f existing assets, on-farm development,
and concomitant institutional development will yield far better economic and social results. The persisting


                                                       64

power sector problems remain a threat not only to government finances, but also to an orderly
reorientation o f public sector priorities. These problems need to be analyzed and resolved with the help o f
appropriate sector work and policy advice.

3.69      Although there could be several ways in which the additional fiscal space could be allocated, in
light o f the government's priorities and commitments, Table 3.4 illustrates one such way o f allocating
additional resources to the high-priority sectors. To meet the government's PRSP targets, substantial
portion o f the additional fiscal space has been assigned to the poverty reduction sectors and programs,
with high-priority given to human development, pro-poor infrastructure and improving the access o f the
poor to justice. In the illustrative table, this i s reflected by increasing the pro-poor expenditure by an
additional 2.2 percent of GDP by 2006/07. As for other expenditures, the government's resolve to correct
the past neglect o f the water resources i s reflectedby showing substantial increase inthe short-, medium-
and long-term investments in the irrigation sector.                     Increases in other infrastructure investments are
mainly to improve the efficiency o f port operations, railways, road and power sectors. A considerably
higher allocation for general administration i s mainly to cater for higher salaries and capacity building
needs o f the civil servants, especially at the district level.

                           Table 3.4: IllustrativeAllocationof AdditionalFiscalSpace

                                                                       (Percent of GDP)                Real Spending
                                                                                                   (Rs. billion) " Percent
                                                                     0011022006/07 Increasr !001/02 2006107 Increase

           'riority poverty related spending                          4.8      7.1     2.21       175.5       336.6       91.8
               Education                                              1.8      2.6      0.73       66.0       121.8       84.4
               Healthandpopulationplanning                            0.6      0.9      0.29       20.4        40.6       98.6
               Law andorder                                           1.0       1.2     0.27       34.5        58.3       68.7
               Rural electrification                                  0.1      0.3      0.15        3.5        11.9      241.1
               Water supply and sanitation                            0.1      0.2      0.13        2.3          9.1     293.5
               Ruraldevelopment                                       0.3      0.5      0.17       12.3        24.4       97.6
               Irrigation                                             0.3      0.5      0.22       10.4        24.4       134.6
               Roads, highways andbridges                             0.4      0.5      0.16       12.8        24.4       90.8
               Other "                                                0.4      0.5      0.10       13.2        22.0       66.4

           Ither non-interest non-defensespending                     6.1      7.3      1.21      220.1       347.3       57.8
               Generaladministrationandpensions                       2.5      2.8      0.30       89.3       131.8       47.6
               Large scale water development                          0.1      0.6      0.48        3.6        27.7      665.0
               Railways, ports, nationalroadsand power                0.8       1.0     0.27       27.3        48.7       78.4
               Other                                                  2.8      2.9      0.16       99.9       139.2       39.4

           rota1                                                     10.9      14.3     3.42      395.5       684.0       72.9
           'Constant 2001/02prices.
         2/Social security and social welfare,food subsidies,food support program, natural calamities and other disasters,
         and land reclamation




                                                                   65


         CHAPTER 4: IMPROVING EFFECTIVENESS                        OF PUBLIC EXPENDITURES

Introduction

4.1.     Delivery o f most public services inPakistan, including education, health, water and sanitation and
law and order, i s poor after several decades o f secular decline. This decline has been only partly the result
of inadequate spending on essential public services inthe face o f a very highrate o fpopulation growth. It
has also been the consequence o f a lack o f well-functioning institutions and erosion over time o f rules,
procedures, and organizational arrangements that ensure the effectiveness o f public expenditures. Poor
budgetary practices, lack o f accountability, inadequate management and organizational capacity, and poor
governance have all contributed to poor outcomes. The issues o f civil service reform, which are at the
heart o f institutional strengthening, are discussed in the next chapter. This chapter focuses on budgetary
and planning processes, monitoring mechanisms, the incentive structures for implementing government
programs and policies, and the recent devolution decisions, which all will have to play an important role
inimprovingthe deliveryofpublic services, especially inthe social sectors. Itoutlines the important steps
the government has taken to improve the effectiveness o f spending and the tasks that remain in
implementing the reform agenda. It also pursues the issues o f a further redefinition o f the role o f the state,
and public-private partnerships in financing, production, and distribution o f a number o f public services
because the public sector capacity problems cannot be resolved quickly.

Responsivenessto Citizens

4.2.     Before discussing the mechanics o f better and more efficient public service delivery, it i s
important to stress that the ultimate test o f effective service delivery i s whether it meets the expectations
of citizens, especially the poor. The usual yardstick used by the government to measure the effectiveness
of these services i s the amount o f money spent on these services and on creation o f physical assets. Little,
ifany, attempt has beenmadeto monitor andanalyzewhether these services are deliveredefficiently and
whether their provision meets citizens' expectations. In order to determine the level o f consumer
satisfaction with a number o f key public services, the National Reconstruction Bureau (NRB) engaged an
international NGO-the Community Information, Empowerment and Transparency (CIET) to undertake a
consumer satisfaction survey in all the districts on in the country. The survey attempts to measure the
extent to which users are satisfied with these services, aspects o f service that were not effective, and
groups o f people most affected by the deficiency in service. This study also provides a database o f
information and sets service standards for analysis in future.

4.3.     Being a measure only o f people's perceptions, the results o f the CIET survey need to be
interpreted with utmost care. The results, however, do reveal a strong dissatisfaction within the general
public with the quantity, and more important the quality, o f public services (especially education, health,
water supply and sanitation, and roads), which clearly demonstrates the low effectiveness o f public
expenditure.

Reachingthe Poor

4.4.         Implementing the strategy highlighted in the government's I-PRSP i s important for
alleviatingpoverty. It however, needs to be recognized that I-PRSP was prepared against the background
of strict fiscal stringency. With more fiscal space likely to be available in future, the poverty reduction
strategy can be funded more adequately and the approach to poverty reduction can, as discussed in
chapter 3, be broadened. However, there remains the need, within the broad framework o f poverty
reduction, to design programs specifically targeting the poor.

4.5.      The World Bank's Poverty Assessment4' has shown access o f the poor to the basic public services
i s low. For example, net primary enrollment rates are 37 percent for the poor, as compared to 59 percent
for the non-poor; the incidence o f medical consultation for diarrhea among children i s 79 percent for the
poor compared to 84 percent for the non-poor; only 15 percent o f married women o f ages 15 to 49 in the
poorest households have ever used contraceptives, compared to 25 percent o f those in the richest
households; as many as 24 percent of the poor rely on potentially unsafe sources for drinlung water,
compared to 19 percent of the non-poor; about 52 percent of the poor live in households connected to
electricity, compared to 76 percent o fnon-poor households.

4.6.      A determined effort is therefore required to focus these services toward the poor through
enhanced outreach and better targeting o f these services. It should, however, be kept in view that in
Pakistan, supply side bottlenecks are only one side o f the problem o f limited access of the poor to public
services. The opportunity and actual costs, social norms and cultural traits all contribute towards less than
full utilization o f these services by the poor. But researchhas also shown that improvingthe quality o f
services can effectively offset most of the demand side obstacles. Institutional and governance reforms
and adequate allocations for key non-salary expenditure items, like textbooks, class material and
medicines would be required to enhance the quality of education and health services. Appropriately
designed incentive schemes, like scholarships, edible oil and school lunches (as planned under Taiwana
PalustanProgram) can thus help alleviate some o f the demand side problems.

4.7.      Apart from developingprograms in education and healththat specifically target the poor, it i s also
highly desirable that the government aligns the planned large water sector public investments with its
poverty reduction strategy. The phasing o f these investments should be determined by the extent to which
the poor can benefit from these projects. Also, the land development aspect o f the Thal, Rainee and
Katchi Canals require a strong pro-poor focus. Most o f the additional arable land that may become
available as a result o f these projects should be distributed among the rural poor.

Improving Budgetary Processes

4.8.      Despite some recent improvements, the budget processes continue to suffer from a number o f
weaknesses that reduce the efficiency of the budget and therefore the effectiveness o f public expenditure.
The mainweaknesses o f the process include (1) lack o f strategic focus, and a lack o f clear articulation and
ownership o f sectoral priorities; (2) lack of information on the costs of policies, programs, and services
and an intertwining in a single agency o f policy advice, regulation, and service delivery; (3) a primary
focus on inputswith performance judged largely by how closely spending matched budget appropriations;
(4) a short-term horizon for budget decision mahng that fails to account for longer-term costs and
benefits; (5) an artificial and unwarranted separation o f development and recurrent budgets, making it
impossible to prepare an integrated and coherent strategic budget framework; and (6) few incentives for
agencies to save budgetary resources because current year spending i s the starting point for the next
year's allocations.

4.9.      In the past, the budgetary practices suffered also because the planning and programming
mechanisms were unable to withstand the political pressures that resulted in a number o f economically
unviable projects being included into the development program (and budget).. As many o f these were
large projects they not only pre-empted a sizable portion o f available development funds but also
imposed an unnecessary burden on the federal or provincial recurrent budgets (see para [4.77] for some
examples o f these projects). Most of these projects were started by ignoring or short-circuiting the regular
approval process, with some projects getting included inthe development program without even a proper
feasibility study. Removal o f key infrastructure sectors (energy, highways, telecommunication, etc.) from


45 Poverty in Pakistan: Vulnerabilities, Social Gaps, and Rural Dynamics, The World Bank, October 2002.



                                                               67

the purview o f the Planning Commission's approval process was one important reason leading to
inclusion o f these projects into the development program. The decision o f the government to restore
Planning Commission's authority over project approval process for all sectors i s a step in the right
direction. Fiscal prudence and adherence to the established processes and procedures in approving
development projects also need to be ensured. In this context, it i s advisable that the government
undertake a through review o f all large new projects which are presently in the program. In future, no
project should be undertaken without proper feasibility and need assessment.

4.10.        The problems with the budgetary processes were less apparent in earlier years, when the
overall levels o f public spending were growing rapidly, planning processes were better integrated with
economic decision malung, and the governance problems and the weaknesses inthe civil service had not
yet assumed alarming proportions. During the past decade or so, the weakness o f the overall fiscal
situation and stagnant or declining economic and social spending exacerbated the structural rigidities in
the budget process and compromised the allocative efficiency o f expenditure. Arbitrary expenditure cuts
undermined program- and agency-level performance, operation and maintenance expenditures were
severely under-funded, and development resources were spread too thinly as the overhang o f ongoing
projects became very large.

Recent Initiatives

4.11.       Faced with particular weaknesses in the budget processes and general problems o f
effectiveness o f public spending, the government has taken a number o f initiatives to improve planning
and budgeting, reduce leakage, and target public spending on key services and the poorer groups in the
society. It has moved strongly to strengthen fiduciary and accounting controls and improve the workmg
of Public Account Committees (see Box 4.1). It has a set up a Debt Policy Coordination Office to
strengthen the macroeconomic work in the Ministry o f Finance, integrate debt reduction goals with the
medium-term economic framework, and monitor the implementation o f the government's Debt Reduction
and Management Strategy. It has also initiated work toward a more comprehensive and detailed Medium-
Term Budget Framework (MTBF) with the help o f technical assistance from the UnitedKingdom. It has
made arrangements for estimating the considerable fiscal contingency obligations with respect to civil and
military pensions. Last but not least, the government has adopted a poverty reduction strategy with a clear
focus on direct poverty reduction expenditures and initiated the processes o f costing o f programs and
monitoring o fboth spending andoutcomes.

Medium-Term Budgetary Framework

4.12.       The move toward a comprehensive and detailed MTBF i s especially significant because an
appropriately designed MTBF can bring the various initiatives for better fiscal planning together and help
remove many o f the problems inthe budgetingand programmingprocess.




                                                   68

                              Box 4.1: Budgetaryand FinancialManagementReforms

In recent years Pakistan has undertaken a series of reforms to improve the effectiveness of public expenditure through
improvementsinitsbudgetmanagement andfinancial systems. The main elements ofthis reformincludethe following:

Devolution: The devolutionplanannouncedby the govemment aims at strengtheningthe local govemments. By taking the
government closerto the stakeholders,this moveis amajor step to enhance the accountabilityof public servicemanagement,
and thereby improvethe efficiency of the govemment. Giventhe low institutionalcapacity of the govemment, however, the
reform is not free of risks. The devolutionprocess, if not appropriately handled, may lead to marginalizing the status and
role of provincial govemments; elite capture of local govemments, breaking down o f subnational personnel management,
andconsequentlyofthe service delivery systems. Althoughwith some expected teethingproblems,to date the reformseems
to beworkingquite adequately.

Budgetary reforms: The government has launched an ambitious reform program to address the numerous budgeting
problems. The most important action in this regard is the initiation of a Medium-TermBudget Framework (MTBF). In
addition, accepting the recommendation of the Debt Committee, the govemment has established the Debt Policy
CoordinationOffice, which underthe Draft FiscalResponsibilityandDebtLimitationLaw would beresponsiblefor framing
the overallmacroeconomicframework, a functionrelatedto the MTBF.

Inorder to improvebudgetimplementation,significant financialautonomyand responsibilityhas been devolvedto the line
ministries throughintroductionof aNew System of FinancialControl and Budgeting(NSFCB), which delegates significant
financialpowersto the administrativeheads of spending agencies. This stephasbeenaccompaniedby the impositionofhard
budgetconstraintson the lineministries anddivisionsthroughexpenditureceilings. At the same time, the spendingagencies
have also beenmaderesponsiblefor ensuringfiscal andfinancialdiscipline, includingtimely reconciliationoffinancial data.
At the provincial level, the provinces of Sindh and NWFP, under their economic and structural reform programs have
ensured better implementationof provincial budgets through timely release of funds and removal of redundantprocedural
constraints. A better systemof monitoring andreportingofbudget implementationhasbeenput inplace. Moreover,the use
of personalledger accounts (PLAs) has beenseverelyrestricted.Fisca1transparency: The govemmenthas announced aplan
to improve the quantity and quality of fiscal information disseminated to the public to ensure fiscal transparency, and
thereby strengthen financial accountability. The govemment has endorsed a review of fiscal reporting and transparency
underthe IMF's Code of GoodPracticeson Fiscal Transparency. Consolidated fiscal accounts ofthe federal andprovincial
govemments (on the basis of reconciled data) are posted on Finance Ministry's web site on a quarterly basis. The
government of Sindh has started to post even its fiscal monitoringreport, which analyzes the implementationof the budget
andidentifieskey problemareas ofthe provincial government, onits web site every quarter.

Financial accounting: The govemment has made significant progress in establishingan integratedfinancial management
system, which can meet intemationallyaccepted standards of completeness, accuracy, reliability, timeliness, and usefulness.
The medium-term objective i s to accelerate the adoption of the new public chart of accounts and financial reporting
requirementsprescribedby the president in December 2000, all under anew computerized accountingsystemsupportedby
the Bank-fundedProjectfor ImprovementinFiscalReportingand Auditing (PIFRA). The govemment seeks to establishthe
public accountingsystemas arobustplatformfor transparencyinthe management ofpublic funds for: (a) economic decision
making within the govemment to promote program efficiency and effectiveness; and (b) provision of auditable financial
informationto various stakeholderswithin the countryas well as to donorsandintemationalfinanciers.

Auditing and legislative oversight. The federal and provincial Public Accounts Committees (PACs) have been made more
effective. For the first time inmany years, they reviewed the financial transactions of an incumbentgovemment and were
also able to significantlyreducethe backlogof auditreports. The quality of auditshas beenimprovedthrough systems-based
auditing proceduresand attentionto reportingonly on significant matters. And transparency of the PAC processhas been
enhancedby makingPACs' proceedingsopento the public.

Expenditure monitoring: The evolving PRSP includesa strongprogramof monitoring, including capacitydevelopment to
gather and analyze information for impact assessment of public policies and investments.The I-PRSP includesinstitutional
mechanisms for the regular/monitoring and evaluation of poverty reduction expenditures and corresponding outcome
indicators at the federal, provincial, and district levels (see 4.28-4.35). These institutional arrangements are presently
evolving and will be finalized after the round of consultations with provincial govemments is completed. Key outcome
indicators have also been suggested, and final benchmarks and targets will be set after consultations with national and
provincial statisticalagencies, federal line departments,andprovincial governments.




                                                             69

4.13.   The rationale for an MTBFi s that it provides the "linlung framework" that allows expenditures to
be "driven by policy priorities and disciplined by budget realitie~."~~tries to reduce the disconnect
                                                                                  It
between policy malung, planning, and budgeting. MTBF "consists o f a top-down resource envelope, a
bottom-up estimation of the current and medium-term costs o f existing policy and, ultimately, the
matching of these costs with available resources.. .in the context o f the annual budget pro~ess."~'The
"top-down resource envelope" i s fundamentally a set o f macroeconomic projections that indicate fiscal
targets and estimates of revenues and expenditures, including government contingency obligations. To
complement the macroeconomic projections, the sectors engage in "bottom-up" reviews that begin by
scrutinizing sector policies and activities with an eye toward optimizing inter- and intra-sectoral
allocations. The real value added o f the MTBF approach comes fi-om integrating the top-down resource
envelope with the bottom-up sector programs. (The steps required inpreparation o f an MTBFare given in
Box 4.2.)

                                    Box 4.2: Steps inPreparationof anMTBF

      A typical MTBFwould entailthe following steps:

          +Initiating a process of rigorous costing and analysis of existing policy commitments, with a view
               to improving efficiency and developing a fmbaseline cost for essentialcontinuing services of
               government. Maintaining a continuous medium-term view of these elements provides an
               essentialfoundation for a fullMTBF.
          +Developing a three-year macroeconomicframework as a startingpoint for the MTBF.
          +Utilizing improved techniques for macroeconomic analysis and revenue forecasting, and
               publishing the basis and assumptionsfor medium-termmacroeconomicandrevenueforecasts.
          +Establishing expenditure review teams with the task of reviewing all o fthe government's existing
               commitments (Part A of the budget) through a series of intensive studies of different sectors.
               The overall objective will be to establish a baseline cost for the continuing essential services of
               government. Once established, this baseline canbe projected forward andpublished as forward
               estimates inthe MTBF framework.
          +Developing broad aggregates for revenues and sectoral (and, for important subsectors,
               subsectoral) expenditure ceilings.
          +Making all ministries and departments responsible for extending their permanent and temporary
               budget forecasts beyond the budget year by another two years, and separately identifying
               additional costs over the three-year period that arise from implementation of government
               policies anddevelopmentprojects.
          *Strengthening capacity for sectoral and expenditure policy assessments. These assessments
               shouldtake into account relevantperformanceinformation.
          QMaking the budget for the following year in a "rolling" format. That is, inthe following year's
               budget, the existing policy estimate for the first outyear will be taken as the baseline estimate
               for the ministry innegotiating the budget, and similar procedureswill be followed eachyear.



4.14.    The MTBF i s intended to deliver a number o f ambitious outcomes: greater macroeconomic
balance; improved inter- and intra-sectoral resource allocation; greater budgetary predictability for line
ministries; and more efficient use of public monies. An improved macroeconomic balance, including
fiscal discipline, i s attained through good estimates o f the available resource envelope, which are then
used to make budgets that fit squarely in the envelope. MTBFs are expected to improve inter- and intra-
sectoral resource allocationby effectively prioritizing all expenditures (on the basis o f the government's
socioeconomic program) and dedicating resources only to the most important ones. Greater budgetary
predictability i s expected as a result o f commitment to more credible sectoral budget ceilings. The MTBF
                               4


46WorldBank(1998a: 32). The World Bank uses theterm Medium-TermExpenditureFramework(MTEF) rather thanMTBF.
47Formore onthe MTEF concept, see World Bank (1998a), Asian DevelopmentBank (1999), andDean(1997).


                                                          70

i s also expected to make public expenditures more efficient and effective, essentially by allowing line
ministries greater flexibility in managing their budgets in the context o f hard budget constraints and
agreed upon policies and programs.

4.15.     While the three pillars o f an MTBF are clear (projections o f the aggregate resource envelope, cost
estimates of sector programs, and a political-administrative process that integrates the two), the blueprint
for designing and implementing an MTBFmustbe developed ineach country context.

4.16.     InPalustan several planning and budget process issues will needto be revisited and resolved
before the MTBF can become a truly operational document. The relative responsibilities o f federal and
provincial governments, the sector ministries and departments, the Planning Commission and provincial
planning departments, and the Ministry o f Finance and provincial finance departments in relation to
various elements o f an MTBF estimation o f the resource envelope, the costing o f sector programs, the
choice o f intersectoral priorities, and the adoption o f specific targets must be looked at afresh and agreed
upon because they are in a state of flux. The arrangements for puttingtogether an MTBF will also need
to be harmonizedwith the preparation and implementation o f the final Poverty Reduction Strategy Paper
(PRSP).

4.17.     Given that Pakistan i s a federation, it i s critical for improved budgetary outcomes and enhanced
effectiveness o f public expenditure that MTBFprocess i s "rolled out', to all the provinces (and ultimately
to all the districts).  Ideally, the national MTBF would be a consolidation o f subnational and federal
MTBFs. Itis essential, therefore, that the process o fpromotingprovincial adoption o f MTBFbe ledby a
strong commitment from the center.

4.18.     Fortunately, two provincial governments (Sindh and NWFP), under their economic reform
programs (supported through World Bank structural adjustment credits), have already made a start toward
formulating detailed MTBFs. Although the federal govemment, with technical assistance from the British
Department for International Development (DFID) (see Box 4.3) has embarked upon expanding its MTBF,
there i s a strong need to base all three MTBFs(and indeed the future MTBFs o fPunjab and Balochistan) on
a sound institutional footing.

4.19.     The introduction of the new MTBF and PRSP processes provides an important opportunity for
strengthening overall planning in Palustan and improving policy coordination among the key economic
ministriesand departments both at the center and inthe provinces andbetweenthe federal government and
the provinces. This change will not, however, happen automatically. The role o f the government needs to be
redefined inthe light o f new approaches. The objectives should be to ground the new processes innot only
better estimation o f resource availabilities and flows, but also greater knowledge about developments inthe
real economy, sectoral trends, and realities o fproject andprogramimplementation.

4.20.     The role o f provinces in developingrealistic budgetary frameworks and viable poverty reduction
strategies can hardly be overemphasized. As discussed later, the federal government must devolve greater
authority to the provinces with respect to economic and social development and eliminate or limit federal
projects and initiatives in a number o f sectors, such as education and water, as part o f the overall process
o f moving services closer to the people. This devolution from center to the provinces will give impetus to
the major initiative o f devolution to the district level. Inany case, it is clear that it i s the provincial- and
district-level spending on education, health, law and order, provision o fjustice, irrigation, rural and urban
water supply, and roads that will account for bulk o f the increase in public spending in the near future.
The institutional arrangements that will integrate financial, policy, project, and sectoral plans at the
provincial and district level clearly need to be strengthened. The role o f the provincial planning
departments should be broader than the present excessive focus on development budgets alone, which are
a fraction o f provincial spending. At the same time, planning departments will need to work closely with


                                                         71

both finance and implementing departments and agencies in costing programs, developing yardsticks o f
spending, and monitoring o f expenditure implementation, impact, and outcomes.


                   Box 4.3: DFIDAssistance to Strengthen the Medium-Term Budget Framework

   The British Department for Intemational Development is providing technical assistance (TA) to the govemment of
   Pakistan (in the form of intemational and local experts) to strengthen its attempts to develop a Medium-TermBudget
   Framework (MTBF). At present the budget framework does not go beyond presentingwith the budget documents a
   three-year macroeconomic forecast, which is updated annually. The TA, which is expected to be spread over thirty
   months, will initially focus on the health and population sectors but subsequently will be rolled out to cover other key
   sectors.

   The Objectives
   The main objectives of the TA will be to help the govemment improve the quality, transparency, and accuracy of
   Pakistan's medium-termbudgeting, provide a firm basis for transmittingpolicy into effective expenditures, and thereby
   improve federal expenditure by bringing credibility and stability to budget plans, along with improved estimation of
   transfers and allocations to subfederal tiers of govemment. The TA i s expected to improve the quality of short- and
   medium-termbudget forecasts, including revenue forecasts and rolling estimates of the costs and liabilities of existing
   andnewpolicies, drawingon intemationalbestpractices. Moreover,the TA will helpthe govemment developnecessary
   institutionalcapacity to prepare, implement, expand, anddeepenthe MTBF inthe future.

   Institutional Setting
   In the first year the MTBF will be piloted in the Ministries of Health and Population Welfare. In subsequent years
   MTBF planningwill be rolled out to other line ministries such as education, and transport and irrigation, which require
   adequate recurrent resources for the operation and maintenanceof large physical assets. As such, althoughthe Budget
   Wing of the Ministry of Finance will be the main counterpart agency for the TA, the process is expected to intimately
   involve the Expenditure Wing of the Ministry of Finance, the Planning Commission, the Central Board of Revenue
   (CBR), the Ministry of Health, the Ministry ofPopulationWelfare, andother key federal line ministries

   The TA Components
   (a) Strengtheningthe Resource Framework: The TA will help the Ministry of Finance, the CBR, and the Planning
   Commission develop sound analytical foundations for the macroeconomic framework, better estimate contingent
   liabilities, estimateandincorporatequasi-fiscalactivities,andstrengthen short- andmedium-termrevenueforecasting.
   @) EstablishingInterSectoralPolicy and Strategy: The TA will help the govemment undertakebettertechnicalwork
   and help establish stronger institutional mechanismsto link technical work with political adjudication of intersectoral
   policypriorities.
   (c) Supporting Line Ministries: The TA will help with line ministries (initially focusing on the Ministries of Health
   and Population Welfare) to: (1) develop sectoral policies and strategies; (2) translate the policies and strategies into
   programs with well-defined objectives and outputs; (3) develop techniques and procedures to estimate program and
   projectcosts; (4) preparecomplementaryandconsistentrecurrentbudget anddevelopment budget proposalsinanMTBF
   format; and (5) prioritize allocationsto programsaccordingto resourceavailability.
   (d) Securing Improvementsin Budget Implementation: The TA will also helpthe Ministry of Finance, jointly with
   line ministries, to review budget implementation procedures, specifically the new system o f financial control and
   budgeting, and secure amoreefficient flow of funds mechanism.
  I


4.21.     At the federal level the Finance Ministry in general, and the Budget Wing, inparticular, will take
the lead in malung the federal MTBF, but much improved coordination would be required between the
Ministry of Finance, the Planning Commission, the sectoral ministries, the Central Board of Revenue
(CBR), the Economic Affairs Division (EAD), and the Accountant General o f Pakistan Revenue (AGPR)
for improved formulation and execution o f MTBF. The respective roles o f the Planning Commission and
the Finance Ministry in developing a macroeconomic framework and sectoral plans and targets need
clarification. Though some overlap in functions between economic ministries may be unavoidable, and
indeedhealthy, the overall lack o f capacity inall parts o f govemment and systemic problems with the civil
service suggest the need for a better delineation o fresponsibilities and clearer coordination arrangements.

4.22.     Preparation o f macroeconomic frameworks i s a case in point. The Planning Commission has
traditionally had a mandate to prepare medium- and long-term development plans and for that purpose


                                                             72

also prepares a medium-term economic framework. Over time, however, the operational relevance o f
these medium-term frameworks declined as fiscal crises forced more ad hoc approaches to resource
estimation in the Ministryo f Finance. Recently, the Ministry o f Finance has undertaken two initiatives to
upgrade its macroeconomic work. First, a macroeconomic framework i s to be prepared as a part o f the
MTBFinthe expenditure wing. Second, the newly created Debt Policy Coordination Office (DPCO) has
also been charged with the task o f preparing a macroeconomic framework under the Draft Fiscal
Responsibility and Debt Limitation Law. There i s the obvious need for better coordination within the
Ministry of Finance on the macroeconomic framework and between the Ministry and the Planning
Commission, The latter can bring to bear its expertise in the real economy and the sectors in the
preparation o f macroeconomic projections. To further this coordination, it i s important that at the present
juncture, the Ministry o f Finance and Planning Commission remain under a unified leadership, preferably
under the same Federal Minister. A unified leadership approach to planning and budgeting has been tried
inquite a few EastAsian countrieswith considerablesuccessinacceleratingeconomic decision-malung.

4.23.   Development o f an MTBFinvolves much more than the estimate o f the resource envelope. It also
involves substantial work on sectoral and subsectoral priorities, costing o f programs, monitoring o f public
spending, and assessment o f outcomes. In these and related areas the Planning Commission, working
closely with the provincial planning departments, can play a major leadership role, quite distinct from its
responsibilities for appraisal and approval o f large development schemes.

Re-invigoratingthe PlanningProcesses

4.24.   The preparation, monitoring, and evaluation o f MTBFs and PRSPs, at both the federal and the
provincial level, provide a unique opportunity for invigorating Palustan's planning processes, which in
their heyday inthe 1960swere considered a model for developingcountries (see Box 4.4).

4.25.   Duringthe past two years progress has been made inproviding the Planning Commission with
greater authority in project appraisal, shifting the focus o f macroeconomic and sector work away from
Five-Year Plans to a ten-year perspective, and setting up rolling Three-Year Plans. Project monitoring has
been made systematic and i s shared with public, and the poverty work has been substantially
strengthened. The recent improvements in the planning processes have been grounded in the general
desire o f the Musharraf govemment to give a larger role to technocrats in policy and decision-making.
Butclearly the institutionalcapacity across economic ministries andplanning agencies remains extremely
weak. There i s excessive reliance on outside agencies for analysis. Significant improvements in sectoral
planning and macroeconomic and budgetary expenditure frameworks are neededboth at the center and in
the provinces.

4.26.   H o w can the planning processes both at the center and the provinces be strengthened? First, at
the center there must be both a closer collaboration and a clearer division o f responsibility between the
finance division and the planning division. For instance, the planning division could play a more modest
role in preparing a macroeconomic framework but should have the major responsibility for setting the
sectoral priorities and development allocations.

4.27.   The PlanningCommissionenjoys two distinct advantages vis-a-vis other ministries:

        (a) It still has more competent technical staff to deal with all the sectors o f the economy and has a
             long tradition o f interactionwith line ministries and provincial departments.
        (b) Iti s not subject to the day-to-day pressures that tend to preoccupy the Finance Ministry.




                                                       73

                                Box 4.4: PlanningProcesses: Sourcesof Weaknesses

     Pakistan's strong economic record o f the 1960s was attributable in part to the strength o f its planning
     processes and the unique role o f its planning institutions. Contrary to popular perceptions, it was not the
     existence o f Five-Year Plans as such that was the main contributor to the quality and speed o f economic
     policy adjustments and generally sound resource allocation decisions during the 1960s. More important
     factors were the strong political support to the planning processes, the high level o f professionalism and
     independence o f groups o f planners both at the center and in the provinces, and the economic policy
     coordination mechanisms, which, notwithstanding the centrality o f the MinistryiDepartment o f Finance,
     provided a clear role for the planning agencies as well as the sector ministries and departments. Serving as
     high-powered economic and technical staffs, the Planning Commission and the provincial planning
     departments were very effective in analyzing key economic development issues, recommending
     adjustments in policies and programs, and assisting the top leadership and the finance minister in both
     economic policy coordination and foreign aid mobilization.

     There is general agreement that the planning processes have greatly weakened over time even though Five
     Year Plans have formally been in existence throughout most o f the past three decades: the EighthPlan
     covered the period 1993-98, and the NinthPlan for 1998-2003, though prepared, was never finalized.

     Factors responsible for the decline in the effectiveness o f both the planning processes and the planning
     .
     institutions have varied over time but include the following:

       Reduced political willingness over time to accept the financial and economic discipline o f independent
       economic and project analysis; in the 1990s large projects in key sectors were taken out o f the
     . jurisdiction o f the Planning Commission, and only inthe past two years have the powers o f the Planning
       Commission to review all large projects have beenrestored.
     . Excessive focus o n preparation o f Five-Year Plans and not enough attention to analyzing the causes of
       continued failures to acheve planning targets innational savings, exports, and social sectors.
       Frequent lack o f realism o f the macroeconomic frameworks prepared by the Planning Commission,
       contributing to excessive new commitments on projects and subsequent spreading o f development
     ..expenditures too thinly and frequent inadequate finding o f foreign-aided projects.
       Insufficient attention to sector plans and the mix betweenpolicy requirements and investment needs.
       Poor links betweenresource needs and physical targets.
       Lack o f balance o f effort between planning o n the one hand and monitoring and evaluation o n the other;
     . even where monitoring has taken place, it has been excessively concerned with inputs o f money rather
       than achevement o f set goals, and impact analyses o f development spending have beenrare.
       Last but not least, professional capacity in all agencies concerned with development has become very
       weak; little new professional staff has been inducted into the Planning Commission during the past ten
       or fifteen years.



4.28.     The Planning Commission should cede the existing de facto role o f overall economic
coordination to the Finance Ministry but keep primary responsibility for formulating sectoral policies,
plans, and strategies. These sectoral plans need to be developed in collaboration with the line ministries
and departments. The Planning Commission (and provincial planning departments) would be responsible
for intersectoral priorities inthe MTBF for development. This implies that the Planning Commission (and
planning departments) should be responsible for determining the allocations at the sector and agency
level. For this purpose, while the Public Sector Development Program (PSDP) needs to be formulated
within the resource envelope provided by MOF, the recommendations of the Planning Commission on
budgetary allocations for individual projects, programs and sectors may be treated as final.




                                                           74

4.29.     The Planning Commission and the planning departments should lead the effort to integrate
recurrent and investment expenditures in a coherent MTBF. The preoccupation o f planners with the
development budget i s hampering the pursuit o f balanced strategies and speedy social development. To
ensure progress, the policies formulated under the major initiative o f PRSP need to be consistent with
sector-specific policies and total sector allocations. Full integration requires the following: (1) the same
strategic objectives should drive the formulation o f the recurrent and investment budgets; (2) the budget
planning process at the sector level should consider options to achieve these objectives and implications
for recurrent or investment expenditures based on a predefined resource envelope for the sector; (3)
project selection should be based on a range o f economic and non-economic criteria, especially on the
defined   role o f the government within the sector; (4) the best projects to realize the sector programs
should be selected on the basis o f adequate cost-benefit and cost-efficiency analysis; and (5) monitoring
and evaluation o f the budget programs should be based on a unifiedbudget classification structure that
allows sensible analysis. The full integration o f the recurrent and capital budget planning processes i s a
gradual and long-term process. As a first step, however, there i s a need to improve the estimation o f the
recurrent cost implications o f the ongoing investment projects from the line ministries to be reflected in
the MTBF sector expenditure proposals. Equally important, organizational arrangements for recurrent and
investment planning inthe line ministries, the Ministry o f Finance, and the Planning Commissionneed to
be realignedto eliminate the present separation. Only inthat way will investment projects be an integral
part o f the MTBF.

4.30.     The Planning Commission should develop a rolling program o f sector reviews, possibly three to
four sectors a year, which will provide depth and credibility to its long-term plans. Such sector reviews
could be a vehicle for worlung with the World Bank, the Asian Development Bank, and other donors on
analytical issues

4.3 1. More concerted efforts to use monitoring and evaluation at the project, sector, and economy level
are needed to monitor progress toward key goals and to better understand the constraints hampering
economic and social progress. Systems being put inplace to monitor PRSP can provide an excellent step
inthis direction.

4.32.     The goals listed cannot, however, be achieved without substantially restoring professional
strength. The Planning Commission should opt for a leanbuthighly skilled staff. For some of the routine
functions o f project appraisal, it could rely on specially constituted technical panels drawing in part on
highlyprofessional experts from outside.

Strengthening Expenditure and Program Monitoring

4.33.     One key factor behind the declining effectiveness o f public expenditure i s a dysfunctional
expenditure monitoring system. The traditional budgetary monitoring system has focused on ensuring the
spending o f the budgetary allocations and the procedural correctness o f these expenditures. Little, if any,
attention i s paid to service delivery and client satisfaction. Nevertheless, appropriate systems were put in
place within the Ministry o f Finance (finance departments at the provincial level), the EAD, and the
Planning Commission (planning departments inthe provinces) to achieve at least the limited objective o f
public expenditure monitoring. For example, the Budget Wing o f the Ministry o f Finance reviews the
current and development expenditures on the basis o f data received from the Controller General o f
Accounts (CGA) and project entities on monthly basis. The EADundertakes a periodic review o f donor-
assisted projects and programs.       The Programming Section and the Projects Wing o f the Planning
Commission undertake monitoring o f financial and physical progress o f the development projects. Not
only was the quality of data usedby these monitoring exercises declining steadily overtime, but also the
interest o f decision makers in using the results o f these exercises in their planning and budgetary




                                                        75

decisions. Partly because o f falling demand, but mostly because o f the inadequate human and financial
resources allocated to it, the quality o fmonitoring itselfhas also been going down.

4.34.    The Planning Commission has recently revived monitoring o f development projects through in-
depth quarterly reviews o f key projects and programs. Although a step in the right direction, this process
has yet to factor prominently ininducing the overall economic decision making.

4.35.     Meanwhile, the I-PRSP has proposed a more elaborate mechanism to monitor poverty-related
expenditures, along with their intermediate indicators and outcomes, vis-A-vis the specific targets set by
the government for human development and poverty reduction. When fully functional, the proposed
mechanism can provide improved and continuous monitoring o f PRSP inputs and outputs, which would
help to validate the choices made inthe prioritizationprocess and can reveal the reasons for the success or
failure o f the various interventions chosen. Further, it can offer warning signs to those responsible for the
management o f the Poverty Reduction Strategy (PRS) when interventions are not having the desired
effect or are having undesirable side effects. If undertaken correctly and diligently, the proposed
monitoring system can facilitate management decisions about continuation, acceleration, or amendments
o f various interventions. Moreover, an effective monitoring system o f this kmd can help develop greater
accountability inthe use o fpublic resources.

         ..
4.36.    The proposed mechanism includes following areas to be monitored:

         ...Budget implementation and public expenditure
            Policy reforms at a number o f different levels (macroeconomic, structural, and sectoral)
            Public action choices
            Public services (quality, availability, access, satisfaction)
            Investment programs

4.37.    With technical assistance from DFID and the World Bank, the government has identified
indicators at a number o f different levels along the "impact chain," starting with inputs and outputs,
followed by outcomes and impacts. Special emphasis has been given to identifying the intermediate
outputs, which bring into the government strategy elements o f accountability, learning and changing
actions, and intervention. The system requires monitoring PRSP expenditures on quarterly basis at the
federal, provincial, and district

4.3 8.   Regardingmonitoring longer-term outcomes, the I-PRSP presents benchmarks on the present-day
status of social indicators and establishes medium-term targets for a variety o f indicators, such as infant
mortality rates and immunization rates. For these benchmarks and targets to have any meaning, however,
they need to be underpinned with a reliable and credible database, as well as backed by the capacity to
interpret and analyze the data. While Pahstan has both administrative and survey data sources, but both
have their limitations, and in any case show wide differences inthe present status o f health and education
levels. Administrative data are incomplete and unreliable, and household survey data are representative
only to the level o f the province (not district). There i s a considerable lag between data collection and
data availability.

4.39.    In order to remove these deficiencies, a new survey instrument Core Welfare Indicator
Questionnaire (CWIQ) has been developed to monitor changes inkey social and poverty indicators on an
annual basis at the national, provincial, and district level. The primary objective i s to keep track o f the
short-term changes in both the intermediate outputs and outcomes. The Pakistan Integrated Household



48Inaddition, the MOF has alreadystartedreportingontwo indicatorsofintermediateoutputs on aquarterly basis.



                                                         76

Survey (PMS) remains the key vehicle for providing information on medium-term changes inhuman and
poverty indicators.

4.40.    Because PRSP monitoring i s a multidimensional task, it i s being undertaken by a collaborative
effort o f a large number o f government agencies, including the Ministry o f Finance, the Controller
General o f Accounts and the provincial accountant generals, the line ministries and departments, and the
Federal Bureau o f Statistics. Given its vast comparative advantage inpoverty research and monitoring o f
public expenditure and projects, the Planning Commission has been assigned a key role in monitoring
PRSP outputs and outcomes. It i s essential, however, to improve the database and strengthen the
analytical capacity of, and cooperation between, the involved agencies to institute an effective monitoring
and evaluation system for the government's poverty reduction strategy.

Redefiningthe Roleofthe State

4.41.    Improving the effectiveness o f public services and expenditures requires not only mending the
institutional framework, but also reviewing the structure and functions o f the government. First, the role
o f the state in the economy must be tailored to its institutional and financial capability and the market
economy alternatives available, because inmany areas o fprovision o f even public goods and services, the
private sector works more efficiently. Second, even in functions that should ideally be in the
government's domain, it may be possible to move away from the over-centralization o f responsibility at
the federal level that has taken place.

4.42.    Apart from defense, foreign affairs, and natural resource management, the federal government
has a strategic role in economic management that has become and will become highly complex in an
integrated world economy. The state at the central level must be responsible for stabilizing the economy,
ensuring fiscal discipline, containing inflation, balancing the external accounts, and keeping the debt
burden within sustainable limits. It must protect the consumers' and stakeholders' interests by regulating
the private sector, facilitate investment, provide a level playing field to economic actors, curtail rent-
seeking behavior, preserve its autonomy against intrusion by pressure groups, and ensure a competitive
environment, vital for survival and growth in a rapidly globalizing world. Inaddition, it must provide an
effective and reliable legal framework and recourse institutions for conflict resolution. It must curtail
transaction costs by promoting transparency and integnty and imparting greater predictability to the
system. It must also reduce information rentals by facilitating access to policies, plans, and data on
economic outcomes and trends.

4.43.    Just this strategic agenda i s formidable and even the best would be hard put to deliver. Hence,
extending its energies and resources to activities and functions which could be well-performed by the
private sector andor sub-national levels o f government, the central government undermines not only the
development o fprivate sector, but also its own ability to effectively deliver on its strategic agenda.

4.44.    According to the 1973 Constitution the federal government i s responsible for, key areas like
defense, foreign affairs, banlung and currency, postal service, transportation (ports, airports, railways),
international trade, macroeconomic management and industrial development. The provinces share with
the federal government responsibility for population planning, irrigation, curriculum and syllabus
planning, and social welfare, and are solely responsible for roads, highways, law enforcement (police)
and dispensing o fjustice, education and health. Under the LGOs 2001, the provincial governments have
devolved a large portion o f their responsibilities to the local (district, tehsil and union) governments.

4.45.    InPalustan, while the legislativedivision ofpowers, as set out bythe Constitutionandprovincial
statutes, compares favorably to the assignment based on principals o f fiscal federalism, in practice the
federal government has expanded itself in areas o f shared responsibility (and even in fields which


                                                        77

constitutionally belong to sub-national governments), and have even taken over some o f the traditional
private sector functions.

Reducing theRole of the Government in Producing and Providing Private Goods and Sewices

4.46.      Successive recent governments have already attempted to reduce the role o f the state in the
economy through privatization o f state enterprises and encouraging private sector into areas that had
exclusively been reserved for the public sector. Nevertheless, the government's role in the economy still
transcends beyond the optimal and its involvement in many sectors i s still excessive vis-his the
appropriate role defined by economic principal^,^' its administrative capacity and fiscal resources. There
i s significant room for greater private sector involvement in the fields o f telecommunications; electricity
generation and distribution; gas and oil exploration, production, and distribution; electronic media;
banking, finance, and insurance; transportation (including air, sea, and rail traffic), and others. Indeed, the
present policy o f the government i s to retain only the regulatory functions and to gradually vacate all o f
these areas for the private sector. The government should accelerate the process o f privatization to
minimize the deterioration in performance and asset-stripping that may occur once the privatization
process i s started. Lack o f an adequate regulatory framework should not be used as an excuse for delaying
privatization. Although appropriate regulation o f the private sector, especially o f the privatized state
monopolies, i s essential, it should be possible to develop the regulatory capacity o f the government
concurrently with privatization.

4.47.      Efficiency of non privatized organizations can be enhanced by: (1) giving them autonomy over
their input and output decisions, thereby increasing their allocative and productive efficiency; (2)
protecting them from political interference, to enhance the efficiency o f their personnel decisions; (3)
subjecting them to a hard budget constraint to ensure that the resources are used efficiently without any
waste or leakage; and (4) providing their management and staff a performance-based incentive structure.

4.48.      Other than the sectors already cited for reduced public sector involvement, it may be appropriate
for the government to reassess the needs and roles o f different kinds o f public sector entities (such as
government departments, attached bodies, autonomous bodies o f various types, and commercially
oriented public enterprises) with a view to making the overall size o f government more consistent with
available resources and focused more intensely on the govemment's strategic agenda. The decision-tree in
Figure 4.1 can help in making a quick assessment o f whether a particular governmental function should
be shed or continued.

4.49.      There i s also a need to re-examine the role o f government as a producer/provider vis-A-vis as a
financier and facilitator o f services. An area, which has not been fully explored in Palustan, i s the
possibility o f inducting the private and non-government sector in service delivery through public-private
partnership. Despite the emerging possibility o f larger fiscal space becoming available, both from
financing but more importantly from efficiency point o f view it i s necessary that the government should
actively pursue private sector in all those sectors where it can deliver service more efficiently. Inall these
areas, the government responsibility may be limited to regulating, and in some cases financing, the
delivery o f service through private providers, so as achieve its equity and other social objectives. Inmany
cases, it i s more efficient for the government responsible to provide only finances to the private sector to
provide the service. The financing mechanisms therefore need to be developed that attract the private
sector inprovision of public services but at the same time are able protect the poor and other vulnerable
groups, for being excluded from private delivery system. Box 4.5 provides examples o f such financing
mechanisms inthe case o f delivery o f education.



49 Stepsrequiredto decide onwhat arethe core functions ofaparticulargovemmentare highlightedinFigure4.1.


                                                          78

                   Box 4.5: Optionsfor PublicFinancingof Private Delivery of Education

Public funding for private schools: In many countries, such as Canada, Chile, and the Netherlands, the
government promotes education by providing fiances to the private schools. The recent spread o f private
schools in the urban and, to a lesser extent, rural areas o f Pakistan offers a possibility for the government of
better use the private sector to further its objective o f a rapid spread o f school education.      This objective
requires providing public financing to the private schools in targeted poor and backward areas and
neighborhoods on a per-student basis. A s these payments are based on the average cost o f education in the
public sector, while expenditures per student (for comparable education) are significantly less in the private
sector, the private schools have a strong incentive to vie for new students, thereby raising school enrollment in
private schools.

Public schools under private management: The government may choose to contract one of more qualified
nongovernmental organizations (NGOs) to manage a certain number o f public schools. An ideal contract would
give the managing NGO the right to appoint the head teacher and school staff. The teachers would have
qualifications and salaries comparable to those of the contract teachers o f public schools. The government
would bear the salary cost and provide to the NGO the nonsalary operational cost on the basis of a per-student
average o f the public schools' cost. Such a mechanism would imply no additional budgetary impact to the
government, while the NGO would have every incentive to runthe school in a most efficient manner and raise
the quality o f education to bring in more students. Such an arrangement would be ideal for enhancing the
outreach o f education to remote or backward areas where it i s difficult for the public sector to operate.

Giving parents a choice (the voucher system): One option for improving school quality is to give parents a
choice through a voucher system. Under this system individual students would be given vouchers funded by
public revenues but redeemable ineither public or private schools. Schools would have to compete for student
patronage, mainly through raising the quality o f education. Such a system has been tried with considerable
success inmany developed countries (such as the United States) as well as developing countries (such as Puerto
Rico).

All these means of public-private partnership require strong monitoring to protect the interests o f the
stakeholders. Such monitoring canbe established only through the participation o f the community.




                                                      79

                   Figure 4.1:Decisiontree for identifying core governmental tasks


                              I Disaggregate functions
                                     Policy
                                     Co-ordinanon
                                     Service deli\ ery
                                     support
                                     Reeulation




                                                                   No                      I
                                 Does thefunction support other
                                    governmentalpriorities?




                                 be achieved through regulation,                 No
                                    taxation, transfers to sub-




                                   oes national or internationa
                                 law or the Constitutionrequire




                 (iii)                                              No
                         Yes




Source: Methodologyfor FunctionalReviews, Paperproducedby the RussianFederationGovemment, World Bank, July 2001.



                                                              80

4.50.     Because o f their public goods nature, many health services (preventing and treating infectious
diseases, providing safe drinkingwater, generating and disseminating health information, and preventing
non-communicable diseases) will be underprovided if left exclusively to the market forces, as public
benefits far exceed private returns. The government has a natural role in providing all these services.
Moreover, the government should also get involved inproviding a safety net for the poor and for hard-to-
reach and difficult-to-treat, unprofitable     segments o f the population. Thus the public sector role in heath
care should center on activities such as providing immunization, controlling disease vectors such as
mosquitoes and rats, and guaranteeingbasic care for the poor and excluded. Beyond these areas, there are
ample examples o f high-quality health care services beingprovided and financed efficiently and equitably
by both public and private sectors. The government has only to ensure appropriate incentives to the
providers, public or private, to guarantee efficiency. A recent attempt to promote public-partnership in
health care i s inthe area o f detection and treatment o f cases o ftuberculosis (see Box 4.6)

                            Box 4.6: Public-Private Partnership inTuberculosis Control

 Pakistan has more than 250,000 new cases o f tuberculosis every year, o f which only 20 percent are detected. In
 1994/95 the World Health Organization recommended a new strategy, "DOTS" (Directly Observed Treatment - Short
 Course) to eradicate TB globally. DOTS is an inexpensive and effective approach, but unfortunately it covers only 10
 percent o f TB patients in Pakistan. Only 20 percent o f Pakistani TB sufferers seek care from the public sector, and about
 40 percent go to the private sector, mainly to general practitioners and NGO health units. The remaining 40 percent
 either do not seek any care or go to other nonprofessional providers. Unfortunately, the nonstandardized TB practice in
 the private sector leads to very low cure rates. In addition, the infectious nature o f the disease leads to increased
 transmission o f the TB germs to others, and inferior-quality treatment creates the risk o f multi-drugresistant (MDR)
 TB.

 Giventhe nature of care sought by most TB patients, it is suggestedthat a public-private partnership would be an ideal
 approach for treatment of TB. The National TB Program (NTP) i s seeking acceptable and feasible mechanisms to
 ensure closer working between the government and private partners to extend TB care according to national standards.
 Poolinghuman and financial resources inthe two sector has enormous potential for nationwide TB control.

 Involvement o f private sector providers in implementing DOTS in Pakistan has been recommended in a nationally
 agreed strategic framework.      Consultative processes are ongoing, and there are a few good experiences in some
 provinces. Successhl implementation of the DOTS strategy through collaboration between the Mary Adelaide Leprosy
 Center (MALC) and the Government of Azad Jammu and Kashmir, under the World Bank-financed Northern Health
 Project, i s one such example.

 The key components o f the public-private partnership project in TB control, to be implemented mainly at the district
 level, include the following:

      Capacity-building ofthe private and public sectors: Because the majority o f TB patients seek care from private
     practitioners, the "IT aims at training the private doctors, paramedics, and laboratory technicians in standardized
      good-quality TB treatment. Inaddition, the private sector will be provided with the latest equipment (microscopes)
      and chemical reagents to facilitate detection.     The public sector, too, will be upgraded: district health offices,
      supportedbyprovincial authorities, will be equipped to develop and sustain partnerships neededto increase DOTS
      implementation.
   1 Advocacy: Media messages to promote awareness about health issues will be developed at the federal level.
   1 Research and evaluation: Provincial officials will participate in development, research, and evaluation activities in
     their respectiveprovinces.



4.5 1. Similarly, given the importance of agricultural extension in enhancing agricultural productivity,
and the virtual collapse o f public extension services, there i s considerable scope for public financing o f
private delivery o f extension services. In the Palustani context, such private extension service providers
might include unemployed agricultural graduates, agricultural consultants, consultancy firms, progressive
farmers, farmers' organizations, cooperatives, nongovernmental organizations, agribusiness companies,



                                                             81

inputdealers, newspapers, private television channels, andprivate sector banks. The key is to involve all
or some o f them in the agricultural development process through appropriate incentive mechanisms,
including direct financing from the budget.

4.52.      The private sector can also be inducted in the road sector through appropriately designed
concessions. Information technology i s a fast-developing area in Pakistan, having relatively large claims
on public funds. According to international experience, the sector requires public involvement in
regulation and facilitation more than in development and provision o f information technology, which can
be well and efficiently performedby the private sector.

4.53.     A number o f services even in the subnational domain might be better provided by the private
sector, including fire protection, refuse collection, street maintenance, traffic management, local transit
service, refuse disposal, and electric power. Inaddition, private sector participation i s useful in primary
and secondary education, transportation, water supply, sewerage, and health services (see World Bank
1994 for details on such participation).

Furthering Devolution

4.54.      Perhaps the most important initiative o f the                       Figure 4.2: The 2-step logic of decentralization
government i s its comprehensive plan o f restructuring the
                                                                                         -1
government through devolving the state authority Table
4.1 presents the areas o f responsibility o f each level o f                             Structural changes:

local g~vernment.~' The plan attempts to create full-
fledged district and municipal (tehsil)                 governments                          administrative
having legislative and financial powers, with the main
objective o f changing the political and administrative
culture of the government by taking government closer to
the people and therefore getting communities better                                     N e w accountabilities for
integrated with the government. This, it i s hoped, that the                             senior staff
                                                                            I
devolution process would improve decision-making and
government's        accountability, particularly o f service
providers, to the citizens.                                                              Service delivery improvement     I
4.55.       The overall objectives o f the current devolution                          I arrangements:                    I
program in Palustan can be discemed from the various                                         supply-side
                                                                                             demand-side
publications       o f the     NRB, the local government
ordinances, and their amendments. Some interpretation
i s required, as the objectives are not set out in any single document. One point i s clear from government


50 Inaddition, city districts may set updistrict municipal offices for the integrated development and management o f the following
responsibilities jointly with the Tehsils: (1) water source development and management, storage, treatment plants, and macro-
distribution; (2) sewage tertiary and secondary network, treatment plants, and disposal; (3) storm water drainage network and
disposal.; (4) flood control protection and rapid response contingency plans; (5) natural disaster and civil defense planning; (6)
solid waste management, treatment, and disposal, including landfill cities and recycling plants; (7) industrial and hospital
hazardous and toxic waste treatment and disposal; (8) environmental control, including control o f air, water, and soil pollution in
accordance with federal and provincial laws and standards; (9) master planning, land use, zoning and classification, and
reclassification; (10) urban design and urban renewal program; promulgation o f building rules and planning standards; (11)
parks, forests, playgrounds, and sporting and other recreational facilities; (12) museums, art galleries, libraries, and community
and cultural centres; (13) conservation o f historical and cultural assets; (14) landscape, monuments, and municipal
omamentation; (15) urban and housing development, including urban improvement and upgrading, and urban renewal and
redevelopment, with care being taken to preserve historical and cultural monuments; and (16) regional markets and citywide
commercial centres. Moreover, unions can receive functions from districts and Tehsils, provided that they are also given the
funds for managing those functions.



                                                                  82

descriptions o f devolution: it i s intended to improve service delivery and i s not an end initself. Keeping a
focus on the intended end product--improvements in service delivery--is crucial, as the literature i s clear
that decentralization, per se, offers no guarantees o f improvement in accountability or service deli~ery.~'

4.56.     Devolution i s an attempt to change incentives within the public sector through a pragmatic
combination o f administrative, fiscal, and political changes. These changes are intended to change how
senior staff o f subnational governments understand their tasks-that is, changing their perceived
accountabilities so that they feel responsible for achieving service delivery improvements. Those service
delivery improvements are likely to be variable between districts and across sectors and thus difficult to
specify precisely in advance. They are also likely to include both supply-side reforms, improving the
quality o f health or education services that are produced, and demand-side, strengthening citizens' ability
to pressure government for better-quality services.

4.57.     The key assumption i s not that these service improvement arrangements could not happen within
a centralized system. The logic i s that malung any such arrangement work requires effort and ingenuity
on the ground, and changing the accountabilities o f senior staff in local government motivates them to
install and sustain these arrangements.

4.58.     Primary education offers an example o f this chain o f connections. Passing authority over teacher
recruitment and discipline to competent local school management committees can be a step toward better
educational service delivery. Establishing local govemance arrangements for schools could in principle,
however, be undertaken by central government as well as by subnational government.                                                The
decentralization argument i s that a judicious mix o f political and fiscal changes, coupled with some
decentralization o f hire and fire authority over teachers to local government, will be the trigger that makes
nazimeen and senior district officials more concerned with making these arrangements work.

4.59.     The same example illustrates why service delivery improvements require better local policy as
well as improvements in service quality. Improvements in the quality and attendance o f primary school
teachers are rather marginal ifpatronage and elite capture significantly direct public spending toward
secondary education for the middle classes, rather thanprimary education for the poor.

4.60.     Initial evidence concerning community oversight at local level i s encouraging. Many school
management committees seem quite well developed and have had considerable success in reducing the
cost o f school construction. InSindh, under the Khairpur Shelterless School Scheme, thirty-three schools
have been completed at a third of the cost of a comparable government school, in quicker time, and
without compromising on the quality of construction. It seems that the elected officials were crucial for
mobilizing the communities and ensuring the effective functioning o f these institutions. Inaddition, some
early signs show that the education monitoring committees are having an impact through surprise visits to
schools and checks on teacher attendance through direct observation. It i s important to build on these
early successes and, mindful of the risks of overclaiming, to publicize concrete examples o f successes if
some momentum i s to be maintained.


5'A recent review of the literature found that the empirical evidence that decentralization increases allocative efficiency inpublic
expenditure, accountability, or cost recovery i s quite meager.          The literature suggests that decentralization can have
disadvantages, including failures o f policy coordination, "elite capture" o f local govemments, and overburdening o f weak local
govemments.     Overall, it seems that while moving political authority over services and over resources to subnational
govemments can indeed improve service delivery, positive results are contingent on a large number of other factors, not least
ensuring the ability o f civil society to exercise discipline over govemment both through voice and exit as mediated by the
electoral arrangements and access to information; preventing monopoly control o f information being exercised by community
leaders; locating functional responsibilities for public goods at the right levels; limiting the retention o f hierarchical authority by
national govemments (even ifthis i s rarely used); and improving the flow o f information from governments to their constituents.
(See Azfar, Omar. 2002. "Conditions for Effective Decentralized Governance: A Synthesis o f Research Findings"
(http://www.inform.umd.edulIRIS/IRISIPEG/synthesisgaper.pdf).



                                                                 83

4.61.      Although Pakistan has made considerable progress in devolution, there i s a considerable distance
still to travel. Interms o f distribution o f state responsibilities among various levels of government, it i s
apparent that the actual assumption o f responsibilities at the higher levels has departed substantially from
the dejure assignment, with the federal government increasingly encroaching on provincial governments'
responsibilities. For example, although education i s a subnational responsibility in Palustan, the federal
government has assumed responsibility for formulating national policy, planning, developing curriculum,
managing centers o f excellence, and financing university education through the Higher Education
Commission(formerly through the University Grants Commission).

                    Table 4.1: FunctionalResponsibilitiesof Districts,Tehsils andUnionCouncils
       City District                    District             I    Tel~siYTaluka/Town~~                        Union
  *Public     transportatio:  *District planning                *Spatial plans                     *Statistical information for
  and mass transit            *Schools (boys & girls)           *Control over land-use, land-      socio-economic surveys
  *Passenger and freight      *Girls Schools                    subdivision, land development      *Consolidation of village and
  transit terminals           *Technical Education              and zoning by public and private   neighborhood development
  *Traffic planning,          *Colleges (other than             sectors for any purpose            needs
  engineering and parking     professional)                     *Prevention o f encroachments      *Union-wide development
  *Industrial estates and     *Special Education.               *Regulation o f sign-boards and    *Coordination for removing
  technological parks         *Literacy Campaigns &             advertisements                     deficiencies inthe delivery o f
  Cottage, small and          Continuing Education              *Municipal infrastructure and      TMA services
  medium sized enterprise     *Vocational Education             services                           *Registration o fbirths, deaths
  promotion                   *Public Health                    *Water supply and control          and marriages
  ohvestment promotion        *Basic & Rural Health             *Development o f water sources     *Libraries
  and protection              *Child & Woman Health             *Sewerage, treatment and           *Sports toumaments, fairs,
                              *Population Welfare               disposal                           shows and other cultural and
                              *District and Tehsil (HQ)         *Storm water drainage              recreational activities;
                              hospitals                         *Sanitation                        *Cattle fairs and cattle
                              *Agriculture (Extension)          *Solid waste collection and        markets
                              *Livestock                        disposal                           *Public open spaces, public
                              *Farm Water Management            *Roads and streets, other than     gardens and playgrounds
                              *Soil Conservation                roads falling under the            *Public sources o f drinking
                              *Soil Fertility                   jurisdiction of, and maintained    water, including wells, water
                              *Fisheries                        by, the District Government or     pumps, tanks, ponds and othei
                              *Forests (local)                  Government and streets             works for the supply of water

                              *Community Organization and       maintainedby the Union             *Street lighting, public ways

                              registration                      Administration or Village          and public places (through

                              *Labor safety regulation          Council                            mutual agreement with the

                              *Social Welfare                   *Traffic planning, engineering     TMA)

                              *Sports and Culture               and management (including          *Facilities for the
                                                                traffic signalling systems, signs  handicapped, destitute and
                              *Cooperatives                     onroads, street markings,          poor
                              *Enterprise and Investment        parking places, transport station  *Protection against stray
                              Promotion                         stops, stands and terminals)       animals and animal trespass
                              *Civil Defense                    *Street lighting                   Cattle pounds
                              *Information Technology           *Fire fighting                     *Regulation o f grazing areas
                              Development & Promotion
                              Environmental regulation          *Parks, playgrounds, open spacr    *Assistance indisasters and
                                                                and arboriculture                  natural calamities, and relief
                              *Land Revenue & Estate            *Slaughterhouses                   activities (including de-silting
                              *Excise and Taxation.             *Sports, cultural, recreational    o f canals)
                              *Spatial Planning and             events, fairs and shows
                              Development                       *Regulation o f markets and
                              *District Roads and Buildings     services
                              *Energy and Transport           I




4.62.      Although the devolution program provides the best opportunity inyears for improving the structure
o f governance and the effectiveness o f public services, the system needs considerable fine-tuning to achieve
these objectives. Some o fthe areasthat require early attention o f the govemment are the following:


52Inurban areas, these responsibilities are undertakenbycitydistricts, not the city towns.


                                                               84

     . Extending administrative and fiscal autonomy: Local governments do not yet have full control
       over the staff and finances o f devolved departments.                      For efficient delivery o f service, local
       government should have full autonomy over personnel decisions (hiring, posting, performance
       monitoring, and rewarding) o f facility-level and other related staff.                   Local governments at all
       levels will only start takmg their responsibilities for workforce management seriously as and when
       they consider themselves to be fully the employer o f their staff. Completing devolution means,
       first and foremost, transferring budgetary responsibility for salaries to the district governments.
     . Thus salary budgets should be transferredto the districts via account IV.

       Extendingfacility-level oversight: Useful steps have been taken in the development o f school
       management committees, but few significant employer functions have been transferred to the
       school level yet. Currently SMCs only have very limited funds, resulting from the transfer to the
       districts o f some nonsalary operating budgets, although in a few comparatively richer areas SMCs
       are raising their own funds as well. The devolution program presents an opportunity to further
       delegate decision making authority over critical facility-level matters to bodies such as SMCs,
       which can ensure better oversight over service facilities, perhaps reducing staff absenteeism. Staff
       absenteeism i s not a major issue for private sector schools and health facilities as immediate
       accountability i s ensured. For public facilities, community monitoring provides one avenue to
     . ensure facilities remain functional with their staffpresent.

       Improvingpolitical careers. It was hoped that the new system would introduce a new class o f
       politicians who had not previously stood for office and who would permeate the entire political
       system. Certainly, a large number o f women have for the first time participated in the political
       process. One issue, however, concerns the limited time that citizen activists will have to contribute
       to a large and growing range o f participative bodies, including Citizen community boards, citizen
       police liaison committees; village, union, district, and tehsil councils; participation inrural support
       programs and in district development committees for Khushal programs.                           Since the number o f
       new elected members at various levels i s large, allowances paid are very modest--a problem o f
       special concern to many low-income councilors (especially those elected on reserved seats). The
     . travel costs they must incur to attend a council session can be a much as a day's wages.

       Reducing institutional complexity: To be motivated to deliver public goods, politicians must be
       able to effectively claim credit from voters for providing such goods to them. The institutional
       complexity of the new system suggests that voters mighthave little knowledge about which tier o f
       local govemment i s responsible for which functions. The high proportion o f politicians that are
       indirectly elected complicates the linkage between voters and their representative^.^^ Enhancing
       awareness and transparency o f roles o f each level o f government i s crucial for effective delivery o f
     . public services.

       Improving capacity: The NRl3 and local civil society groups have been providing training support
       to local councilors. To date training has focused on constituency and coalition building, social
       mobilization, strategic negotiation, legislative oversight processes, and legal rights awareness for
       elected representatives.           Further training could usefully include skills building for new local
     . government politicians inpublicizing local achievements.

       Raising the eficiency offiscal management: Only a few provincial taxes, including the property
       tax, and some user charges have been devolved to the local government. This situation i s in
       distinct contrast to the well-accepted logic that revenue assignments should be closely matched to


53One third o f tehsil and district council members are indirectly elected by the 126,462 union councilors. (The ballot contained
126,462 seats, 21 seats for each of the 6,022 unions. Not all seats were filled, however.)


                                                                 85

       revenue needs.s4There are also many local tax administration issues to resolve, such as clarifying
       the mechanism of returning property tax revenue from the district to tehsils and education and
     . health sector user charges from the province to the district governments.

       Improving PFC formulae: The interim Provincial Finance Commission Awards need to be
       improved to avoid adverse incentives that they might create. For example, according to the Sindh
       PFC Award, districts will be given an additional allocation matching their baseline collection of
       certain items. This arrangement, however, may motivate districts to raise taxes regardless o f how
       well they are spent. Districts with wider tax bases and lower efficiency may gain more than those
       with a narrow base and higher efficiency. The PFCs could be encouraged (or required) to transfer
     . additionalresources throughmechanisms that convey federal and provincial policy priorities.

       Clarifiing mechanisms for conveying federal and provincial policy priorities: Provincial
       micromanagement i s enabled by the involvement o f provincial and federal governments in
       personnel management decisions in the districts (and to a lesser extent in the tehsils). It i s partly
       justified because there are few transparent avenues available to the federal government and to
       provinces to influence local government policy priorities. Inorder for the devolution initiative to
       be in complete sync with I-PRSP, mechanisms, national and provincial policies, local priorities,
       and financial resources must be better integrated. To achieve this, incentive mechanisms need to
       be instituted to help local governments prepare and implement their polices and plans in
       accordance with the national and provincial priorities, without compromising the autonomy o f
       local governments.        Increasing the tenure of senior district staj? Inter-district transfers o f
       provincial cadre staff are an enormous problem. As long as provincial and federal cadres retain
       positions in local governments, then some minimumtime must be placed on the tenure o f these
     . officers. Without this, the temptationto use transfers as a covert policy instrument i s too strong.

       Improvingfinancial management andfiscal reporting: One approach to reducing the incentive to
       use development funding for achieving political benefit i s to improve the quality o f public
       reporting on the effectiveness o f development expenditures. For more effective service delivery,
     . the financial management system shouldbe improved(see CFAA for detailedrecommendations).

       Reviewing alternativefunding mechanisms: Decentralization o f services i s usuallyjustified on the
       basis improving the delivery o f services that are valued by the community and for which they are
       willing to pay. Inthe case o f water and sanitation, for example, devolving capital expenditures
       from the provincial level to the district or tehsil level, accompanied by good intentions to tie them
       to capacity for operations and maintenance, i s unlikely to be sufficient to discipline investment.
       Without a relationship between the quality and quantity o f services delivered on the one hand and
       the cost o f those services on the other, infrastructure will suffer from underinvestment and
       worsening problems in operations and maintenance. Thus reliance on block grants to finance
       water i s clearly influencing the institutional options that city districts or tehsils are contemplating
     . for delivery o f services.

       Strengthening the Mushawarat Committees: The weak management links between the districts,
       tehsils, and union councils do seem to provide some challenges in collaboration.                         The only
       coordinating mechanism between districts and tehsils i s the Mushawarat Committee. Since each
       local body can make development decisions independently, without consulting or informing the
       others, there i s potential for considerable waste due to duplication o f efforts and development


54The argument is that govemments have few incentives to maximize their effort inraisingrevenues that they do not spend, and
  when they spend revenues that they do not raise, they have few incentives for spending them efficiently. See Shah, Anwar.
  "Fiscal FederalisminPakistan: Challenges and Opportunities". World Bank. 1996.


                                                            86

        schemes. The Mushawarat Committee can be strengthened to serve as a coordinating body for
        budgetary decisions.

4.63.    As the system beds down, it will be important to harness the potential that it offers for
strengthening user control over local policy and over the supply o f local services. This will require that
government take a distinctly pragmatic approach to the development o f citizen community boards,
amending their structure and their governance arrangements as practical experience i s gained.

4.64.    More strategically, devolution will require some adjustments in the new political environment.
An active role for the provinces must be found within the new arrangements. One area that the provinces
might well be encouraged to pursue is in actively choosing the form o f devolution that they encourage.
They could choose among several different paths. Early signs are that Sindh i s likely to seek to maintain
stronger guidance over policies at the local government level, whereas NWFP may feel more comfortable
in delegating authority. If this is the case, then the key is clarity and honesty about what approach
provinces are taking. Radical devolution mightnot be the approach chosen by each province.

4.65.    Most important, however, local governments need adequate fiscal and human resources to fulfill
their duties appropriately and efficiently. Considerable efforts need to go into developing the managerial
and technical capacity o f the local governments and providing them with adequate facilitating office
technology. Nevertheless, without a significant increase in fiscal resources o f local and provincial
governments, there i s little hope for any major improvement in public services through the devolved
delivery system.

Federal- Provincial Responsibilities

4.66.    With the recent devolution, the distribution o f responsibilities at lower tiers o f the government
has been corrected to a large extent. Yet there i s significant scope for devolving additional functions from
the federal to the provincial governments, which in many ways remains overly centralized. Over-
centralization o f functions i s also reflected inthe planning process, which still exhibits a strong top-down
format. A large portion o f decisions regarding development expenditure are taken by the federal
government. A clear example o f this i s inclusion of a large number o f schemes in the PSDP that are
regional innature.

4.67.    The simplest way to proceed on this front i s to have a serious review o f the actual responsibilities
in relation to spirit of devolution and, encourage the provincial governments to take back as many of
those functions that have been assumed by the Federal Government during the last three decades either
because o f financial constraints or lack o f technical and administrative capacity (actual or perceived) at
the provincial level. Devolution o f some o f the functions presently carried out by the Federal Government
would reduce the provincial government concern o f becoming redundant inthe wake o f large devolution
to the district governments, will lift the morale o f the provincial civil service and would enable the federal
government to focus more intently on its key strategic functions. In most o f the federal - provincial
concurrent list areas, the federal government should remain concerned with providing strategic direction,
setting national minimumstandards and providing financing to the provinces conditional on their compliance
withthese standards.

Inter- Government FinancialFlows

4.68.    The adequate funding o f the devolution initiative i s only a part o f the important issue o f inter-
government financial flows. W e have argued above for much greater devolution o f responsibility to the
provinces in order to underpin and strengthen the devolution to the district level. But even with the
present rather ad hoc distribution o f responsibilities, it i s clear that it i s the provincial/ district level


                                                        87

spending on education, health, law and order, provision o f justice, irrigation, rural and urban water
supply, androads that will account for the bulkofthe increase inrealpublic spending inthe near future.

4.69.    The share o f provincial/ district level public spending is likely to rise from around 23 percent o f
the total public spending in 2002 to over 35 percent in 2007. Interms o f GDP, provincial/ district level
spending should increase from 5.6 percent to over 8 percent over the period, if our assumptions about the
fiscal space and priorities materialize. This has major implications for allocation o f national financial
resources. At present, the sub-national governments are facing a severe fiscal squeeze. Whereas the
National Finance Commission (NFC) has been seriously considering a significant increase inthe share o f
provinces inthe federal divisible poll from the present 37.5%, greater flow o f funds would be required to
achieve the targets o f the PRSP. Clearly, mechanisms will need to be developed to allow a large portion
of the fiscal space created at the federal level to be passed down to the sub-national levels. This will
require two things: (1) new NFC and PFCs arrangements which assign much greater formula-based
transfers to the lower levels o f governments than at present and (2) puttinginplace a system o f financial
incentives and special transfers to the provincial and local governments so that national priorities can be
protected consistent with decentralization and devolution.

4.70.    Although institutional frameworks under the NFC and PFC exist to devise appropriate revenue
sharing formulae, presently there i s no such mechanism that can help the federaYprovincia1 governments
to induce the local governments to pursue the national/provincial policy objectives.           For example,
achievement o f PRSP targets, require a sharp increase in sub-national spending inhigh-priority areas like
education, health, water supply and sanitation, etc. It is, therefore, imperative to strengthen the
administrative and technical capacity and financial management systems at the sub-national levels and
develop mechanisms to channel sub-national expenditures towards meeting national objectives (e.g.
PRSP targets). One such mechanism could be institutionalizing the conditional grants, whereby the
federal government provides additional funds to the sub-national governments conditioned upon them
meeting some clearly definedtargets. For instance, the federal government can finance full or part cost o f
an adequately designed scholarship or any other incentive program to promote girls education in the
province or district.    Care must be exercised to make the grant criteria very transparent to avoid
misunderstandings and frictions among the governments. There i s a particular need for targeted national
programs to generate employment and income among the poor and for promoting specific national agenda
inhealthandeducationsectors.

ImprovingGovernanceandFightingCorruption

4.71.    Poor governance and rampant corruption have been responsible for significant wastage of, and
leakage from, public resources (see Box 4.8). Table 4.2 provides key indicators of governance for
selected South Asian countries. Although based on perceptions rather than real data, these indicators
show that Pakistan performs rather poorly in terms o f accountability, effectiveness o f government, and
corruption. All these weaknesses have a strong bearing on the effectiveness o f public expenditure. Areas
where governance has been a particular problem are personnel and resource management; these include
staff absenteeism, hiring o f staff on grounds other than merit, frequent transfers, and maldistribution o f
staff to the disadvantage o f rural areas. Closely related to weak governance i s the phenomenon o f ghost
facilities (especially schools and health centers) and staff, which siphoned a significant amount o f public
resources with no service being delivered.

4.72.     One aspect o f weak governance in Pakistan has been the widespread political interference in the
planning and budgeting process, which has eroded the quality and credibility o f the government's
expenditure decisions. InPahstan, the informal role o f politicians in the budgetingprocess went beyond
the ideal o f establishing the policy agenda for prioritizing the expenditures, malung final decisions on



                                                       88

broad expenditure allocations consistent with these priorities, and monitoring the use o f budgetary
resources and associated outcomes (such as via the Public Accounts Committees)."

4.73.     Senior government officers operate in a political environment, where absence o f job security
forces them to rely on their political patrons. The 1973 civil service reforms removed constitutional
protection given to civil servants against being prosecuted for unintended adverse consequences o f their
official decisions. Another significant change was the shift o f disciplinary proceedings from the Federal
Public Service Commission (FPSC) to the Establishment Division, which, although was later retrieved by
the Federal Shariat Court, had left a sense o f insecurityamong the civil service.

4.74.     The varied and unstructured responsibilities o f government officers; complex, out-of-date rules
and procedures; and unclear institutional structures create a system that on the one hand gives government
officers immense discretionary powers while on the other i s extremely opaque and within which it i s
difficult to determine responsibility and accountability. This situation creates an ideal environment for
corruption to flourish. In addition, the financial management and accountability system has broken down
(see para 4.80).

4.75.     Furthermore, according to anecdotal evidence, politicians frequently interfere in recruitment,
transfers, and posting o f government employees. More important, they exert pressure, and often with
success, to incorporate projects and schemes in the development program that do not merit inclusion on
grounds o f technical, locational, or economic reasoning. Lahore-Islamabad Motorway, Saindak Mineral
Project, and Chashma Nuclear Power Plant are some o f the many projects that were undertaken mainly on
the insistence o f politicians or for political reasons. Similarly, the Peoples' Works Program (PWP) and
Tameer-i-Watan (TJW) Programs inthe PSDP are standard examples o fpolitically motivated expenditure
programs. These programs were included in the PSDP to provide the elected representatives with
development resources to undertake projects o f their liking. Although justified on the grounds that these
programs offer an ideal mode o f bringing local preferences into the planning and budgeting process, the
schemes undertaken under these programs circumvented the regular screening processes o f the
government and thwarted the policy frame~ork.'~




''For  a variety of reasons, the Public Accounts Committees (PACs) failed to meet regularly, leading to a large backlog of
undiscussedaudit reports. As a result, two- to three-year delays between when expenditures are incurred and when the PACs
take up the relatedaudit reportsfor discussion are quite common.
56For example, the rural water supply schemes initiated by the govemment departmentsask the benefiting community to sign a
memorandumof understandingwith the department indicatingthat it will assume responsibilityfor operatingandmaintainingthe
scheme. Rural water supply schemes started under PWP and TIW, however, did not adhere to this sectoral policy, affectingnot
only the sustainabilityo fthese schemes, but alsomakingapplication ofpolicy onregularschemes thatmuchmore difficult.



                                                             89

                                          Box 4.7: CorruptioninPakistan

According to a Transparency International perception survey, Pakistan falls among the bottom few countries in
prevalence o f corruption. With the establishment o f national and regional accountability institutions, however, and
the military govemment's stated objective of reducing corruption from the society, the situation has improved
significantly over the past three years. Presently Pakistan scores about the same as India and significantly better than
Bangladesh on the corruption perception index. Nevertheless, there i s significant room for further improvement. A
recent survey by Transparency International gives a stark review o f corruption in Pakistan. Based on perceptions and
experiences o f individuals, the survey attempts to explore the level o f corruption in seven main departments o f the
government. The results o fthe survey are reproducedbelow:

Education: About 70 percent o f respondents who interacted with educational institutions cited the existence of
irregular methods o f gaining admission. Teachers and members o f the management committee were cited as the most
involved. The main causes o f corruption identified were a lack o f accountability and low salaries.

Health: Sixty-five percent o f all patients visiting a hospital reported irregular admissions, and 96 percent o f those
who were admitted said they were victims o f corruption. Hospital staff were identified as the key facilitators of
corruption by 65 percent o f the users, and direct extortion was reported in 60 percent of the total cases o f corruption.
Lack o f accountability and monopoly power were quoted as key contributing factors.

Electricity: A very high percentage (65 percent) o f users with access to electricity reported irregular processes in
acquiring it; a much higher percentage reported corruption in regular interaction with the department (96 percent).
Meter readers and billingemployees were identified as the key facilitators; extortion was reported by 72 percent of
the victims. A lack o faccountability andlow salaries ofemployees were identified as major contributoryfactors.

Land administration: Only 8 percent o f respondents reported interactions with the land administration department.
All o f those who used the services o f the department, however, reported incidences o f corruption. Surveyors and
revenue officers were identified as the major facilitators. Bribes were directly demanded in64 percent o f cases, and a
lack o f accountability, discretionary powers, and low salaries were cited as the key causes o f corruption.

Tax administration: Eight percent o f respondents interacted with the tax department duringthe past year. Corruption
was perceived to be rampant, with 99 percent o f users reporting corruption. A third o f all incidences o f bribes were
paid to reduce a customer's assessedtax. Tax officers and employees o f the department were identified as the major
facilitators, and in most cases (61 percent) bribes were extorted. A lack o f accountability was identified as the major
reason for the prevalence o f corruption inthe tax department.

Police: Nearly a third (32 percent) o f all respondents reported using the services o f the police duringthe past year.
All o f those who interactedwith the police reported encountering corruption; one intwo victims identified the police
officer as the key perpetrator. Extortion was reported by 74 percent o f the respondents. A lack o f accountability and
low salaries were quoted as the major contributory factors.
Judiciary: Only 6 percent o f respondents reportedusingthe services o f the judiciary duringthe past year; almost half
of the users interacted in their capacity as a complainant. Ninety-six percent o f users reported encountering
corruption; 44 percent o f cases involved bribingcourt officials. Court officials (41 percent) andjudges (24 percent)
were identified as key facilitators; in most cases (61 percent), bribes were demanded directly. One-third o f the
respondents said that a lack o f accountability was the main factor contributing to corruption inthe judiciary.




                                                           90

                              Table 4.2: Key GovernanceIndicators in South Asia
                           Polity       Press              Voice &            Law &       Govemment        Graft
                           Score      Freedom           Accountability         Order      Effectiveness  Corruption

Bangladesh                6.00         60.00               -0.20               2.00           -0.54        -0.64
India                     9.00         42.00               0.66                4.00           -0.17        -0.39
Nepal                     6.00         57.00               -0.06               NIA            -1.04        -0.3 1
Pakistan                  -6.00        57.00               -1.43               3.00           -0.48        -0.79
Sri Lanka                 5.00         74.00               -0.23               3.00           -0.44        0.00
Source: Transparency International, Corruption in South Asia, December 2002.


4.76.    Recently, the newly elected government has announced it will provide development funds (Rs. 10
million for each member o f the national Parliament). Not only are there concerns that this program will
follow the same route o f PWP and TIW and lead to wasting o f public resources, but in view o f the
devolution program, the argument that this programbrings local preferences into the budgeting process i s
muchweaker and may even go against the devolution objectives o f the government. Moreover, unlike the
past when the deputy commissioner (generally a federal government employee) coordinated all
development activities, including PWP and TIW schemes, in the district, the present district coordinating
officer (DCO) i s a local government employee under the district Nazim-a politician. There are therefore
concerns that coordination of the M " P A                  schemes with other local and provincial development
activities may not materialize, leadingto duplication and further wastage o f development funds.

4.77.    Governance i s a multidimensional issue, and improvements in all dimensions are required to
improve the effectiveness of government. Given the correlation between increase in corruption and the
erosion o f real income o f government employees, an increase inthe salary o f civil servants would play an
important part in reducing corruption from within the government. In the short run, the govemment
should ensure that rules and regulations governing financial transactions are strictly adhered to.
Deviations from these rules should be kept to a minimum, with the need for doing so adequately spelled
out. The decision to allow such a deviation should rest with the responsible agency and not with the
politicians (for instance, the Planning Commission in terms o f including development projects in the
development program), and the decision should be made only after thorough consultation with all
concerned parties. The National Accountability Bureau (NAB) needs to develop a medium-term
anticorruption strategy, which should not only significantly improve present methods o f combating
corruption, but also formulate a more effective and systemic approach toward monitoring and prevention
o f corrupt practices. For any anticorruption strategy to be effective, it needs to promote the following
features in the government's operations and decisions: transparency, accountability, meritocracy,
deregulation, standardization, automation, and community and public involvement.

Strengthening FinancialAccountability

4.78.        Although, Palustan has a reasonably sound accountability system on paper, in fact weak
accountability i s the factor underlying most, ifnot all, the problems that impact the effectiveness o fpublic
expenditure. The procedures governing financial accountability are neither overly cumbersome (given
the circumstances of Palustan) nor obviously flawed.              Nevertheless, accountability has been a serious
problem, for the following reasons:

         a) Parliamentary and executive control o f the public purse has been ineffective; audit reports on
            the annual accounts are heavily qualified;          and federal, provincial, and district accounts do not
            currently meet the expectations o f the Constitution and associated laws and regulations.




                                                              91

              Substantially strengthening the effectiveness o f the oversight function provided by the public
              accounts and other parliamentary committees can improve this control.
           b) The accountability o fthe executive at present is not sufficiently supported by informationthat
              would enable it to become focused on results and outcomes. This improvement infocus can be
              accomplished by progressively introducing a performance-based management system with
              incentives that reward outputs and outcomes and hold the secretaries o f govemment
              departments to account for meeting departmental objectives and performance standards. The
              responsibilities need to be enforced by strong performance reporting and review functions and
              intemal control responsibilities.

           c) The separation o f audit from accounting functions i s still not complete. The Auditor General o f
              Palustan (AGP) continues to exercise control over staff o f Pakistan Audit and Accounts
              Service (PAAS), which include the accounting staff. This implies that the Controller General
              of Accounts (CGA) does not have clear administrative authority over his own staff. This
              problem could be effectively resolvedby separating PAAS into an auditing cadre, controlled
              by AGP, and an accounting cadre, controlledby CGA.

           d) The devolution process has brought to light the institutional wealmesses in the accounting
              system. The initiative for a more devolved system o f governance for the central, provincial,
              and local governments and their financial reporting arrangements needs substantial further
              strengthening in financial accounting, reporting, and auditing and in arrangements for
              improvingpublic scrutiny.

           e) The public audit function does not currently meet the standards required by the Constitution
              and supporting authorities. The audit reports are too narrowly focused and do not contain any
              information on the efficiency and effectiveness o f program or service delivery. The focus i s
              mainly on pointing out irregularities in individual transactions without sufficient attention to
              significance, and systemic weaknesses, o f financial management system. The legislative
              oversight o f accountability system too i s weak. The PAC meets too infrequently and lacks the
              expertise to undertake an effective review o f audit reports. Public scrutiny of the govemment's
              activities and financial transactions i s seriously impaired because o f lack o f timeliness and
              comprehensiveness o f legislativereviews o f audit reports and because o f poor follow-up to the
              PAC's recommendations. The auditing function can be strengthened by focusing it in the
              short term more directly on attest audit and improving intemal controls, ensuring its
              independence by malung the Auditor General more independent o f the executive; improving
              the PAC with institutional support; and making the govemment responsible for following up
              on the PAC's recommendations.

           f) Although a penalty system exists for misdemeanors in public service (such as unauthorized
              expenditures and misuse or theft o f resources). A range o f sanctions are applicable, including
              dismissal. The procedures for exercising such sanctions, however, are inordinately
              cumbersome and lengthy, with the result that they are ineffe~tive.~' There i s a need to
              simplifythe penalty procedures.

           g) While dismissal o f staff through regular procedures appears to be virtually unachievable,
              anecdotal evidence suggests that informal intervention from politicians, at least in some cases,


57 Incontrast to the lack o f timeliness o f such sanctions, staff can be transferred to different departments and different areas o f the
country at relatively short notice. Transfers are regarded informally within the public service as an effective power of sanction,
although this method is not used officially for this purpose. Instead, it increases the scope for breakdowns in accountability by
virtue of the power that the authority to transfer bestows.


                                                                    92

            can result in swift dismissal (and also in appointments that circumvent procedures), with
            adverse implications for accountability and morale.
         h) Procedures for recruitment on merit at the senior level appear to be sound, but promotions
            depend on Annual Confidential Reports (ACRs), which apparently have failed to establish a
            sound performance system. A very large proportion o f staff i s given an "outstanding" rating by
            managers, irrespective o f their performance and in contravention to the ACR guideline.
            Moreover, the completion o f ACRs i s not without bias.

         i)Therecentreversalofthepreviousgovernment's decisiontorecruitallfederalgovernment
            staff in BPS 11-16 through the FPSC i s a step in the wrong direction. The FPSC's purview
            over this recruitment should be restored, or alternatively a foolproof system should be
            developed by which this recruitment could be made by the government departments without
            compromising merit.

4.79.    The government has recently embarked on a comprehensive reform program to restructure its
financial management system.58Although significant steps have been taken in this regard (see Box 4.1),
the government must carry the program forward and involve the subnational governments, public
corporations, and autonomous bodies more intently in the reform process. The continuation, expansion,
and deepening o f the reform program can ensure substantial improvement in the government's financial
management system within the next three to five years. The CFAA-Phase-I1 has outlined (among other
things) the followingkey actions for improving the financial management system.

4.80.    The immediate actions include (1) enhancing the transparency o f the budget by disclosing fixed-
cost and contingent liabilities and the transparency o f the accountability system by publishingannual and
quarterly financial reports o f all levels o f government; (2) developing provincial budget implementation
systems along the lines of the federal government's Revised System of Financial Management and
Control and making the federal andprovincial systems fully operational; (3) establishing a strong internal
control mechanismfor each level o f government; (4) buildingthe capacity o f the auditing and accounting
establishments and o f the PAC; and (5) renewing the government's commitment to implementation o f
PIFRA.

4.81.    The medium-term measures include: (1) developing and operationalizing the MTBF; (2)
undertakmg civil service reforms; (3) implementing a phased transition from cash to accrual accounting;
(4) giving provinces control over their own accounting system by redefining and clarifying the role o f
CGA and AGs; (5) ensuring a transition to international auditing standards; and (6) malung public
accountability more viable through formal PAC reviews and training the public and the media on the
power and utilityo f financial information.




   These reforms are based on issues and recommendations that appeared in the IMF's Accountable Fiscal Management
Framework and the World Bank-supported Projectfor Improvement in Fiscal Reporting and Auditing (PIFRA; and the Country
FinancialAccountability Assessment (CFAA )-Phase I and II.




                                                         93

                     CHAPTER5: STRENGTHENINGTHE CIVILSERVICE

Key Issues

5.1      As discussed in the previous chapters, Pakistan has a unique opportunity to expand and
restructure its public sector spending in support o f its economic and social goals. But while public
expenditure on infrastructure and social development must increase significantly for reviving growth and
reducing poverty, this effort could be wasted if public funds are not utilized effectively. Inthe previous
two chapters, we discussed the principal issues, which normally influence the effectiveness o f public
spending, strategic approaches to public interventions, the role o f the various levels o f government,
sectoral allocations, budgetary, planning and monitoring procedures and fiduciary controls. But all these
aspects o f design and implementation o f public service delivery require functioning institutions including
a clear framework o f laws and regulations. A competent, honest and independent civil service provides
the backbone o fthese institutions.

5.2      In Palustan the strength of public institutions has been greatly eroded over time. Many inter-
related factors have contributed to this decline including increasing centralization and non-transparent
decision malung, lack of accountability, low or declining compensation, and the dilution o f the role o f
meritinselection andpromotion of civil servants withundue political interventions innormal functioning
o f government departments and agencies. This has contributed to declining technical and professional
competence o f officials, and a culture o frisk-averse behavior.

5.3      Some steps have been taken to reform the civil service aimed at improving the performance o f
public servants and reducing corruption. In the interest o f merit based recruitment, a more autonomous
Federal Public Service Commission (FPSC) was given responsibility for recruitment for all middle level
and highlevel professionals. Promotion panels for higher grades are now chaired by the Chairman FPSC
and to ensure quality of high grade staff, acceptance to advanced training programs are now subject to
entry examinations. The first phase o f pay and pension reforms was implemented in January 2002. All
federal civil servants including military were given new pay scales which restored 75 percent o f the lost
purchasing power after 1994 and also revised pension benefits (lower commutation rates and more
actuarially fair commutation factors). The pay reform also reduced the share o f monetary allowances from
40 to 27 percent and de-compressed the pay scales. The ratio o f the highest to the lowest increased from
1:9.2 to 1:9.9 including all monetary allowances.

5.4      Palustan shares the symptoms o f poor performance and low accountability stemming from an
unresponsive bureaucracy with many other low-income countries. Indeed, the problems o f compressed
pay scales offering few incentives for slulled senior managers to join the public sector, the excess o f
unslulled lower level employees relative to senior staff, and the extensive patronage and politically
motivated appointments and transfers are found in many settings.             However, some o f Pakistan's
administrative reform challenges are distinct - as are the opportunities that are available for addressing
them.

5.5      The structural rigidities in the federal and provincial administrations stem from strongly
vertically-oriented occupational groups. These groups foster the distinctively deep-seated concern with
status. Inter-cadre rivalries significantly determine the shape o f federal divisions and attached
departments, as structures are created to meet the needs o f the various groups and cadres for promotion
opportunities -rather than on the basis o fpolicy priorities or operational needs.




                                                      94

5.6      However, government in Pakistan also has some significant advantages. Above all, the public
sector is not bloated in the absolute sense o f too many employees representing an insupportable fiscal
burden. The wage bill i s not highinabsolute terms, and has been reduced over the last decade as real pay
has declined. This i s not to say that it has not crowded out other non-salary recurrent expenditure                          - but
this owes more to the reductioninthese expenditures than to an increase inemployment costs. Also, this
i s not to offer false comfort public employment i s certainly excessive relative to what the public sector
                                   -
delivers and it i s undoubtedly bottom-heavy, with a disproportionate number o f employees at the lower
levels. The government also finds itself with three `low-hanging fruit' that can be picked at relatively
modest political costs.          Specifically, government has: some consensus between the Establishment
Division and the NREI on an approach to restructuring; the fiscal space to afford pay reforms; and a well
developed, but incomplete devolution agenda.

5.7      This chapter sets out the need for administrative reform, but argues against an over-ambitious or
over-elaborate set of proposals. History suggests that such proposals will generate many opportunities for
lengthy reports, but will lead to little sustained progress. Instead, it argues for three steps. First, it
suggests that there i s an opportunity to move ahead rapidly on the creation o f a National Executive
Service that would greatly increase incentives both for improved performance and upgrading o f s h l l s
through learning and education. The two critical elements o f this reform, on which there i s seemingly a
substantial consensus, are the much greater emphasis on merit and performance for selection to top
grades, including a highbar beyond grade 18, and opportunities for higher education and learning for civil
servants especially early in their careers. Second, it suggests sharp pay increases (and monetization o f
many o f the large in kindbenefitss9) for higher levels o f civil servants so that the serious decompression
of pay that has been going on for decades can be reversed. Finally, it notes that completing administrative
devolution, by transferring budgetary responsibility for salaries to the district governments and by
progressively passing appointment, promotion and transfer (APT) authorities to the districts, will address
a necessary, if not sufficient, condition for local governments to value professionalism more than
patronage.

5.8       These or similar steps are central to the achievement o f both the narrower objective o f more
efficient use o f public resources and the broader objective o f "reviving growth and reducing poverty". In
a very basic sense, civil service strengthening i s on the critical path o f the reformprocess.

Structure, notsize,is the problem

5.9        Government in Palustan starts from a significant advantage                   -the public sector i s not bloated in
the sense o f too many employees or excessive fiscal weight.                       Total public employment i s about 3.2
million (Figure 1) and i s around 2.3% o f the population6'. Historically, provincial employment in the
provinces has been driven more by political and patronage concerns rather than identified social and


59 Inkindbenefits, subsidized housing, free cars, telephones, etc are largeparto fcompensation at the highest levels. Such inkind
benefits are wasteful because they generally lead to excessive use o f facilities provided, free telephones, cars, gasoline, utilities,
etc. An equivalent cash compensation will both increase the discretion o f civil servants about the pattem o f their consumption
and reduce wasteful spending.
60Total Public Employment includes both state-owned enterprise (SOE) employees and General Govemment. General
Govemment refers to employment in "all government departments offices, organizations and other bodies which are agencies or
instruments of the central or local authorities whether accounted for or financed in, ordinary or extraordinary budgets or extra-
budgetary funds. They are not solely engaged in administration but also in defense and public order, in the promotion of
economic growth and in the provision o f education, health, cultural and social services." (International Standard o f Industrial
Classification o f All Economic Activities (ISIC), Series M No. 4, Rev 3- 1990.) By tradition, General Govemment contains six
mutually exclusive categories: Armed Forces; Civilian Central Govemment; Subnational Government; Health employees;
Education employees; and Police. It excludes State-Owned Enterprise (SOE) employees: employees o f enterprises that are
majority owned by govemment.



                                                                95

development priorities. Sanctioned employment inprovincial governments grew dramatically inthe early
to mid 1990s but has since leveled off or reduced slightly. Federal employment i s only about 20 percent
o f general government employment (excluding military) reflecting the federal character o f the
government.

                     Figure5.1: The Main Componentsof Public SectorEmploymentinPakistan

                                                                                                                               Powlation (0001        134,790
                                                                                                                                                       37.2%
                                                            TOTAL PUBLIC EMPLOYMENT                               Total publicemp. as % of pop.         2.3%
                                                                        3,164,900                                Generalgovt emp. as % of pop.          2.1%




                                                382,800                                          2,782,100


                                                                                                                                  I
               Armed Forces (ii)                                                                                         Total subnational
                                                                     government (iii)                                     government (iv)
                                                                                                                             1,765,600



                                                                      Education (v)                                          Education

                                                                          10,700


                                                                        Health (iv)                                            Health




                                                   I                      Police                                               Police
                                                               1         106,700

                                                                                                                            Subnational
                                                                 Governmentexcluding                                  Government excluding
                                                                   education, health.                                   education, health,
                                                                        B police                                              B police


         ources: (i) Government of Pakistan,ManagementServicesWing. EstablishmentDivision,"AnnualStatisticalBulletinof Employeesof AutonomousISerni-
        AutonomousCorporationslBodiesunderthe Federal Government. 2001-2002; (ii) InternationalInstitutefor StrategicStudies, The Military Balance 1996'99; data
        for 1997; (iii) Governmentof Pakistan,ManagementServicesWing. EstablishmentOiyision. "AnnualStatisticalBulletinof FederalGovernmentEmployees,2001-
        2002"; and "AnnualStatisticalBulletinof Employeesof AutonomouslSemi-Autonomous CorparationsIBodiesunderthe FederalGovemment,2001-2002"; (iv) Data
        for2000 providedby the Ministryof Finance. Government of Pakistan;(v) Governmentof Pakistan,PlanningCommissionReport,April 2001; (IV)Data does not
        include70,000 communityhealthworkers hiredon one-yearrenewablecontracts underthe "NationalProgramfor FamilyPlanning& PrimaryHealthcare."




5.10        Excluding state-owned enterprises, general government employment i s some 2.1 percent of
population, which               -  bearing in mind that international comparisons are only indicative because of
practical and conceptual problems relating to the accuracy and consistency o f all public employment data
- compares to           an average of some 2.3 percent for other low-income countries inthe region6'

5.11        The wage bill, at Rs 160 billion, is about 4.3 percent of Pakistan's GDP. This is similar to the
level o f this ratio in India, China, and Bangladesh although higher than Korea and possibly other similar


61Source: (de Tommaso, Giulio andAmitabhaMukherjee.2002. "World BankAdministrative& Civil Service ReformWebsite:
Cross-NationalData on GovemmentEmployment& Wages"; World Bank.
(http://wwwl .worldbank.orglpublicsector/civilservice/cross.htm.Access Date: March24,2003).
Data on public sector employment are notoriouslyslippery and other sources make the regional average rather higher. Schiavo-
Campo et a1indicatethat general civilian govemment employment(Le. public sector employment excludingthe military as well
as state owned enterprises) has accounted for about 2.5% of the population in Asia and the Pacific in recent years (Schiavo-
Campo, Salvatore and P.S.A. Sundaram (eds). 2001. To Serve and Preserve: Improving Public Administration in a Competitive
World. Manila: Asian DevelopmentBank.)




                                                                                 96

emergent economies (Table 5.1). It i s not highinany absolute sense and it has gone down inthe 1990s as
real pay declined (notwithstanding recent adjustment). However, the share o f establishment costs intotal
recurrent spending increased during the last decade as overall spending was squeezed. As a result, non-
salary expenditure in the social sectors such as education and health, and O&M spending on irrigation,
roads etc. declined.

5.12     The average public sector wage as a multiple o f per capita GDP at 2.9 for Pakistan i s about the
middle o f the range between the high level in Korea, Malaysia, and Bangladesh and low level in China,
Indonesia, Sri Lanka and the Philippines.




          (in percent of total population)                                      Ei
                                                        ",
                                                         F9  g5     E&'       ;a o                                                    E:
                                                                                i;'     2
Wage bill as % ofGDP                                      3.7  3.8      4.2      3.0     2.2      5.9     7.7         4.3     5.5      5.1
Average wages as multipleofper capitaGDP                  4.1  1.3     N.A.      1.6     4.6     2.0      3.4       2.9       2.3      1.8




Bottom heavy

5.13     Public sector employment in Palustan may not be large in this comparative sense but it i s
certainly excessive in many areas inrelation to what it delivers and certainly it became large in the first
halfofthe 1990sinrelation to the fiscal resources (see Table 5.2).

                                  Table 5.2: SanctionedPostsby Province(selected years)
                                           1988189               1993194                     1996197                      2000d
     Punjab                               722,916               818,647                      880,781                  888,796
     Sindh                                285,042               424,974                     445,577                   457,494
     NWFP                                  177,106b l           227,679                     254,587          1              n.a.
     Balochistan                            98,942               122,921                     127,069         1        128,132

            ai Figures are actual posts as ofJanuary 2001.
            b/ NWFP data for 1988/89 has been estimated

                                                                          Figure5.2: Compositionof public employment
5.14     Also, public employment i s heavily
skewed towards lower grades (Figure 5.2),
which often typify low productivity levels:
about 95 % o f the public employment i s in
grades 1 to 16. For just the Federal level-this
share i s 97 percent.

Rigid                                                                                              I 1
5.15     kgidities arise because decades o f
patronage-based employment and inattention to
skill mix in the civil service has undermined the                  .grades 17-22  11,491  7.378    61,334     29,807    3,894   24,125     ~
work ethic throughout the civil service and has                    agrades 12-16  13,050  8.309    124,537    38.760    6,371   68,912
                                                                   Ogrades 1-11  203,138  112,445  702,925   376,813    53,197  741,710

                                                                 Source: WB staff files

                                                               91

created an entitlement culture, in which status considerations and inter-cadre rivalries far outweigh any
concerns about service to the public or to government. Various reforms have attempted to counter what
are considered to be the self-serving nature o f the bureaucracy. Most radically, the 1973 reforms sought
to simplify the grading system, encouraging mobility between different parts o f the service and
encouraging lateral entry for talented individuals from the private sector. The reforms seem to have had
surprisinglylittle impact on the vertically-oriented occupational groups, and despite these reform efforts,
occupational groups remain vertically separated, distinct career streams.

5.16      The structure o f federal divisions and attached departments i s unnecessarily restricted by the need
to ensure that the various occupational groups that have a claim to positions within that division or
department are satisfied. Positions are created to satisfy the needs o f the various groups and cadres for
promotion opportunities -rather than on the basis o f policy priorities or operational needs. Insome cases
set proportions o f various groups have also to be maintained.

5.17      Recently lateral induction on fixed term contracts have been introduced (MP grades systems).
However, it i s unclear how many organizations are talung advantage o f this flexibility. It appears that the
proportion o f these positions inthe senior management remains low. Generally, the appointments made
at senior levels reflect the need to accommodate generalists who tend to be promoted in large numbers
because their cadres are larger and thus promotion i s faster. At times the career paths define different
years o f experience for different cadres malung promotions slower or faster inthem. Also for some o f the
technical cadres the promotion path narrows down with seniority and ends before reaching the highest
levels. For this reason, senior positions tend to exclude staff who have technical backgrounds in specific
disciplines.

MismatchBetweenSkills andRewards

5.18      Compensation has become increasingly insufficient to attract skilled entrants at the higher levels
but may be too generous at lower levels. The average pay masks a serious problem with pay
compression.62 As noted earlier, and as shown in Table 5.3, the present ratio o f the highest to the lowest
paid (BPS22:BPSl) i s about 1:lO.              In OECD settings, this ratio might be reasonable because o f the
structure o f the labor market. At higher levels, public sector pay inboth developing and OECD countries
may be compared with slulled private sector positions. However, at lower levels, while in industrial
countries the usual comparators for public sector pay are clerical jobs in private companies, this
comparison may not be relevant indeveloping counties where the true employment alternatives o f many
public sector workers are inthe informal sector o fthe economy.




  "Compression Ratio" is used here to refer to the ratio o f the highest salary to the lowest on the central government's main
salary scale. This definition differs from that adopted by the OECD, which measures wage compression in OECD countries as
the mean o f ninth decile salaries divided by the mean o f first decile salaries. The OECD approach ensures that a handful of
salaries will not dramatically skew the compression ratio. However, the data quality needed to calculate the OECD compression
ratio makes it impractical for extensive cross-country work.



                                                               98

                                   Table 5.3: Impact of the January 2!002 Pay Reforms

                            Before pay reform (1994 2001)
                                                      -(monthijpay,    Rs)

                   Base Day a/I  Estimated  I   Total     IAllowances


~~

    ade 2




  Source: WE3 staff files
  a! These ratesshowthe pay of staffwho are 2 incrementsbelowthe scale maximum

  5.19      There i s substantial evidence that real compensation for the senior levels has declined over time
 and the gap between public and private salaries and emoluments has steadily increased. For the highest
  grade (BPS22), the nominal salary and monetary allowances have increased from Rs. 3,000 per month in
 the 1950 to Rs.38,000 at present while consumer prices have arisen thirty fold over the period.
 Undoubtedly, the share o f in lund benefits at higher levels has increased but certainly not enough to
  compensate for the sharp drop inreal pay. Real total compensation for the highest civil servants probably
  decreased over the last half century by at least one-third. 631nthe meanwhile, real incomes in the private
  sector have at least doubled reflecting the general increase inper capita incomes. Even ifone assumes that
 the gap between the top management positions in the private and the public sector was small initially, it
 would appear to have grown several fold. The increasing use of MP grades, originally meant for top
 managerial positions in the public corporations and approximating more closely to private sector pay, to
  attract managerial and technical skills to govemment also indicates the problems with the present levels
  of compensation. Finally, the fact that there has been a historic decline o f the pay compression (excluding
 in kind allowances), estimated at 1:4664in 1962, also seems to confirm the hypothesis that pay at the
  lowest levels has increased significantly over time while it has declined at the highest levels.65

  5.20      This pattern o f compensation is undoubtedly contributing to two significant problems. First,
 relatively generous remuneration at lower levels increases the pressure for patronage-based employment
 o f unskilledstaff. Second, below market rates at senior levels are unlikelyto be attracting the quality o f
  entrant that i s needed at the highest levels o f govemment.



 63 See Hasan, 1998 (p.359).
 64 Table 7.8, "GovemmentThat Works: Reformingthe Public Sector inBangladesh" 1996. World Bank.Dhaka.
    Indeed if one assumes that there has been no real pay decline at the highest levels, the lowest level pay must have increased
    very sharplyto explain.


                                                               99

                                         Box 5.1: MonetaryAllowancesand inkindBenefits




                                                                           Beforepay PostDec.
                                   Rs (million)                              reform         2001
                                   Annual Base Pay                              11,728        18,774
                                   Annual Allowances                               7,765       6,85 1
                                   Total                                        19,493        25,625


                                   Notes: This refers to federal ministries and attached departments-
                                   it does notincludeautonomousbodies

     Table 5.4 shows that January 2002 payreforms provided consolidation o f monetary allowances within base pay. The
     table shows the impact o f consolidation on the federal civil service wage bill. The impact on the pr
     will have beenbroadly similar.

     Even after the January 2002 pay reforms, inkind benefits for BPS-22 (value of accommodation, transport, etc.) were
     estimatedto be much higher than the basic pay. Further, monetization o fthese allowances and in-kindbenefits i s highly
     desirable as has been done inthe State bank of Pakistan and i s under consideration inthe Central Board o fRevenue.

PersistentPatronage

5.21      Politically-motivated patronage appointments to the civil service and transfer decisions have had
a damaging impact on performance. For this reason, the decision o f the Federal Cabinet inearly February
to recruit grades 11-16 without involving the Federal Public Service Commission sends a disturbing
signal. Delegationo f lower level recruitment to line departments, per se, i s not problematic. Most Public
Service Commissions delegate recruitment decision for junior and middle-level to line departments.
However, generally they do this while maintaining tight regulatory supervision over the recruitment
process.66Thus while they are not themselves recruiting, they maintain full responsibility for the quality
of the recruitment process.

5.22      The recent decision in Palustan seems to remove responsibility for the process as well as the
function from the Commission. Since the original decision was made with the intention of providing
some safeguards against identified problems of patronage in a range o f more sensitive federal bodies, it
begs the question as to why this will not simplyrecur. However, in addition this will send an unfortunate
signal to the provincial Commissions. InNWFP, for example, PSC oversight has been extended down to
grade 11 in some corruption-prone departments (mirroring the equivalent reform at the federal level). It
i s improbable that such arrangements will remain in force given the federal changes.

5.23      In addition to appointments, political pressures are also behind the continuing high level of
transfers o f senior staff. While data on the rate o f transfers o f senior staff are not available at the federal
level, the problem i s clearly chronic at the provincial levels. InSindh, the average tenure o f Secretaries i s
some 15 months. InNWFP, it i s 7.5 months.


66 See Polidano, Charles and Nick Manning. 1996. Redrawing the Lines: Service Commissions and the Delegation o f Personnel
   Management. Managing the Public Service     - Strategies for Improvement Series. London: Commonwealth Secretariat.




                                                                  100

5.24     Similar problems pertain at district levels. The Provinceretains the ability to transfer senior staff
in any of the provincial occupational groups (i.e. above grade 16 and thus not transferred inDecember
2002 to the districts).    This i s a highly politicized arrangement to influence transfers throughout the
district. By the threat o f an unpleasant transfer, or the promise o f an attractive one, the senior staff
member can be put under pressure to accede to a transfer o fthejunior staff.

A Pragmatic Approach

5.25     The track record o f incomplete and ultimately ineffective civil service reforms in Palustan
suggests that there i s little advantage in devising yet another large and elaborate strategy. This report
suggests instead that some momentum towards reform i s created by taking advantage o f three fortuitous
circumstances. The government finds itself in the unusual position o f having: some consensus between
the Establishment Division and the NRB on some restructuring; the fiscal space to afford pay reforms;
and a well developed but incomplete devolution agenda.

5.26     By seizing the low-hanging fruit that these provide, government can set in motion some
movement towards a longer-term reform agenda. Other steps can be taken that will achieve short-term
efficiency gains and send significant signals at no political cost. For example, on recruitment, the FPSC
should be given the unambiguous responsibility for oversight o f the recruitment process at all grades,
even if the decision stands that it should not itself undertake direct recruitment for positions below
BPS16.

5.27     Similarly, since present compensation levels are inadequate at senior levels, use o f fixed term
consultants with maximum term o f say three to four years at near market salaries should be greatly
encouragedboth for professional and managerial positions. Liberal use o f consultants i s especially needed
to strengthen the federal ministry o f finance (including the debt office and the economic adviser's wing),
the federal planning commission and the ministry o f commerce as well as planning and finance
departments o f the provincial governments

Seize the emerging consensus

5.28     The NRBhas proposed the creation of five new broadly-based services:

                 ...
                 9
                 . The Federal Civil Service
                    The National Executive Service
                   The Provincial Civil Service, including the ProvincialExecutive Service
                    The District Civil Service
                    The Tehsil Municipal service

5.29     Within the NRB proposal, both the Federal and Provincial Civil Services will continue to be
divided into occupational groups below grade 18.           The National Executive Service (NES) would,
however, be an All Palustan Service comprising a pool o f officers selected on merit from amongst
officers o f the Federal, Provincial and District Services for critical policy and management positions
beginning at Grade 19. A maximum o f 20 percent o f the entrants would be recruited laterally from the
private, NGO and academic sectors through a competitive process. The NES posts would total between
1,000 and 1,500 and would encompass all grade 19 to 22 positions in the federal secretariat, heads o f
important departments in the federal government, and senior positions in provincial and district
governments (e.g. Provincial Chief Secretaries, some posts o f provincial secretaries and additional
secretaries, and DCOs inDistrict governments).




                                                      101

5.30     The NRBpropose that the NES should have three broad streams o fprofessional specialization:

                 .   Economic Management
                      Social Sector Management
                      General Management

5.31     The compensation package for NES will be attractive. It i s proposed that NES officers (and
officers from the proposed parallel Provincial Executive Service) would be required to undergo
specialized training in foreign and domestic universities and later inthe newly created National School o f
Public Policy.

5.32     The Establishment Division's proposed reforms are similar. The Establishment Division has also
proposed the creation o fthree streams o f professionalized civil servants to some 1,000 policy formulation
positions in the Federal Government. These civil servants would be competitively selected and specially
trained, and would enter at grade level 19 with up to 20 percent recruited laterally. They would be
attractively compensated    - the Establishment Division i s proposing to double the basic salary plus
relevant grade allowances. Establishment Division also proposes three broad professional streams for
these senior staff.

5.33     These reform proposals are appropriate and can be a significant contribution towards increasing
institutional capacity at the higher levels o f government. It i s not clear, however, that it i s necessary to
create a dual structure o f compensation. Compensation for grades 19 to 22 should be raised substantially
and combined with monetization o f most in-kindbenefits. Those selected for N E S will have a fast track
advantage for promotion so there need not be separate pay scales. The key to both proposals i s the
National Executive Service, with a high bar before grade 19, sufficiently robust to deter any attempt to
use seniority as a justification for access, and a system o f very broad occupational groups beyond that
point. The proposals will encourage improved performance and greater s h l l s development because
promotion beyond grade 18 will not be at all automatic. The pursuit o f the professionalization objective
will be greatly helped by the full development the new National School o f Public Policy that has been
created with the following functions:
  .. To provide training and researchinpublic policy, law, economics, finance, and management science
     To provide consultancy and advisory services for the government
     To award degrees inPublic Policy, public sector management and related disciplines

5.34     The broad agreement between the NRB and the Establishment Division provides a rare
opportunity to move ahead on a pragmatic first step. The National Executive Service i s undoubtedly the
priority. Over time the need for provincial executive services may emerge. Government should seize the
initiative and decide the outlines o f the NES quickly. We also suggest that the reform should be
comprehensive that i s it should create the three agreed occupational groups immediately.

Managing the Transition

Addresspay

5.35     For the reasons noted above, the retrenchment o f large numbers o f employees i s not an overriding
priority. Large-scale retrenchment may not, in any case, be politically feasible at this time because o f
slow job creation in the rest o f the economy. However, the government finds itself in a situation when
significant pay reform, combined with highly selective future hiring, can be achieved largely through
fiscal space and attrition inthe low priority civil service groups.




                                                      102

5.36      There i s a prima facie strong case for significantly adjusting upwards the compensation for
higher managerial and professional grades, say grades 19 to 22. The numbers involved are small. For
instance, Federal employees in the grades 19 to 22 are only about 3,200, less than 1 percent o f total
Federal employment. The share i s somewhat higher inthe provinces. But overall these grades account for
less than 3 percent o f the government wage bill. So financial costs o f say a 100 percent real pay increase
over the next five years would be only 0.1 percent o f GDP. A substantial part o f this increase could be
financed fkom natural attrition o f lower grade staff focused on areas where employment i s excessive and
services are no longer needed. Restraints on recruitment coupled with attrition might be easier politically
than retrenchment as a method for creating fiscal space, but it i s recognized that the de-compression in
compensation that i s being recommended here i s nevertheless politically challenging.                          It is however
urgentlyneededifhigh-level managerial andprofessional slulls are to be attractedto the public sector.

5.37     It is much less clear whether there should be a general pay increases at lower levels without
loolung at job content, slulls required, and the comparator salaries inthe private sector. The recent moves
both at Federal and provincial levels to selectively give large special pay increases (for instance to tax
officials and doctors inrural areas) also suggest the advisability o f tailoring compensation to the nature o f
the responsibilities, hardship o f posts and reducing the incentives for corruption. There are other cases,
primary school teachers, lady health workers, and junior police officials who are extremely poorly paid
because their jobs are graded low. Detailed comparator studies o f private sector equivalents will, o f
course, be a necessary to guide future pay reforms for all grades.

5.38     Whether created by attrition and restraints on recruitment or otherwise, the opening o f the fiscal
space can provide an opportunity for reversing the squeeze in real salaries and de-compressing the
compensation differentials and further consolidating monetary allowances and in lund benefits into base
pay (See Box 5.1). However, before a new compensation policy can be formulated the future pension
liability must be assessed and actions taken to avoid a crisis in the future. The Ministry o f Finance has
initiated a comprehensive review o f reform options for the pension system. Such a review should identify
alternatives for controlling pension expenditure and designing and financing the new system, including
proposedmethods for data collection and for an actuarial valuation of cash flow and l i a b i l i t i e ~ . ~ ~

5.39     Provided future pension liability can be contained and civilian employment growth can be kept
down to say 1percent per annum, there should be room for increasing real salaries significantly over the
next five years. The challenge will be to use the fiscal space, on the one hand, for restoring compensation
closer to private sector for the higher grades, and on the other to adjust pay scales for groups for instance
primary school teachers, policemen, and lady health workers who are particularly poorly paid. Flat rate
increases, that do not decompress salaries or provide motivational important occupational groups, will not
serve the objectives o f the civil service reform.

Complete devolution

5.40     The unfinished administrative decentralization agenda provides a further opportunity for
government to initiate movement on an otherwise intractable issue. Completing devolution means, first
and foremost, transferring budgetary responsibility for salaries to the district governments. Thus salary
budgets should be transferred to the districts via account IV. In theory, from that time, they can select


67 Further work should include a review of the proposal by a Commissioner of the Securities and Exchange Commission o f
Pakistanthat would move employeesof autonomousbodies onto privately managed definedcontributionpensionschemes. The
scheme i s a little unclear, without a clear definition of coverage or o f the fate of existing (largely defined benefit) schemes.
Apparently the proposed law seeks to replace all existing schemes with a mandatory defined contribution scheme to be
administeredby PensionAdministration Companiesset up for the purpose. The Pakistan Society o f Actuaries has written to the
govemment expressingsome concern that the above andother significant issues havenot beenaddressed.




                                                               103

more o f one type o f staff or skill and less o f another as once the salaries are in account IV, then the
interim PFC awards to the districts require a single line itemtransfer that i s up to the districts to allocate -
withno reserve powersto the provinces to interfere.@

5.41     This is not a sufficient condition for local governments to value professionalism more than
patronage, but it i s undoubtedly a necessary one. Local governments at all levels will only start talung
their responsibilities for workforce management seriously as and when they consider themselves to be
fully the employer oftheir staff.

5.42     Furthering devolution might also assist in reducing the scale and frequency o f politically-
motivated transfers. One source o f the problem i s the willingness o f higher levels o f government to
intervene in the affairs o f lower levels. There are many reasonable arguments for maintaining a federal
presence at the provincial level, and similarly for maintaining a federal and provincial presence in the
districts. However, this provides a point o f entry for manipulation and, particularly for the new district
governments, leaves them subject to provincial pressures to transfer staff. As feasible, the appointment,
promotion and transfer (APT) authorities for staff should be progressively passed to the districts.
Currently, staff at grade 17 and above are appointed, promoted and transferred by the provincial
governments. APT authorities for grades 17 and 18 can be transferred to the districts within a year,
although this will require some clarification o f the role o f the Provincial Service Commissions or, more
logically, by the creation o fDistrict Service Commissions.

5.43     In addition, minimumtimes in post for senior federal, provincial and district staff should be
clearly specified with a requirement that public explanations be provided if staff are to be moved before
they have completed the minimum. By themselves, however, such specified minimum tenures are
unlikely to be sufficient to withstand the considerable political pressures for inappropriately rapid
 transfer^.^' Also, as noted inthe discussion o f devolution, one pretext for provincial micromanagement                        in
the districts can be removedby providing any increases inprovincial-district fiscal transfers inthe form o f
specific purpose grants. This would counter the argument that, since transfers are made on a block grant
basis, provinces must resort to various methods to ensure that districts address federal and provincial
policy priorities.




   The SLGO establishes as a "function and power o f the Zila council" to "approve the proposals o f the District Govemment for
   changes in the number o f posts o f officials and employees o f the decentralized offices o f District Administration and Taluka
   Municipal Administration and Town Municipal Administration as part of the budget statement" (Art. 39).

69 A requirement that staff remain inpost for at least 3 years is already included inthe Govemment Rules of Business (Rule 35
   (ii)
      Schedule 9). This i s not implemented however.



                                                                104

                                                Box 5.2: IncompleteDevolution

 OnAugust 14,2001, District, Tehsil andUnionCouncils were createdas administrative entities. These entities
 inherited a complex array o f staff, with complicated and overlapping authority relationships. The consequence
 of the creation o f the districts inAugust 2001 was a transfer o f responsibilities for various fimctions - it was
 not initially a transfer o fthe `employer' role from provincial to district governments.

 The newly created districts found that they had inherited three groups o f staff:
      1. They found themselves with some staff from federal employment groups (primarily District
           Management Group (DMG) and the Audits and Accounts Group).
      2.   In addition, employees who had earlier belonged to the rural district councils were posted to the
           districts.
      3.   The bulk o f their employees came from provincial employment groups - particularly Public Health
           Engineering, Rural Development, Local Government, and Education. Most provincial groups had
           some staff inthe districts.

 The "employer" of the first group remained undoubtedly the federal government. The "employer" o f most of
 the second group became unambiguously the district, although some staff at lower grades were assigned as
 administrators to the Union Councils (maximum o f 3 staff per UC at Grade 9). The employer for the third
 group i s distinctly ambiguous.'

 Performance evaluation responsibility for staff was transferred to the district as soon as they were created, by
 the simple logic that Annual Confidential Reports (ACRs) are initiated at the same level o f government as the
 administrative location o f the staff. District officers (DOs) heading each o f the functional service categories in
 district government are to have their ACRs initiated by their respective EDOs, countersigned by the DCO.
 Below this level, ACRs are initiatedby an employee's immediate superior.

 Career managementresponsibilities have just beentransferred. Legally, all the formerly-provincial staffbelow
 grade 16 now have their appointment, promotion and transfer (APT) decisions made through the districts.
 Senior staff o f the districts (Executive District Officers (EDOs) and other staff at above BPS16) will remain
 subject to the province or the federal govemment for appointment, promotion and transfer decisions.

 The ambiguities arise because budgetary responsibility for salaries has not yet been fully transferred.
 Currently, the staffs o f the Public Health Engineering, Rural Development, and Local Government employment
 groups - as well as other formerly provincial employee groups posted to the districts - are paid by the province
 through the mechanism o f Account I(a provincial account). Under this arrangement, sanction is sought in
 advance from the province for the release o f the funds for their salary payments. Thus, the salaries are
 effectively paid from the provincial budget.            Until district governments consider themselves to be
 unambiguously the employers of their own staff, then they will have few incentives to manage their workforce
 effectively.


5.44     If education and health expenditures are to be increased and MDG outcomes achieved (see
Chapter 3), there will necessarily be a significant increase in the number of teachers and health care
personnel. Clarifying the employment authorities for district governments, and gearing up their capacity
for the significant challenges that lay ahead, will be crucial.




                                                           105

                               REFERENCESD3IBLIOGRAPHY

Azfar, O., Kahkonen, S. and Patrick Meagher. 2001. "Conditions for Effective Decentralized
Govemance: A Synthesis of ResearchFindings". Center for Institutional Reform and the Informal Sector,
UniversityofMaryland.

Government o f Pakistan. 2001a. A Debt Burden Reduction and Management Strategy. Debt Reduction
and Management Committee, Finance Division. March2001. (Pakistan DebtReport).

Government o f Palustan. 200lb.Interim-Poverty Reduction Strategy Paper (I-PRSP). Jointly preparedby
the Ministryof Finance and PlanningCommission, Islamabad. November 2001.

Government o f Palustan. 2001c. Summary of the Ten Year Perspective Development Plan 2001-11 and
Three YearDevelopment Program 2001-04. PlanningCommission, Islamabad. September2001.

Government of Pakistan. Economic Survey. Various Issues.

Halcrow and ARCADIS. 2002. "Pakistan Water Resources Strategy Study". ADB TA 3130 PAK, (Draft
Report), June.

Hasan, Parvez. 1998. Pakistan's Economy at the Crossroads - Past Policies And Present Imperatives.
Karachi. OxfordUniversityPress.

InternationalMonetary Fund."Accountable Fiscal Management Framework". Washington, DC.

Shah, Anwar. 1996. "Fiscal Federalism in Pakistan: Challenges and Opportunities". World Bank.
Washington DC.

State Bank of Palustan. 2002. Annual Report 2001-2002.

World Bank. 2002a. Pakistan Development Policy Review: A New Dawn?. Report No. 23916-PAK.
South Asia Region. Washington, DC.

World Bank. 2002b. Pakistan Poverty Assessment - Poverty in Pakistan: Vulnerabilities, Social Gaps,
and Rural Dynamics. Report No. 24296-PAK. SouthAsia Region.Washington, DC.

World Bank. 2001a. Reforming Punjab's Public Finances and Institutions. Report No. 20981-PAK.
South Asia Region. Washington, DC.

World Bank. 200lb. "Methodology for Functional Reviews". Paper produced for the Russian Federation
Government. Washington, DC.

World Bank. 1998a. Pakistan: A Framework for Civil Service Reform in Pakistan. Report No. 18386-
PAK. South Asia Region. Washington, DC.

World Bank. 1998b.Pakistan Public ExpenditureReview: Reform Issues and Options. Report No. 18432-
PAK. South Asia Region. Washington, DC.

World Bank. 1998c. Pakistan - Towards A Health Sector Strategy. Report No. 16695-PAK. South Asia
Region. Washington, DC.



                                                 106

World Bank. 1995. Pakistan Poverty Assessment. Report No. 14397-PAK. South Asia Region.
Washington, DC.

World Bank. 1994. Pakistan - Irrigation and Drainage: issues and options. Report No. 11884-PAK.
South Asia Region.. Washington, DC.

World Bank. 1968. Water and Power Resources of West Pakistan: A Study in Sector Planning. John
Hopluns Press.




                                              107

                                                                                                   Annex A
                                                                                                Page 1of 4

                                      GOVERNMENT PAKISTAN
                                                        OF
                                           FINANCEDIVISION

                                              MARCH    14,2001

                             DEBTREDUCTIONAND MANAGEMENT             STRATEGY

1.       Pahstan faces grave debt issues, which have been in the malung for decades and threaten the
economic future o f the country. Moreover, the country faces not one but two distinct debt problems: the
overall public debt problem and the external debt problem. The external debt problem has greater
immediacy because o f the concentration o f large external debt service payments, amounting to nearly
US$21billion over July 2000 -June 2004, which cannot be metwithout exceptional financing from IMF
and other multilateral institutions and further debt rescheduling from the Paris Club.

2.       Pakistan's total external debt obligations (including debt obligations to residents in foreign
exchange) increased from less than US$ 10 billion in June, 1980 to US$ 20 billion in June, 1990 and
peaked at US$43 billion in May 1998, just before the imposition o f economic sanctions forced Pakistan
to freeze all individual foreign currency accounts. Even after these large-scale foreign exchange liabilities
(totaling approximatelyUS$ 9 billion) were extinguished, external debt and foreign obligations, including
those to residents, stood at US$ 37 billion at end-2000, equivalent to nearly 300 percent o f foreign
exchange eamings (the normal sustainable level i s 200 percent). External debt, narrowly defined, stands
at US$ 34.5 billion. The nominal value o f overall public debt grew from Rs. 155 billion in 1980 to Rs.
800 billion in 1990 and rose to Rs. 3,570 billion at end-2000, raising the burden o f public debt from 66
percent o f GDP in 1980 to well over 100 percent at present. Interms o f a more relevant indicator, public
debt i s now around 625 percent o f government revenues compared to 400 percent in 1980 (whereas the
sustainable limit is approximately 350 percent). Pakistan has a higher public debt burden, by a large
margin, than any large developing country. Moreover, the above figures do not include any estimate of
contingent liabilities o f the Federal Government, which comprise guarantees to P P s and other private and
public bodies; and large unfundedlosses o f public corporations including WADPA, KESC, Pakistan Steel
Millas well asnationalized banksandpublic financial institutions.

3.       Given the absolutely unsustainable level o f the country's debt and urgent need to reduce the debt
burden, the Chief Executive announced in hispublic address on December 15, 1999 the establishment o f
the Committee on Debt Reduction and Management (Debt Committee) headed by Dr. Parvez Hasan, a
noted Pahstani Economist and former director o f the World Bank. The Debt Committee's task was to
analyze (a) factors underlying debt growth, (b) impact o f high public debt interest payments on
Government development and social spending (c) implications o f large contingent liabilities o f the
Government for future debt and fiscal management; and (d) short-term consequences o f large gross
borrowing requirements for balance o f payments management. The Committee was also asked to review
the existing framework for debt contracting, recommendmedium and long-term goals for the reduction o f
burden o f public as well as external debt and specify institutional arrangements for a debt management
system. The Committee has worked for nearly a year. It submitted its interimrecommendations to the
Minister for Finance in April 2000, and has now finalized its Summarv Report. The Chairman o f the
Committee briefed the Cabinet on November 1,2000 and February 15,2001 upon submission o f the draft
Summary Report.

4.       The Committee has analyzed the evolution o f debt burden in Palustan and found that
extraordinarily highpublic debt and very large external debt have stemmed from the following factors:

         (a)   Large and persistent fiscal and current account balance o f payments deficits.

                                                                                                     Annex A
                                                                                                  Page 2 of 4


         (b) Imprudent use of borrowed resources (such as wasteful government spending, resort to
             borrowing for non-development needs, undertaking o f low economic priority development
             projects and poor implementation o f foreign aided projects) resulting in low productivity o f
             resource use.
         (c) Weakened debt carrying capacity in terms o f stagnation or decline in real government
             revenues and exports.
         (d) Risingreal cost o f government borrowing, bothdomestic and foreign.

5.       The work o f the Debt Committee has confirmed that the high and growing debt burden i s a
fundamental cause o f many o f Palustan's problems, including the slowdown o f overall economic growth,
increase inpoverty incidence, the low levels o f social welfare indicators, declining investor confidence in
the long-term prospects o f the country and the recurrent financial crises. The Committee has especially
highlighted the close links between high debt burden and slow economic growth. First reduced net
external resources are constraining investment. Since reliance on net external flows inthe past was totally
unsustainable, the Committee has concluded that national savings must finance 90 percent o f investment
at least for the next decade. Second, large government domestic borrowing i s keeping real interest rates
highand discouraging private investment. Third, private investment is also being adversely affected by
the shortage o f complementary infrastructure and inadequate s h l l development, attributable directly to
shortage o f fiscal resources.

6.       Since the debt problem has arisen because o f poor economic decisions, postponed reforms and
weak governance over a sustained period, there i s no easy remedy for debt burden reduction. The
satisfactory resolution o f debt issues will take both time and aggressive policy action. It will also involve
painful choices. The Debt Committee recognizes that, desirable though it may be, reduction o f debt to a
sustainable level cannot be the only economic goal. There i s also an urgent need to revive economic
growth, which on a trend basis has been only around 4 percent per annum over the last five years. The
Committee has therefore stressed that Pakistan needs a comprehensive and integrated economic revival
and debt reduction strategy. The twin goals o f this strategy should be a notable reduction in the debt
burden and a significant increase inthe rate o f economic growth over the medium term. This implies that
inthe short-run (that is, for up to two to three years), some difficult tradeoffs between debt reduction and
economic growth will be unavoidable.

7.       The present Government has already been vigorously pursuingmany elements o f the strategy
recommended by the Debt Committee, notably, reducing fiscal deficits, increasing government revenues,
expanding exports, accelerating privatization and strengthening the oil and gas and information
technology sectors. It i s also attaching high importance to maintaining a program with the IMF, with a
view to moving towards a three-year Poverty Reduction and Growth Facility (PRGF), which will provide
the basis of exceptional assistance from the World Bank and Asian Development Bank and further
rescheduling from the Paris Club till 2004.

8.       Nevertheless, the Committee's work has provided many fresh insights concerning the nature o f
Pakistan's debt problem and the economic challenges Pakistan faces inthis decade. These include:

         i) A focus on productivity enhancement through both structural reforms and improved
             governance i s crucial for reviving economic growth, given financial constraints.

         ii)Whilesteadyreductioninthefiscaldeficitandsharpincreaseingovernmentrevenuesare
             critical, the pattern o f fiscal adjustment and effectiveness o f public spending will determine
             whether government investment in human and physical capital can increase significantly and
             whether the economic and social returns to public borrowing will become significantly

                                                                                                  Annex A
                                                                                                Page 3 of 4

             positive. The balance between development and defense expenditure needs to be restored and
             the Government must eliminate borrowing for non-development spending, which i s lowering
             national savings and compounding the debt problem.
       iii) RealcostsofdomesticandforeignborrowingbytheGovernmenthaverisentoveryhigh
             levels inrecent years and needto bebrought down sharply.

       iv)   The huge public debt burden poses as big a threat to Pakistan's long- term financial and
             economic health as does the large external debt.

9.       The Debt Committee is confident that with strong financial discipline, forceful structural reforms
and improved governance, economic growth can be revived to a soundly based 5.5 percent per annum by
2003-04. At the same time, with adequate international support, the real external debt burden can be
reduced by 30 percent and the need for exceptional assistance from the IMF and Paris Club rescheduling
phased out by 2004. If Palustan can safely navigate the turbulent economic waters during the difficult
transition period o f the next three or four years, and if major unanticipated exogenous shocks do not
occur, it will have laid the basis o f high, self-sustaining and equitable economic growth, a combination
which it has never achieved in the past. Pakistan's population growth rate i s expected to drop gradually
from the present level o f 2.2 percent to 1.8 percent per annum by 2009-10, reflecting the demographic
transition that has commenced. There are thus good prospects o f achieving a per capita GDP growth rate
o f over 4 percent per annum inthe second half o f this decade -higher than inthe 1980s and with a much
better quality o f growth.

10.      The Committee has examined the issues involved in debt default and debt reduction. For clearly
stated reasons, the Committee believes that default i s not a real option. Debt reduction also faces many
practical problems. Takmg a pragmatic approach, therefore, the Committee has not ruled out seelung
softer terms at Paris Club reschedulings. It hopes, however, that Palustan can resolve its external debt
problems while maintaining and improving its creditworthiness.

11.      The Government has accepted the broad thrust o f the Debt Committee's recommended strategy
and has decided on a number of steps to implement its key recommendations. To this end, the following
steps are being taken to formalize a Debt Reductionand Management Strategy:


     1)    Public and external debt reduction goals will be made an integral part o f the Government's
           economic revival and poverty alleviation strategy. Specifically, the Government has adopted
           the goals of reducing external debt burden to 200 percent o f foreign exchange eamings by 2005
           and reducing public debt burdento 350 percent o f government revenues by 2010. These targets
           will become part o f the ten-year macroeconomic framework being prepared by the Planning
           Commission.

     ii) ThebuildingofgrossforeignexchangereservestoalevelofUS$5billionbymid-2004willbe
           a part o f external financing plans.

     iii) Theattainmentoftheadmittedlyambitiousrevenueandexportgrowthtargetsover2000-2004
           (60 percent and 50 percent respectively) will remain a top Government priority.

     iv)   The Government will also aim at eliminating the imbalance between current revenues and
           current expenditure (that is, borrowing for non-development purposes), which at present i s 3.7
           percent o f GDP, by 2004. Consequently, the overall fiscal deficit will be reduced to 3 percent
           of GDP by 2004 and further to 2 percent by 2010.

     v)    To facilitate the above goals, defense spending will be kept constant inreal terms, for the next
           four years.

                                                                                              Annex A
                                                                                            Page 4 of 4



vi)  As the overall fiscal deficit is gradually reduced, the share of interest payments in total
      expenditure will decline from 33 percent in 2000 to 16.4 percent by 2010. The fiscal revenues
     thus freed from lower interest payments and defense will be used to sharply increase the share
     o f social and development spending intotal expenditures.

vii) Bringing down the burden o f external debt significantly by 2005 will permit an increase in
     normal extemal borrowing thereafter. However, net reliance on extemal resources will be kept
      to less than 10 percent o f total investment during 2000-10. This compares with net borrowing
      from extemal resources o f over 20 percent inthe 1980s and over 30 percent inthe 1990s.

viii) The Government will make utmost efforts to obtain US$ 6 billion in exceptional assistance
      from the IMF, World Bank and ADB on soft terms in order both to avoid debt default and
     reduce the onerous burden o f extemal debt payments.

ix)   Duringthe next four years normal term borrowing from all sources, including IBRDand ADB,
      will be kept to a minimum.

x)    The Govemment will aim at reducing domestic borrowing costs in real terms to three to four
      percent per annum, from the average o f nearly 9 percent during past two years, by sharply
     reducing its net domestic borrowing and restructuring its borrowingplans.

xi)   The Govemment will be extremely selective in undertaking fresh foreign aided projects in the
      short to medium term and apply strict criteria o f financial and economic viability for screening
      these projects. Terms and conditions o f foreign loans will also be disclosed to the public at
      large.

xii) Borrowing for local currency from international sources will be phased out in the next two
      years. Thus, development projects requiring mainly rupee funding will rely on indigenous
      financial resources and local consultants.

xiii) The Govemment will set up a Debt Policy Coordination Office (DPCO) in the Ministry of
      Finance, which will report to the Cabinet every year on progress in attaining debt reduction
      goals, provide broad leadership on debt management issues to all parts o f govemment and
      closely monitor and evaluate contingent fiscal liabilities.

                                                                                                    Annex B
                                                                                                Page 1of 20

                                  MEDIUM-TERM FRAMEWORK

TechnicalAnnex

1.      This annex provides technical information about the medium-term macroeconomic and fiscal
framework inthe two scenarios in Chapter 2, and a sensitivity analysis o f the base case.

2.      The medium-term budgetary framework presented in this report consolidates the budgetary
accounts o f both Federal Government and the Provinces. The presentation o f a consolidated fiscal
framework i s for illustrative purposes and allows for an assessment o f the total fiscal impact on
macroeconomic balances and expected progress towards public debt sustainability. The consolidated
fiscal space estimates allow for an assessment o f the room for increased spending in priority areas by all
levels o f government with progress towards the government's goals o f growth facilitation, poverty
reduction, and social development. Drawing on this framework we present findings about the total
countrywide resource availability, illustrate the public expenditure level and pattern for the country as a
whole and highlightfuture expenditure options.

3.      Inorder to operationalize the consolidated budgetary framework for use by the central and
provincial governments, the authorities would need to break it down into separate real budgetary
frameworks for the central government, the provinces o f Punjab, Sindh, North West Frontier, and
Balochistan, and the areas o f A X and the Northern Areas. This will include making overall and sector-
specific spending projections at the provincial and district level consistent with the consolidated totals we
present and the associated necessary transfers from the center to the provincial governments.

I.THEBASELINESCENARIO

4.      The baseline scenario for macroeconomic policies for the period FY 02/03 - FY 06/07 assumes a
continuation o f the reform-oriented policies pursued by the GoP in recent years. The gains seen in
achieving macroeconomic stabilization and reducing fiscal imbalances are expected to be consolidated
and strengthened further. However, it also envisages a concerted approach to reach the goals for poverty
reduction and facilitation o f economic growth. The macroeconomic framework for the first projection
year, FY 02/03, i s the same as inthe PRGFprogram agreed with the IMF.

11. KEYASSUMPTIONSFORTHEPROJECTIONS

A.      Overallmacroeconomic framework (TableA.1):

        Real GDP growth (at market prices) rises from 4.4 percent inFY 01/02 to 5.9 percent o f GDP in
        FY 06/07.
        Inflation stays at 4.0 percent throughout FY 02/03 to FY 06/07, up from 2.7 percent inFY 01/02
      w Gross national savings rise from 16.4 percent o f GDP in FY 01/02 to17 percent o f GDP in FY
        06/07, reflecting continuing fiscal consolidation
        Investment (gross fixed capital formation) rises from 12.3 percent o f GDP in FY 01/02 to 16.6
        percent o f GDP inFY 06/07 with proportionally similar increases ininvestment by the public and
        private sectors (the latter includes public enterprises inthe statistics)
      w The ICORs (Incremental Capital Output Ratios) decline slightly from the FY02 level (3.1) to 3.0
        byFY07. The three-year average ICORs decline from 3.5 to 3.1.
        Constant real effective exchange rate        - which in the relatively low inflation environment
        projected here implies gradual adjustment inthe nominal exchange rate insmall steps.

                                                                                                Annex B
                                                                                             Page 2 of 20



B.   Fiscal projections (Table A.2):

5..  Themost critical assumptionsfor creation of sufficientfiscal space are:

     Revenue rises from 16.6 percent o f GDP in FY 01/02 to 18 percent in FY 06/07. The revenue
  .  increase comes entirely from tax collection, in particular tax collection by the CBR (1.4 percent
     o f GDP).
     Contingent liabilities are contained and sharply reduced, reflecting in particular improvement in
  .  operational performance o f public enterprises. Payments o f contingent liabilities decline from 1.4
     percent o f GDP inFY 01/02 to 0.3 percent inFY 06/07.
     Defense spending i s constant inreal terms at the FY02/03 level.
     Interest payments decline from 6.6 percent o f GDP in FYO1/02 to 4.6 percent o f GDP in
     FY06/07. This decline reflects mainly the positive public debt dynamics inthe overall - favorable
     - macroeconomic framework and stagnant domestic public debt in the face o f a growing
     economy. It also reflects slowly declining interest rates on domestic public debt (the weighted
     average interest cost on domestic debt declines by 0.1 percentage point per annum).

6. Thefollowing assumptionsapply to the use of thefiscal space:

     Non-interest non-defense non-contingent liability spending - which includes the high priority
  .. spending on growth revival, poverty alleviation and social development - rises from 11.1percent
     of GDP inFY 01/02 to 14.5 percent inFY 06/07.
     Social and poverty-related spending rises from 3.6 percent o f GDP in FY 01/02 to 5 percent in
     FY 06/07.
     Development spending i s targeted to increase from 3.4 percent o f GDP inFY 01/02 to 4.7 percent
     in FY 06/07.    In this regard it is critical to improve governance (inter alia in procurement),
     implementation capacity, and expenditure control and monitoring. To the extent that such
     improvements cannot be achieved in line with the spending plan, actual spending would need to
     be lower to avoid waste and unnecessarily large public debt

7. Other assumptions:

     The government i s successful in containing overall expenditure pressures. Total spending
  ...declines from 23.4 percent o f GDP inFY 01/02 to 22.3 percent inFY 06/07. This requires strong
     political commitment and improved institutional capacity for effective expenditure control.
     Foreign grants decline from the high level in FY 01/02 but remain in the range o f 1.0 - 1.2
     percent o f GDP (the same as inFY 99/00 - OO/Ol).
     New foreign loans are mainly on concessional terms.
     Net privatization receipts to the fiscal budget average 0.3 percent o f GDP per year

                                                                                                                                   Annex B
                                                                                                                               Page 3 of 20

                               Table A.l: Pakistan: Baseline Scenario, 1999/2000-2006/07

                                                                                                Est.Prov.Est.                Proj.
                                                                   1999100         2000/012001/02 20021032003104 2004/05          2005/062006/07

                                                                                         (annual changes inpercent)
Outputand prices
     Real GDP at factor costs                                           3.9             2.2     3.4       5.1    5.3        5.5       5.7    5.9
     Real GDP at marketprices                                           4.2             2.6     2.8       5.8    5.3        5.5       5.7    5.9
     Consumer prices (p.a.)                                             3.6             4.4     2.7       3.1    4.0        4.0       4.0    4.0

                                                                                            (in percent of GDP)
Savings and investment
     Gross national savings                                            14.1            13.6    17.4     21.3    16.7       18.0      18.3   18.5
            Govemment                                                  -2.5            -2.2    -0.9       1.6    0.1        0.9       1.7    2.2
            Nongovernment                                              16.5            15.8    18.3     19.7    16.7       17.1      16.5   16.3

     Gross capital formation                                           16.0            15.5    14.7     15.5    16.5       17.1      17.7   18.3
            Govemment                                                   3.O             2.6      3.5      3.2     3.5       4.0       4.5    5.0
            Nongovemment l/                                            11.3            11.3      9.6      9.9   11.0       11.5      11.6   11.7
            Change in stocks (nongovt)                                  1.6             1.6      1.6      2.4    2.0        1.6       1.6     1.6

Savings - Investmentbalance                                            -2.0            -1.9      2.7      5.9     0.2       0.9       0.6    0.2
            Govemment                                                  -5.5            -4.8     -4.4     -1.6    -3.4      -3.1      -2.8   -2.8
            Nongovemment(inclchangein stocks)                           3.5             2.9      7.0      7.5     3.6       4.0       3.4     3.0

                                                                                            (inpercent of GDP)
Public finances
     Revenue (including grants)                                        17.3            17.3    19.3     20.8    18.5       19.3      19.6   19.8
     Revenue                                                           16.3            16.2    17.0     17.9    17.3       17.6      18.0   18.3
     Expenditure                                                       22.8           21.4     23.6     22.3    21.8       22.4      22.4   22.6
          Non-interest, non-defense,no-SOE spending                     9.8            10.2    10.9     12.0    11.6       12.6      13.4   14.3
      Overallbalance (including grants)                                -5.5            -4.1     -4.4     -1.6    -3.4      -3.1      -2.8    -2.8
     Primary balance (including grants)                                 2.3             2.8      3.8      3.6     1.4       1.5        1.4    1.0
     Gross public debt 21                                             100.9          108.3    100.1     91.8    83.4       80.3      74.3   71.0
     Public debt inpercent of revenue                                 619.6          670.5    588.9    512.0   481.3      456.9     412.1  388.1

                                                                                            (in percent of GDP)
Extemal sector
      Merchandise trade balance                                        -2.3            -2.2     -0.5     -0.8    -1.4      -1.6      -1.7    -1.8
            Merchandise exports                                        13.4            15.2    15.4     15.9    15.4       15.4      15.4   15.4
            Merchandise imports                                        15.8            17.4    15.9      16.7   16.8       17.0      17.1   17.2
     Current account excl. off. transfers                              -3.5            -3.3      0.2      4.4    -0.9      -0.7      -1.0    -1.3
     Current account incl. off. transfers                              -2.0            -1.9      2.7      5.9     0.2       0.9        0.6    0.2

                                                                           (in percent of exports of goods and nonfactor services)

      Extemalpublic andpublicly-guarantd debt 2/                      308.7          289.4    282.6    229.6   211.3      199.2     184.3   172.5
          Debt service                                                 46.9           28.2     34.0     26.4    27.6       22.7      17.0   15.1

     Grossreserves(in US.dollar ml)                                     908          1,679    4,330    9,486   9,212      9,569    10,679 11,669
            Inmonths ofimports ofgoods and services                     3.8              1.7     3.7      7.0     6.3       6.0        6.2    6.2

Memorandumitems:
     Real effective exch. rate (% change)                              -0.6            -2.6     -1.2      6.0     3.9       0.0        0.0    0.0
     Real per capita GDP ("hchange)                                     2.0              0.3     0.7      3.6     3.1       3.3        3.5    3.7
     GDP at marketprices (Rs bl)                                      3,147          3,423    3,629    4,018   4,400      4,828     5,307  5,845

Source: Data providedby the Pakistani authorities; and World Bank staff
l/Includesstate-ownedenterprises.
2/ Excludes SBP.

                                                                                                                                 Annex B
                                                                                                                              Page 4 of 20



                       Table A.2: Pakistan:ConsolidatedGovernmentBudget,2000/01- 2006/07

                                                                          2000/01   2001102   2002103    2003/04    2004105   2005/06 2006107
                                                                                              Prov. Act.  Proj.      Proj.     Proj.     Proj.

                                                                                                 (in percent of GDP)
Revenueand grants                                                             17.3       19.3       20.8       18.5     19.3       19.6     19.8
      Revenue                                                                 16.2       17.0       17.9       17.3     17.6       18.0     18.3
        Tax revenue                                                           12.9       13.0       13.8       13.6     14.0       14.4     14.6
               Federal                                                        12.3       12.5       13.3       13.0     13.4       13.7     13.9
                  CBR revenue                                                 11.5       11.1       11.5       11.6     11.9       12.2     12.4
                  Petroleumsurcharge                                           0.5       0.81        1.2        1.o       1.1       1.1      1.1
                  Gas surcharge                                                0.4       0.49        0.5        0.3      0.4        0.4      0.4
               Provincial                                                      0.6        0.5        0.5        0.6      0.6        0.7      0.7
      Nontaxrevenue                                                            3.3        4.0        4.1        3.7       3.6       3.6      3.7
Grants                                                                          1.2       2.3        2.8        1.1       1.7       1.6      1.5

Expenditure                                                                   21.7       24.0       22.3       21.8     22.4       22.4     22.6
      Currentexpenditure                                                      19.6       19.0       19.7       17.7     17.9       17.7     17.5
        Federal                                                               14.8       14.2       14.9       13.0      12.9      12.3     11.8
               Interestpayments                                                6.8        6.8        5.2        4.8       4.6       4.2      3.9
                  Domestic                                                     5.4        5.1        4.0        3.9       3.7       3.5      3.3
                  Foreign                                                       1.5       1.7        1.2        0.9       0.9       0.7      0.6
               Defense                                                         3.8        4.1        4.0        3.6       4.0       4.0      4.0
               Runningof the civil govemment                                   2.1         1.6       1.5        1.6       1.7       1.7      1.8
               Pensions for defense and civil govemment                        0.9        0.7        0.9        0.9       0.9       0.9      0.9
               Subsidies                                                       0.6        0.5        1.2         1.4      0.9       0.6      0.3
                  Power Sector                                                                                  0.8       0.6       0.4      0.2
                  Other SOEs                                                                                    0.5       0.2       0.1      0.0
                  Non SOEs                                                                                      0.1       0.1       0.1      0.1
               Grants                                                           0.5       0.6        0.6        0.7       0.7       0.7      0.7
         Provincial                                                            4.9        4.8        4.8        4.7       5.0       5.3      5.7
      Developmentexpenditure andnet lending                                     2.1        3.5       2.7        3.7       4.1       4.5      5.0
               Public Sector DevelopmentProgram                                 2.6        3.5       3.2        3.5       4.0       4.5      5.0
               Net lending                                                     -0.5       0.0       -0.6        0.2      -0.1       0.0      0.0

      Payment of contingentliabilities                                                     1.4                   0.5      0.3        0.2      0.1

Statistical discrepancy                                                        -0.3       -0.4       0.1         0.0      0.0        0.0      0.0

Budgetbalance (excludinggrants)                                                -5.2       -6.6      -4.5        -4.5     -4.8       -4.4     -4.3
Budgetbalance (including grants)                                               -4.1       -4.4      -1.6        -3.4     -3.1       -2.8     -2.8


      Primarybalance(including grants)                                          2.5        2.0       3.6         1.4       1.5       1.4      1.o

 Financing                                                                      4.1        4.4       1.7         3.4      3.1        2.8      2.8
      Extemal                                                                   2.3        1.4      -0.2         1.o      0.1        0.6      0.4
      Domestic                                                                  1.7        2.7       1.6         2.1      2.7        1.9      2.2
      Privatizationproceeds                                                     0.0        0.2       0.3         0.2       0.3       0.3      0.3


Memorandum items:
    Primarybalance(excluding grants)                                            1.6        0.1       0.7         0.2      -0.2      -0.2     -0.5
    Revenuebalance(excludinggrants)                                            -3.5       -2.0      -1.8        -0.4      -0.4       0.3      0.8
    Socialandpoverty-relatedexpenditure I/                                      0.0        4.8       5.1         5.4       6.0       6.5      7.1

     Spendingexcl interest, defenses, subsidies, cont. liabilities, % GDP   10.2      10.9       12.0       11.6      12.6      13.4       14.3
      Share of total spending(percent)                                      47.4      46.0       53.5       52.9      56.2      59.9       63.3
      Nominal GDP (in millionsof Pakistanirupees)                          3,423     3,629     4,018.1    4,400.3    4,828.0   5,307.4   5,845.3

~~Source:Nationalauthorities; and staffprojectionsand calculations.
  11Education,healthandpopulation planning, Law and Order, Ruralelectrification,Water supply and sanitation, Rural Development,Irrigation, Roads,
  highwaysandbridges, and "Other poverty-relatedexpenditure".

                                                                                             Annex B
                                                                                          Page 5 of 20

8. Key results:

  . Reflectingthe totality o fthe above assumptions the primary balance (excluding grants) rises from
    broadly balance inFY 2001/02 to 0.9 percent o f GDP on average inFY03 -FY07.
    The revenue balance (current revenue - current spending) marks an even stronger improvement
    by a full 2.7 percent of GDP over the projectionperiod, from a deficit of 2percent o fGDP inFY
    2001/02 to a surplus o f 0.7 percent o f GDP in FY 06/07. Importantly, this leaves room for
    financing parts o f public investment from savings on current operations which marks an
    important turnaround from many years o f government dissaving.
    The overall fiscal deficit including grants and payment o f contingent liabilities drops from 4.2
  . percent o f GDP in FY 2001/02 to 3.3 percent o f GDP in FY 06/07, consolidating the fiscal
    adjustment achieved so far under the PRGFprogramwith the IMF.
    This scenario shows a steady improvement in public debt indicators. Annual public debt service
    inpercent of fiscal revenue falls fi-om 55 to 32 %by the end ofthe projectionperiod. The public
    debt/GDP ratio drops from 99 to 72 % and the Present Value o f the public debt/GDP ratio drops
    from 80 to about 60 %.
    The structure o f public expenditure changes considerably towards high priority areas. The share
    on non-interest non-defense non-contingent liability spending in total spending rises from 48
    percent inFY02 to 65 percent inFY07.


C.  Balanceof payments(Table A.3):

9.. Key assumptions:

    Stable real exchange rate
    Average annual export volume growth: 6.1 percent. This reflects the exchange rate assumption
  . and continued structural reforms to remove impediments to improvements inproductivity and to
    diversifying into higher-value added exports
    Average annual import volume growth: 9.5 percent, reflecting the pick-up in economic growth,
  . the expanding government development spending and increasing private sector investment-
    relatedimports.
    Private transfers: it i s assumed that the surge in`workers' remittances in FY02/03 i s temporary
    and reflects a portfolio reallocation. Hence workers' remittances inFY03/04 are projected to fall
  . back to the FY01/02 level (about $2.4 billion) and then grow slowly in line with international
    inflation.
    Official grants are projected to decline from 2.2 percent o f GDP inFY 02/03 to 1percent o f GDP
  . inFY 06/07, throughout the period lying considerably below the average for the last three years
    (1.8 percent o f GDP).
    Continued program financing (on concessional terms) from the World Bank and the AsDB
    throughout the periodwhereas the IMF's PRGF financing stops inFY 04/05.
    Total net foreign financing (disbursement o f grants and loans plus rescheduling minus
    amortization payments) i s projected to decline gradually from 4.6 percent o f GDP inFY 01/02 to
    2.7 percent o f GDP in FY 06/07. Actual disbursements o f new loans are projected to remain in
    the 2 % - 3 1/3 percent o f GDP range seen in recent years reflecting willingness from donors to
    provide financial assistance for the government's reform program. The projections imply that
    disbursements grow in foreign currency terms but remain fairly constant in terms o f percent o f
    donors' GDP.

                                                                                                                                 Annex B
                                                                                                                               Page 6 of 20



                     Table A.3: Pakistan: BaselineScenario: Balance of Payments, 2000/01-2006/07

                                                                                             Proj.        Proj.       Proj.     Proj.      Proj.
                                                                    2000101       2001102   2002103     2003104     2004105    2005/06    2006107

                                                                                             (in millions of US. dollars)
Current account (excluding official transfers)                            -1,951         96     3,029        -651        -602       -847     -1,198
Current account balance (including official transfers)                    -1,112      1,591     4,029          190         793       570        239
     Tradebalance                                                         -1,268      -292       -536       -1,040      -1,306    -1,507     -1,762
            Exports f.0.b.                                                 8,934      9,140    10,889       11,759      12,663    13,648     14,738
            Imports f.0.b.                                             -10,202       -9,432   -11,425     -12,800      -13,969   -15,155    -16,500
     Services (net)                                                       -3,143     -2,617    -2,172       -3,230      -2,977    -3,082     -3,239
            of which: interest payments                                   -1,657     -1,579    -1,272       -1,117      -1,053      -902       -799
     Private transfers (net)                                               2,460      3,005     5,737        3,619       3,682     3,742      3,803
     Official transfers (net)                                                839      1,495     1,000          841       1,395     1,417      1,437


Capital account                                                             -623     -2,322    -1,529       -2,481      -1,426      -676       -418
     Public medium- and long-termcapital                                    -652     -1,613    -1,497       -1,590      -1,254      -832       -733
            Project and nonproject loans                                    -332       -983    -1,696       -1,012        -879      -501       -592
                  Disbursements                                            1,463        531       723          788         855       967      1,052
                  Amortization                                            -1,795     -1,514    -2,419       -1,800      -1,734    -1,468     -1,644
            Commercial banks and IDB (net)                                   -76       -224      -158          -33         -16        -16        -16
            Other                                                           -244       -407       357         -545        -359      -315       -125
     Public sector short-term(net)                                           -59     -1,064      -233         -564        -117      -100       -100
     Private medium- and long-term                                           343        -80       221          218         475       670        744
     Private short-term (including errors & omissions)                      -256        435       -19         -551        -530      -415       -329


Overall balance (before debt relief)                                      -1,735       -731     2,501       -2,297        -633      -106       -179


Financing                                                                  1,735        734    -2,501       -2,297         633        106        179
Reserveassets (increase -)                                                -1,088     -3,082    -5,267          -11        -892    -1,541     -1,442
     State Bank of Pakistan (including FE-25s)                              -729     -2,717    -5,916           99        -540    -1,309     -1,194
     Deposit money banks                                                    -359       -365       649         -110        -352      -232        -248
Fundrepurchases                                                             -194       -194      -418         -526        -362       -135       -111
Net exceptionalfinancing                                                   3,017      4,007     3,185        2,842       1,941     1,850      1,800
     Arrears (increase+)                                                    -525           0         0           0            0         0          0
     Rescheduling                                                          1,587      1,210     1,009          100         100        100        100
            of which: Private Sector Involvement                                0          0      100          100         100        100        100

     Rollover o f foreign deposits with banking system                     1,676      1,314       900        1,200         500        500        500
     Programfinancing from 1FIs                                              279      1,367     1,090        1,342       1,121     1,000         950
            World Bank                                                          0       698       213          500         500        600        600
            AsDB                                                                0       185       408          400         400        400        350
            JMF                                                              279        484       469          442         221           0          0
      Privatization receipts                                                    0       117       186          200         220        250        250


Memorandum items:                                                                                (in percent of GDP)

   Current account balance (including official transfers)                    -1.9        2.7       5.9          0.2         1.o        0.6        0.2
   Current account (excluding official transfers)                            -3.3        0.2       4.4         -0.9        -0.7       -1.o       -1.2
   Current account (excluding all transfers)                                 -7.5       -4.9      -4.0         -5.6        -5.2       -5.2       -5.2
   Interest payments1GDP                                                     -2.8       -2.7      -1.9         -1.5        -1.3       -1.0       -0.8
   External primary balance (incl. off. transfers)/GDP                        0.9        5.4       1.7          1.7         2.2        1.7        1.1
   Official grants1GDP                                                        1.4        2.5       1.5          1.1         1.7        1.6        1.5
   Total foreign financing (net) incl. resched./GDP                           4.0        4.5       2.0          1.7         2.1        2.3        2.0
   Trade balance                                                             -2.2       -0.5      -0.8         -1.4        -1.6       -1.7       -1.8
   Exports f.0.b.                                                            15.2       15.4      15.9         15.4        15.4      15.4        15.4
   Imports f.0.b.                                                          -17.4      -15.9      -16.7        -16.8       -17.0     -17.1      -17.2
   End-period gross official reserves                                      1,679      4,330    -9,486        9,312       9,773    10,989      12,089
       (inweeks ofimports ofgoods andnonfactor services)                      7.5       15.9      30.4         27.5        26.7      27.6       33.2

Sources: State Bank of Pakistan;Ministry of Finance; and World Bank staff estimates.

                                                                                                      Annex B
                                                                                                   Page 7 of 20

10. ..  Results:

        The trade deficit widens gradually but remains withinmoderate levels (in all years at less than the
        FYOO/Ollevel of2.2 percent o f GDP).
        The current account balance (including official transfers) returns to deficits - albeit modest - in
        FY03/04 after two years o f surpluses. The surpluses have been driven by the massive surge in
        private transfers that inour assessmentreflect a portfolio reallocationand hence are not o f a long-
        term nature. The current account balance excluding all transfers (official and private) shows a
        much less marked improvement in the last two years and over the medium term, only a slight
        increase inthe deficit i s projected. The current account deficits (including official transfers) are in
    ..  line with the needs o f a fast-growing low-income economy and fully sustainable as shown by the
        rapidly declining external debt burden.
        External public and publically guaranteed debt declines from 54 percent o f GDP in FYO1/02 to
        36 percent o f GDP inFY06/07.
        Gross official reserves continue to accumulate in nominal terms, reaching US$ 9.9 billion by
        FY06/07. They level out at about 5.8 months o f import cover (of goods and services) over the
        medium term.

111. LowCASESCENARIO

11.     The main alternative scenario assumes less stringent reform actions in particular with regard to
the fiscal budget, public enterprises and (other) structural reforms.

D.      Overall macroeconomic framework (Table A.4):

        Real GDP growth (at market prices) rises from 4.4 percent inFY 01/02 to 4.9 percent o f GDP in
    ..  FY06/07.
        Inflation rises gradually to 4.3 percent over the mediumterm, up from 2.7 percent inFY 01/02
        Gross national savings remain at the pre-FYOU02 level o f about 14 percent o f GDP compared to
        the base case reaching 17percent o f GDP by FY 06/07.
        Investment (gross fixed capital formation) remains at the pre-FYOU02 level of about 14 %
    .   percent of GDP; substantially lower than the base case reaching16 % percent o f GDP by FY
        06/07.

    .   The ICORs (Incremental Capital Output Ratios) are somewhat higher (averaging 3.3) than inthe
        base case (average 3.O) since productivity gains to investments are lower due to lagging structural
        reforms.
        Constant real effective exchange rate.

E.      Fiscal projections (Table A.5):

12.  .  Main assumptions:

        Fiscal revenue collection i s considerably lower than in the base case: the government proves less
        effective in tackling challenges in CBR and vested interests. Revenue collection in percent o f
        GDP stagnates at the FY 01/02 level averaging 16.6 percent o f GDP over the mediumterm. By
        FY 06/07 revenue collection is 1.1percent o f GDP lower than inthe base case. Inparticular, the
        improvement in CBR revenue collection seen over the last year ceases and CBR revenue stalls
        slightly below the FY03 level (inpercent o f GDP).
        There i s less foreign financing - o f both grants and (concessional) loans, reflecting less
        confidence by donors in the government's commitment to pursue reforms and address the goals

                                                                                                  Annex B
                                                                                               Page 8 of 20


    ..  for poverty reduction, social development and growth revival through appropriate economic
        policies.
        Improvementsineffectiveness o f government spending, governance and implementationcapacity
        are much slower than inthe base case.
        Total spending levels are only slightly higher than in the base case but the composition i s
    .   radically different with much higher interest payments, defense spending and payments o f
        contingent liabilities than inthe base case and consequently, much less fiscal space.
        Interest payments are much higher: at 4.6 percent o f GDP by FY06/07 they are 0.9 percent o f
        GDP higher than in the base case. This reflects a slower reduction in interest rates on domestic
        public debt because o f less strengthening in domestic confidence, and also the larger fiscal
        deficits and hence higher borrowingrequirements that leadto substantially larger public debt than
        inthe base case.
        Defense costs grow in line with nominal GDP instead o f in line with inflation only in the base
    .   case. They remain at 4 percent o f GDP, and by FY06/07 defense spending i s 1.1percent o f GDP
        higher than inthe base case.
        The government i s less successful in reducing public enterprise losses and other contingent
    .   liabilities. Payments o f contingent liabilities average 0.9 percent o f GDP inFY 02/03 - FY 06/07
        compared to 0.4 percent o f GDP inthe base case.
        The progress on privatizing public enterprises i s slow. Annual average sales receipts are half o f
        those inthe base case.

F.      Key results:

13.     There i s substantially less room for non-interest non-defense non-contingent liability spending
(stable at 11.9 percent o f GDP throughout the medium term) than inthe base case (averaging 13.2 percent
o f GDP and increasing to 14.5 percent by FY07). The structure o f spending remains almost unchanged
from FY03. The share o f non-interest non-defense non-contingent liability spending in total spending
declines slightly from 54 to 53 percent.

        Social and poverty-related spending stagnates at 3.8 percent o f GDP, below the FY03 level.
        Development spending i s cut over the medium term since most of the fiscal adjustment neededto
    .   avoid excessive fiscal deficits  takes place in this category. By FY06/07 development spending
        has fallen to 2.8 percent o f GDP; i.e., 1.9 percentagepoint lower than inthe base case.
        The lower spending on health and education as well as on infrastructure - both O&M and new
        investments (growth-related spending) - result in lower productivity o f public capital stock and
        human capital and slower economic growth than inthe highcase.
        Social indicators, poverty reduction and productivity o f human capital and public infrastructure
        improve less than in the base case since the effectiveness o f government spending improves less

    *   and the government finds less room for spending inthose areas.
        Total public debt i s higher and with less concessional terms on average (since there i s less
        external debt and more nonconcessional domestic debt); hence, the prospects o f reaching fiscal
        sustainability are considerably worse with much slower improvement in public debt indicators.
        Annual debt service in percent o f fiscal revenue rises from 38 percent in FY03 to 41 percent by
        FY07 as compared to 32 percent inthe base case. The debthevenue ratio drops from 602 percent
        inFY02 to 485 percent by FY07 as compared to 396percent inthe base case. Progress is hence
        slow inthe low case on reachingthe sustainability threshold o f 300 percent.

                                                                                                                        Annex B
                                                                                                                    Page 9 of 20

                Table A.4: Pakistan: Low Case Macroeconomic Framework, 2000/01-2006/07

                                                                               Est.Prov.Est.                 Proj.
                                                              2000101     2001102 2002103 2003104 2004105 2005106 2006107


Outputandprices                                                                   (annual changes inpercent)
   RealGDP at factor costs                                             2.2     3.4        5.1       5.0     4.9      4.9       4.9
  Real GDP at marketprices                                             2.6     2.8        5.8       5.0     4.9       4.9      4.9
   Consumer prices (p.a.)                                              4.4     2.7        3.1       4.1     4.1       4.2      4.2

Savingsandinvestment
   Grossnational savings                                              13.6    17.4       21.3      15.1    15.3      14.7      14.3
     Government                                                       -2.2    -0.9        1.6      -0.8    -1.0      -1.4      -1.8
     Nongovemment                                                     15.8    18.3       19.7      15.9    16.3      16.2      16.1

Gross capitalformation                                                15.5    14.7       15.5      15.7    15.7      15.9      16.0
     Government                                                        2.6     3.5        3.2       3.3     3.1       2.9       2.8
     Nongovemment 1/                                                  11.3      9.6       9.9      10.4    11.0      11.4      11.6
     Changeinstocks (nongovt)                                          1.6      1.6       2.4        2.0     1.6      1.6       1.6

 Savings- Investmentbalance                                           -1.9      2.7       5.9      -0.6    -0.4      -1.2      -1.7
     Government                                                       -4.8     -4.4      -1.6      -4.1    -4.1      -4.4      -4.6
     Nongovemment (incl.change instocks)                               2.9      7.0       7.5        3.5     3.7      3.2       2.9


Publicfinances                                                                         (in percent of GDP)
   Revenue(includinggrants)                                           17.3    19.3       20.8      18.1    18.3      18.0      17.7
   Revenue                                                            16.2    17.0       17.9      17.1    16.9      16.7      16.5
   Expenditure                                                        21.4    23.6       22.3      22.2    22.4      22.4      22.4
     Non-interest,non-defense,non-SOEspending                         10.2    10.9       12.0      11.4    11.5      11.4      11.3
   Overallbalance(including grants)                                   -4.1     -4.4      -1.6      -4.1     -4.1     -4.4      -4.6
   Primarybalance(including grants)                                    2.8      3.8       3.6        0.7     0.7      0.3      -0.1
   Gross public debt 21                                             108.3    100.1       91.8      83.6    81.5      78.9      77.9
   Publicdebt inpercent ofrevenue                                   670.5    588.9     512.0     488.8    482.5    473.0     473.2


Extemalsector                                                                          (inpercent of GDP)
   Merchandisetradebalance                                            -2.2     -0.5      -0.8      -1.4     -2.0     -2.5      -2.8
     Merchandiseexports                                               15.2     15.4      15.9      15.4    15.2      14.9      14.7
     Merchandiseimports                                               17.4     15.9      16.7      16.9    17.2      17.4      17.5
   Current account excl. official transfers                           -3.3      0.2       4.4      -1.6     -1.8     -2.5      -2.9
   Current account incl. official transfers                           -1.9      2.7        5.9     -0.6     -0.4     -1.2      -1.7

                                                                          (inpercent of exportsof goods and nonfactor services)

Extemalpublic andpublicly-guaranteeddebt 21                         289.4    282.6     229.6      208.1   196.4     186.7     179.0
     Debt service                                                     28.2    34.0       26.4      27.7    20.0      14.9      10.8
Gross reserves (inmillions of US.dollars)                           1,679    4,330      9,486     7,920   7,024     6,000     4,979
     inmonthsofimportsof goods andservices                             1.7      3.7        7.0       5.4     4.4      3.5       3.5

Memorandumitems:
   Real effectiveexchangerate(% change)                               -2.6     -1.2        6.0       3.7     0.0      0.0       0.0
  Realper capitaGDP (% change)                                         0.3      0.7        3.6       2.8     2.7      2.7       2.7
   GDP at marketprices (billions of PK rupees)                      3,423    3,629     4,018      4,392   4,796     5,242     5,730

Source:Dataprovided by the Pakistani authorities and World Bank staff.
1/ Includes state-ownedenterprises.
2/ Excludes SBP.

                                                                                                                               Annex B
                                                                                                                         Page I O of 20



                 Table A.5. Pakistan: Low Case: Consolidated GovernmentBudget, 2000/01-2006/07

                                                           2000/01     2001102   2002103       2003104      2004105    2005106    2006107
                                                                                Prov. Act.       Proj.       Proj.      Proj.      Proj.
                                                                                          (in.Dercent of GDPi
                                                                                          ,
Revenue and grants                                                17.3      19.3       20.8           f8.l        18.3       18.0       17.7
  Revenue                                                         16.2      17.0       17.9           17.1        16.9       16.7       16.5
     Tax revenue                                                  12.9      13.0       13.8           13.5        13.4       13.3       13.2
        Federal                                                   12.3      12.5       13.3           12.9        12.8       12.7       12.6
          CBR revenue                                             11.5      11.1       11.5           11.5        11.4       11.3       11.2
          Petroleumsurcharge                                       0.5      0.81        1.2            1.o         1.o        1.o        1.o
          Gas surcharge                                            0.4      0.49        0.5            0.3         0.3        0.3        0.3
        Provincial                                                 0.6       0.5        0.5            0.6         0.6        0.6        0.6
     Nontax revenue                                                3.3       4.0        4.1            3.6         3.5        3.4        3.3
  Grants                                                           1.2       2.3        2.8            1.o         1.4        1.4        1.3

Expenditure                                                      21.7       24.0       22.3           22.2        22.4       22.4       22.4
  Current expenditure                                             19.6      19.0       19.7           18.3        18.7       18.8       18.9
     Federal                                                      14.8      14.2       14.9           13.6        14.0       14.2       14.3
        Interestpayments                                           6.8       6.8        5.2            4.8         4.8        4.6        4.5
           Domestic                                                5.4       5.1        4.0            3.9         3.9        4.0        4.0
           Foreign                                                 1.5       1.7        1.2            0.9         0.8        0.7        0.5
        Defense                                                    3.8       4.1        4.0            4.1         4.4        4.7        5.0
        Runningof the civil govemment                              2.1       1.6        1.5             1.6        1.7        1.o
                                                                                                                              1.8        1.9
        Pensions for defense and civil govemment                   0.9       0.7        0.9            0.9         0.9                   1.o
        Subsidies                                                  0.6       0.5        1.2             1.5        1.3        1.2        1.1
            Power sector                                                                               0.9         0.8        0.7        0.6
            Other SOEs                                                                                 0.5         0.4        0.4        0.3
            Non SOEs                                                                                   0.2         0.1        0.1        0.1
        Grants                                                     0.5       0.6        0.6            0.7         0.7        0.7        0.7
     Provincial                                                    4.9       4.8        4.8            4.7         4.7        4.7        4.6
  Development expenditure and net lending                          2.1       3.5        2.7            3.5         3.3        3.1        3.0
     Public Sector Development Program                             2.6       3.5        3.2            3.3         3.1        2.9        2.8
     Net lending                                                  -0.5       0.0       -0.6            0.2         0.2        0.2        0.2

             Paymentof contingent liabilities                                1.4                       0.5         0.5        0.5        0.5

Statistical discrepancy                                           -0.3      -0.4        0.1            0.0         0.0        0.0        0.0

Budget balance (excluding grants)                                 -5.2      -6.6       -4.5           -5.1        -5.6       -5.7       -5.9
Budget balance (including grants)                                 -4.1      -4.4       -1.6           -4.1        -4.1       -4.4       -4.6

  Primary balance (including grants)                               2.5       2.0        3.6             0.7        0.7        0.3       -0.1

Financing                                                          4.1       4.4         1.7            3.4        4.1        4.4        4.6
  Extemal                                                          2.3       1.4       -0.2             1.o       -0.1       -0.3        0.0
  Domestic                                                         1.7       2.1         1.6            2.1        4.1        4.5        4.5
  Privatizationproceeds                                            0.0       0.2         0.3            0.2        0.1        0.1        0.1

Memorandum items:
  Primary balance (excluding grants)                               1.6       0.1         0.7           -0.3       -0.8       -1.1       -1.4
  Revenue balance (excluding grants)                              -3.5      -2.0       -1.8            -1.2       -1.8       -2.1       -2.5
  Spending excl. interest, defenses, subsidies, cont.
      liabilities, % GDP                                          10.2      10.9       12.0           11.4        11.5       11.4       11.3

      Share o f total spending (percent)                          47.4      46.0       53.5           51.2        51.1       50.9       50.5
  Nominal GDP (inmillions o f Pakistani rupees)                  3,423     3,629      4,018          4,392       4,796      5,242      5,730


Source: Nationalauthorities; and staffprojections and calculations

                                                                                                     Annex B
                                                                                                 Page II of 20

G.       Balanceof payments(Table A.6):

14.      Key assumptions:

          Stable real exchange rate
     m   Average annual export volume growth: 4 percent. Slower structural reforms result in slower
         productivity growth. Hence the competitiveness in export and import competing industries i s
         worse.
         Average annual import volume growth: 8.3 percent.
     m   Private transfers: as inthe base case it i s projected that workers' remittances inFY03/04 fall back
         to the FYO1/02 level (about $2.4 billion) - although slightly lower than inthe base case reflecting
         lower confidence -and then growing slowly inline with international inflation.
         Official grants are projected to decline from 2.4 percent o f GDP in FY 02/03 to 0.8 percent of
         GDP in FY 06/07, throughout the period lying 0.1        - 0.3 percent o f GDP lower than in the base
         case, reflecting that donors are less confident about the govemment's commitment to continued
         reforms.
         Lower program financing (adjustment loans on concessional terms) from the World Bank and the
         AsDB throughout the period (on average 0.5 percent of GDP per year comparedto 1.4percent in
         the base case) and consequently recourse to larger financing at commercial terms.

15.      Results.

         The trade deficit widens back to the level inFYOO-FYO1, averaging 2.3 percent o f GDP.
         The current account balance (including official transfers) returns to deficits, reaching 2.3 percent
          o f GDP by FY07 compared to 1.3 percent inthe base case.
          Given the limited financing at concessional terms the increasing current account deficits are
         partially financed by a drawdown o f international reserves. They drop from about US$8 billion
          in March 2003 (about 25 weeks of import cover) to about US$ 4.4 billion (about 13 weeks of
          import cover) by FY06/07.
          Extemal debt relative to exports o f goods and services declines from 309 percent in FY02 to 240
         percent inFY07 but remains still far above the sustainability threshold o f 150 percent.

16.      Figures 2.1 and 2.2 present the results o f a sensitivity analysis for the public debt burden and the
external debt burden, respectively, under the baseline scenario. The robustness o f the projected public
debt burden (as measured in percent o f fiscal revenue) and the external debt burden (measured against
exports of goods and nonfactor services) i s tested with regard to a wide variety o f exogenous shocks and
assumptions about the macroeconomic stance over the mediumterm.

17.       The main conclusion from the stress tests on the public debt burden (Figure 2.1 and Table A.7)
i s that almost none o f the shocks identified here lead to a worsening debt burden over the medium term.
Most scenarios show sustained reduction inthe debt-to-revenue and the debt-to-GDP ratios and remain on
a path toward debt sustainability within the 10-year horizon identified in the Fiscal Sustainability and
Debt Limitation Ordinance. However, one scenario-reversal                  o f macroeconomic policies and
environment to the average of 1992/93 -1998/99--stands            out as unsustainable. In the other cases o f
severe shocks the debt-to-revenue and debt-to-GDP ratios generally still declines although very slowly.
Two sets of stress tests have been run: one where we test the extent o f fiscal policy reversal that could
make the public debt situation unsustainable and one where we test for shocks to the growth rate o f real
GDP, the interest rate, and the GDP deflator within a magnitude o f 2 standard deviations o f the average
over the last 10 years.

                                                                                                                                       Annex B
                                                                                                                                 Page 12 of 20



                           TableA.6: Low Case: Pakistan: Balance of Payments,2001/01-2006/07
                                                                                            Proj.        Proj.        Proj.      Proj.      Proj.
                                                                      2000101     2001102   2002/03       2003104    2004105      2005106    2006107
                                                                                                           (in millions of USDollars)

Current account (excluding official transfers)                            -167          96     3,029       -1,210      -1,487       -2,159    -2,747
Current account balance (including official transfers)                        9      1,591    4,029           -450        -351      -1,020    -1,604
  Trade balance                                                            -92       -292      -536        -1,099      -1,609       -2,165    -2,640
     Exports f.0.b.                                                      2,351       9,140   10,889        11,704      12,364       12,999    13,676
     Imports f.0.b.                                                     -2,443      -9,432  -11,425       -12,803     -13,973      -15,164   -16,3 16
  Services (net)                                                          -675      -2,617    -2,172       -3,230      -3,028       -3,172    -3,337
     Ofwhich: interest payments                                           -326      -1,579    -1,272       -1,115      -1,050         -891       -782
  Private transfers (net)                                                  600       3,005     5,737         3,119      3,151        3,178     3,230
  Official transfers (net)                                                 176       1,495     1,000           760       1,136       1,139      1,142


Capital account                                                            440      -2,322    -1,529       -2,547         -871        -281        299
  Public medium- and long-term capital                                     272      -1,613    -1,497       -1,650         -624        -311        159
     Project and nonproject loans                                          249       -983     -1,696       -1,072         -549        -280          0
        Disbursements                                                      669         531       723           728         735         787        844
        Amortization "                                                    -420      -1,514    -2,419        -1,800       1,284      -1,068       -844
     Commercial banks and IDB (net)                                          16       -224      -158           -33         284         284        284
     Other                                                                    7       -407       357          -545        -359        -315       -125
  Public sector short-term (net)                                           -11      -1,064      -233          -564        -117        -100       -100
  Private medium- and long-term                                            152         -80       221           218         400         545        569
  Private short-term (including errors & omissions)                          27        435       -19          -551        -530        -415       -329

Overall balance (before debt relief)                                       449        -731     2,501        -2,997      -1,223      -1,300    -1,306

Financing                                                                 -449         731    -2,501         2,997       1,223       1,300      1,306
Reserve assets (increase -)                                               -883      -3,082    -5,267         1,392         475         711        691
Fundrepurchases                                                             -40       -194      -418          -526        -362        -135       -111
Net exceptional financing                                                  474       4,007     3,185         2,131       1,110         725        725
  Arrears (increase +)                                                       -8           0         0            0            0          0           0
  Rescheduling 31                                                          211       1,210     1,009           100         100         100        100
  Rollover o f foreign deposits with bankingsystem                         271       1,314       900         1,200         500         500        500
  Program financing from IFIs                                                 0      1,367     1,090           731         400           0           0
  Privatizationreceipts                                                       0        117       186           100         110         125        125


Memorandum items:                                                                                   (inpercent of GDP)
   Current account (excluding official transfers)                           0.0         2.7       5.9          -0.6        -0.4        -1.2       -1.7
   Current account (including official transfers)                          -0.3         0.2       4.4          -1.6        -1.8        -2.5       -2.9
   Interestpayments/GDP                                                                -2.7      -1.9          -1.5        -1.3        -1.o       -0.8
   Extemalprimary balance (incl. off. transfers)/GDP                                    5.4       7.7           0.9         0.9        -0.1       -0.9

                Total foreign financing (net)/GDP
   Official grants/GDP                                                                  2.5       1.5           1.o         1.4         1.3        1.2
   Total public- and publicly-guaranteed extemal debt                     40.9        58.0      48.9           40.9        37.6        34.9       32.7
      01wPublic- andpublicly-guaranteed extemal debt excl. SBP            38.1        52.7      45.8           38.1        35.0        32.5       30.4
  Total foreign financing (net) incl.resched./GDP                                       4.6       2.0           0.7         1.3         1.1        1.3
  Trade balance                                                            -0.2        -0.5      -0.8          -1.4        -2.0        -2.5       -2.8
  Exports f.0.b.                                                            4.0        15.4      15.9          15.4        15.2        14.9       14.7
  Imports f.0.b.                                                           -4.2      -15.9     -16.7          -16.9       -17.2       -17.4      -17.5
  End-period gross official reserves                                     1,679       4,330     9,486         7,920       7,024       6,000      4,979
      (inweeks ofimports o f goods and nonfactor services)                  7.5        15.9      30.3          23.4        19.2        15.2       13.7

Sources: State Bank o f Pakistan: Ministry o f Finance; and World Bank staff estimates.
1/ Assumes early debt retirement o f $1 billion in29003/04, and $500 millioneach in2004105 and 2005106.

                                                                                                                   Annex B
                                                                                                              Page 13 of 20




                         Figure 2.1: Sensitivity Analysis for Public DebffRevenues and Grants

                                                                   I




                 ----      A                                                     -     .c
                                                                                      ~

                           D                             B                      ---_
                                                                                -_--
                           G                             E                              F
                                                         H                              I
                 -Baseline        Projection



A = Ifinterest rate, real GDP growth rate, primary balance and non-debt flows (inpercent ofGDP) are at average ofpast 10 years
B= Ifinterest rate in2003104 and2004105 is averageplustwo standard deviations, others at baseline
C = Ifreal GDP growth rate in2003104 and2004105 is averageminustwo standarddeviations, others at baseline
D=Ifprimarybalance (inpercent of GDP) in2003104 and2004105 is average minustwo standard deviations, others at baseline
E= Combination of2-4 using one standarddeviation shocks
F= Annual additional contingent liability payments of5 percent ofGDP in2003104 and 2004/05, others at baseline
G= Same as A, but averages and standarddeviations calculated upto 1998199 only.
H= Same as D,but shock extending through 2006107
I=SameasE,butshocksextendingthrough2004105

Baseline Projection=Public debthevenues and grants

                                                                                                                  Annex B
                                                                                                              Page 14of 20




                 Figure 2.2: Sensitivity Analysisfor External Public Debt in percent of Exports
                                            of Goods and Nonfactor Services

     400




     350




     300




 $ 250




     200

                                                                            I
                                                                            I
                                                                            I
                                                                            I
                                                                            I
                                                                            I
     150                                                                    I
                                                                            I
                                                                            I
                                                                            I
                                                                            I
                                                                            I
                                                                            I
                                                                            I



     100
      1996/97     1997/98    1998/99    1999/00   2000/01    2001/02    2002/03    200304     2004/05    2005/06     2006/0


           -Baseline         Projection - .- .- A  -       0           C            D ..__E - -   _ _ _ _
                                                                                      ~                   F -         G


BaselineProjection= Ratio Extemalpublic debt to Exports of goods andNF services

A = Ifinterestrate, real GDP growth rate, US$ GDP deflator growth, non-interest current account, and non-debt flows (inpercent
    of GDP) over the mediumterm are at average ofpast 10 years
B=Ifinterestrate on extemal debt in2003104 and2004105 is IO-yearaverage plus two standard deviations, others at baseline
C = Ifreal GDP growth rate in2003/04 and2004/05 i s average minus two standard deviations, others at baseline
D=If GDP deflator in2003/04 is averageminustwo standarddeviations, others at baseline
      US$
E = Ifnon-interest current account (inpercent of GDP) in2003104 and 2004105 is averageminustwo standard deviations, others
    at baseline
F= Combinationof2-5 usingone standarddeviation shocks
G = 20 percent depreciationin2003104 and in2004/05, others at baseline

                                                                                                                   Annex B
                                                                                                              Page I 5 of 20

18.       Intest 1we look at the impact of a fiscal policy reversalback to the average primary balance (in
percent of GDP) over the last 10 years as well as a reversal to a similar macroeconomic environment with
the interest rate, real GDP growth rate, and non-debt flows (in percent o f GDP) all assumed equal to the
10-year average. (This period includes the last four years with reforms and progress in fiscal
consolidation and macroeconomic stabilization.) A policy reversal within this magnitude would not end
up inexplosive debt dynamics butwould rather be characterizedby severe stagnation: inthe debt burden,
economic growth and poverty reduction. The debt burden would decline only very slowly, economic
growth would be slow and progress on poverty reduction, social and growth-related spending would be
minimal.

19.       Inthe second stress test we check the sensitivity ofreturning to the policies and macroeconomic
ffamework inthe seven years preceding the initiation o f reforms, i.e., we use the averages up to 1998/99.
The conclusion i s clear: such a policy regime i s totally unsustainable and with substantially worse debt
indicators than intest 1.

20.       A third set o f shocks (3. - 7.) tests the debt burden sensitivity o f temporary shocks to the
following macroeconomic variables: interest rates (test 3), real GDP growth rate (test 4), primary balance
(test 5) and a combination o f tests 2. - 4 covering a set o f macroeconomic variables (test 6)70but with a
smaller shock (using averages plus 1standard deviation). Test 7 analyzes the impact o f subsidies to public
enterprises and fiscal payments o f contingent liabilities o f 5 percent o f GDP in addition to the baseline
amounts annually in 2003/04 and 2004/05. The general conclusion i s that temporary shocks o f these
magnitudes only slow down the process o f debt burden reduction but a sustainable debt position would
still be reached by 2011/12. Hence, the baseline scenario i s robust to temporary macroeconomic shocks.

21.       A fourth set o f shocks (8. - 9.) tests the debt burden sensitivity o f shocks persisting throughout
the projection horizon o f the same magnitude as above. The general conclusion i s that these permanent
shocks lead to stagnation or very slow decline inthe debt burden even over the whole 10-year horizon and
the debt positionremains unsustainable.

22.       The stress tests on the external debt burden (Figure 2.2 and Table A.8) show that a return to
historic macroeconomic perfonnance i s unsustainable. This emerges from three tests: Test 1 assumes
macroeconomic performance throughout the medium term to be the same as the average over the last 10
years. Test 6 and 7 assume a temporary reversal to historic macroeconomic performance in the first two
years o f the projection period but with an additional worsening o f key macroeconomic variables by 1-2
standard deviations. Such reversals will trigger a worsening o f the debt burden measuredboth interms o f
exports o f goods and nonfactor services (all three tests) and interms o f GDP (tests 6 and 7). Inparticular
the first test shows clearly an unsustainable path. These three tests once again underline the importance
for the Government to maintain the reform momentum, pursue sound macroeconomic policies and avoid
reverting to the macroeconomic policies o f the 19990s.




' OThe shocks are modeledas follows: inthe first two years - 2003104 and 2004105 -the projected variables are set equal to their
historic averages plus aworseningby two standard deviations. Subsequent years use the baseline scenario data.

9 9 9 9 9 9 9 9 9
2 2 2 2 2 2 2 2 2
I A I A L A I A I A ~ I A I A I A

0 9 0 9 0 9 0 9 0 9 0 9 0 9 0 9 0 9 0 9




                                             h2x
                                              b
                                              P
                                              8O
                                             Y



                                           -$b
                                           %:Ea -x
                                                 8h

                                                                                                  Annex B
                                                                                              Page 18 of 20



23.     However, the external debt sustainability o f the baseline scenario is fairly robust to most o f the
other stress tests. None o f the singular shocks identified here (separate tests for temporary shocks to
interest rates (Test 2), real GDP growth rate (Test 3), US$ GDP deflator (Test 4), and the non-interest
current account balance (Test 5)) leads to a worsening debt burden over the medium term. However, if all
these temporary shocks were to come at the same time (Test 6), the combined massive shock to the
economy would induce such a sharp, initial rise in the debt burden that it would still be substantially
heavier after five years. After the initial peak the debt burden would steadily be reduced, reflecting the
underlying strength o f the baseline scenario and would be on the path towards sustainability. However,
debt sustainability would only be reached beyond the 10-year horizon given the substantial size and
comprehensiveness o f the initial shock. An almost identical conclusion applies to the impact on debt
sustainability o f 20 percent depreciation ineach o f the first two years (Test 7).

       2N

       p!
       t-
       N




          .-*0
          2
I    ei

   .
   r-
   0
   u3
   0
   0
   N

   e
   u3

.=2g
   m
0  N
V
'Eom
4 B8
   0
   N

   e
   d

   W
   0
   0
   N

   W




         .-F
         w
          s
         c



         e
         Y

                                                                                                   Annex C
                                                                                                Page 1of 5

         SIMULATIONSBASED DYNAMIC ON             BENEFITINCIDENCEOFPUBLICEDUCATION
                                      EXPENDITURESPAKISTAN
                                                       IN


I.StaticBenefitIncidenceAnalysisanditsLimitation

1.      InPalustan, the social gaps (urban-rural, male-female, and rich-poor) in educational attainment
are prominent. According to the Static Benefit Incidence analysis, in order to narrow these gaps, public
expenditure for primary education i s particularly important, since such expenditure i s pro-poor, and
especially benefits the rural poor children (see Box 3.1). The Palustan governments have appropriately
allocated more public funds to primary education than other educational sectors, although the total
expenditure for primary education has recently stagnated.

2.      Although these results are useful, they may be debatable inthe following sense. The static benefit
incidence measures discussed above are static inthat they do not take into account the dynamic effects o f
spendingan additional dollar on the use o f public services. Inparticular, it is not clear whether spending
an additional dollar benefits the rural poor even though the static benefit incidence analysis suggests it
does.


11. The impact of expanding publicexpenditure for education: Dynamic benefit incidenceanalysis

3.      This section evaluates the incidence o f expanding public expenditure for education. The primary
objective i s to measure the impact o f expanding public expenditure for primary education on the gross
enrollment rate o f the poor, while controlling for expansion in private schooling.         To do so, it i s
appropriate to use the dynamic benefit incidence analysis proposed by Lanjouw and Ravallion (1999) (see
Box 1for a brief description o f the methodology).

4.      The analysis utilizes the variation among districts inthe gross enrollment rate (GER) o f different
income/consumption groups, to infer the impact o f expanding public expenditure on the GER o f each o f
these groups. To do so, one must estimate the marginal odds ratio of enrollment (Lanjouw & Ravallion),
which represents the change in the enrollment o f a given consumption group when there i s an increase in
aggregate enrollment. As long as the size o f aggregate enrollment i s proportional to that o f public
expenditure - an assumption that i s not perfect, but most "neutral" under the circumstances - one can
infer from the marginal odds ratio o f enrollment how an increase inpublic expenditure on education will
affect school enrollment o f each income group. In addition, in order to obtain the true relationship
between expanding public expenditure and enrollment o f the different groups, certain heterogeneous
characteristics deemed important - namely private school availability, literacy rate and rural/urban - are
controlled for.

5.      The first three columns o f Table 5 show the estimated marginal odds ratio o f primary enrollment
for different consumption groups, which can also be interpreted as the estimated impact o f 1 percent
increase o f public expenditure on the GER o f different consumption groups (under the above-stated
assumption that the size of aggregate enrollment i s proportional to that o f public expenditure).    These
results indicate that the lowest three consumption groups receive the largest benefit from public
expenditure expansion.    1 percent increase in total public expenditure increase enrollment rate by 0.9
percent for the bottom quintile, 1.1 percent for the second quintile, 0.9 percent for the third quintile and
0.6 percent for the fourth quintile. On the other hand, expanding public expenditure does not have any
significant impact on the public primary enrollment rate o f the richest group.

                                                                                                                   Annex C
                                                                                                                Page 2 of 5



                          Box 1: A Standard Methodology for Dynamic Benefit Incidence

Lanjouw and Ravallion(1999)71have proposed a method to calculate the marginal odds ratio o f enrollment based on
a single cross-sectional survey including data on school enrollment across districts within provinces. The procedure
is as follows. First, the average school enrollment rate for each quintile and each district i s calculated. Second, an
overall average school enrollment rate is computed for eachprovince. T o estimate the marginal odds o f enrollment
by expenditure quintile, one has to regress the quintile-specific enrollment rates across districts on the relevant
province's average enrollment rate separately for each quintile. Furthermore, to address the correlation between a
quintile-specific enrollment rate and an overall province-specific average rate, the leave-out mean i s used as an
instrument for the province-specific rate (see L&R, 1999for details).
There are several caveats. First, the size of aggregate enrollment i s interpreted as the size o f public expenditure if
the subsidy rate for educational service is constant not only across provinces and income groups,72 but also between
before and after expansion of public expenditure. But this may not be the case.             Second, it is assumed that all
factors other than province-level average enrollment rate are identical across provinces.73This assumption might be
too strong. T o partly address this problem, two province-level variables are included as explanatory variables inall
regressions: literacy rate and private school enrollment rate. Province average literacy rate i s included because
school enrollment is positively correlated with parents' education in Pakistan.74 The effect o f private school
enrollment rate i s includedbecause the impact o f expanding public expenditure might fade out as the private sector
becomes more accessible. Third,the gender gap and the urban-rural gap ineducation outcomes for Pakistan suggest
there are some structural differences between male and female as well as between the urban and the rural. T o
address the urban-rural gap, anurban dummy i s included as an explanatory variable inall regression^.^' To deal with
the gender gaps, the marginal odds ratio of school enrollment rate i s regressed separately for each gender in some
regressions. 76


6.        Although expanding public expenditure increases both male and female enrollment for the
poorest and the second poorest income groups, there i s a gender difference: the impact of public spending
on female enrollment appears to b e significantly          lower than that of male enrollment (Table 1). 1 percent
increase inpublic expenditure for primary education raises the gender gap inthe poorest income group by
0.3 percent, simply by impacting boys' enrollment more thangirls'.

7.        To some extent, the differential impact of public primary spending on boys and girls may b e
explained by the fact that parental decisions to send their children to school may b e affected by
sociocultural factors, over and above the availability of schools (which i s the factor that i s directly
affected by public spending), more inthe case of girls and for boys. Moreover, the decisions to send girls
to school may also depend on the availability of certain facilities in schools, like separate latrines and
female teachers, more than it i s the case for boys for obvious reasons - facilities that may b e quite
independent of the amount of public spending. For policymakers in Pakistan, this indicates an important
issue to address - namely, to explore how to enhance the impact of public spending on girls' enrollments.
To overcome instances of low incentives to send girl children to school due to sociocultural barriers,
targeted scholarships or subsidies can b e instituted - programs like these have been operational in
countries like Bangladesh andNepal with some success.



71Lanjouw, P andM.Ravallion(1999), "Benefit Incidence, Public SpendingReforms, andthe Timing o fProgramCapture," The
WorldBank Economic Review 13: 257-73.
72LanjouwandRavallion(1999) shows that the differences are relatively small
73Or all factors other than province-levelaverage enrollment rate vary in ways that they are uncorrelatedwith province-level
average enrollmentrate.
74Lanjouw et.al. (2001) also includes the literacy rate of province in the analysis for Indonesia's public expenditure reform.
Lanjouw,P., P. Menno, F Saadah, H.Sayed, andR. Sparrow(2001), "Poverty, Education, andHealthinIndonesia:Who Benefits
from Public Spending?"Policy Research WorkingPaper 2739, the World Bank: Washington, DC.
75 Furthermore, all province-level variables are computed separately for urban and rural districts within a province in all
regressions.
76Intheseregressions,all variables are computedseparatelyfor eachgender aswell.

                                                                                                                                    Annex C
                                                                                                                                  Page 3 of 5

8.       The first three columns o f Table 2 report how 1 percent increase in province average private
school enrollment rate affects district average public school enrollment rate. It i s worth noting that the
public primary school enrollment rate for the poorest income group significantly grows by 0.6 percent as
the private school enrollment rate grows by 1 percent. On the other hand, the public primary school
enrollment rates for the richest and the second richest groups decline by 1 percent and 0.3 percent,
respectively. Thus for the lower income groups, a larger private sector inprimary schooling boosts public
school enrollment; whereas, for higher income groups, the effect i s the opposite.

9.       Finally, after the availability o f private schooling and literacy rate has been controlled for, the
urban-rural gap does not seem to be significant.

10.      Loolung at the marginal odds ratio o f secondary school enrollment, the thirdpoorest consumption
group receives the largest benefit fkom expanding public expenditure for secondary education.                                           The
gender gap o f secondary education i s larger than that o f primary education. For the poorest group,
expanding public expenditure for secondary education does not increase female enrollment rate
significantly (Table 1). As Table 2 indicates, availability o f private schooling does not have any
significant impact on public secondary school enrollment.

                Table 1.MarginalOdds of Primaryand SecondarySchoolEnrollmentinPakistan
                                      Public Primary School                                Public Secondaiy School
         Quintile          Boys             Girls              Both              Boys               Girls             Both
         1(poorest)        1.052**          0.773'             0.925**           0.902**            0.055             0.747*
         2                 1.097**          1.092**             1.121**          0.848"             0.475*            0.727**
         3                 0.898'           0.907              0.912*             1.176**           0.886*             1.295**
         4                 0.156            0.773              0.597*            0.096              0.486             0.054
         5                 0.096            0.388              0.145             -0.152             0.676             0.154
       Note: n e table gives the instrumental variables estimates of the regression coefficients of the quintile-specific primary (and
       secondary) school enrollment rates across districts on the average rate by province. The leave-out mean enrollment rate is the
       instrumentfor the actual mean. The numbers in parentheses are t-ratios. + refers to significant at 10%; significant at 5%; *+
                                                                                                                 +
       significant at 1%.
       Source: Authors' calculations based on the 1998-99 PIHS.


             Table 2. The effectofprovinceaverage private schoolenrollment rate on district average
                                                   publicschoolenrollmentrate
       P                              Public Primav School                                 Public Secondary School
         Quintile          Boys             Girls              Both              Boys               Girls             Both
          1(poorest)       0.599'           0.698               0.613'           0.530              -0.365             0.535
         2                 0.410             1.039'             0.560             0.169             0.715              0.210
         3                 0.177             11.982             0.238             1.370             0.063              0.986
         4                 -0.971**         -0.167              -0.711*          0.681              -0.779             -0.810
         5                 -0.366           -0.675              -0.517'          -1.995             -0.689             -1.259
       Note: The table gives the instrumental variables estimates o f the regression coefficients o f the quintile-specific primary (and
       secondary) school enrollment rates across districts on the average rate by province. The leave-out mean enrollment rate_i s the
                                                                                                                                     .
       instrument for the actual mean. The numbers inparentheses are t-ratios. + refers to significant at 10%;   * significant at 5%; **
       significant at 1%.
       Source: Authors' calculations based on the 1998-99 PIHS.


111. Simulationsto RelateTargetswithPublicExpenditures

11.      Using the analysis presented above, this section addresses the following question: what are the
rough and preliminary estimates o f public expenditure required to achieve a target GER (fixed at 80
percent o f gross primary enrollment rates) for the bottom two consumption expenditure quintiles by the
fiscal year 2007/08? An important assumption underlying this exercise is that the impact o f overall
public expenditures on education, on the enrollment o f different incomehonsumption groups will

                                                                                                    Annex C
                                                                                                 Page 4 of 5

continue to be the same - over time and as public expenditures increase - as that revealed by the data o f
1998-99 (and estimated inthe previous section).

12.      Table 3 illustrates the results from this exercise. Usingthe coefficients from Tables 1and 2, it i s
estimated that overall public spending on primary education should be at least Rs.42.2 billion (at 1999/00
prices) to achieve 80 percent o f primary GER for the poorest quintile in 2007/08 FY, while Rs. 49.3
billion (at 1999/00 prices) i s neededto achieve the same for girls inthe poorest quintile in2007/08 FY. It
should be noted that these' spending figures will be enough to ensure GERs of 80 percent or above for
corresponding groups among higher consumption quintiles.

13.      Before examining the numbers in Table 3 in detail, it i s worth describing in some detail the
restrictive assumptions that underlie these calculations. First, as mentioned in the previous section, the
subsidy rate for primary education is assumed to be constant in real term, which implies that public
expenditure i s proportional to a province level primary enrollment rate. Second, the structure estimated
above does not change from 1998/99 (when PIHS data was gathered) to 2007/2008. Inother words, the
impact o f expanding public expenditure (or private enrollment rate) in 2007/08 i s assumed to be the same
as the number reportedinTable 1(or Table 2). Ifthe estimate o f coefficient i s not significant, it i s treated
as zero. Third, the private enrollment rate of a particular quintile i s assumed to grow from 1998/99 to
2007/08 at the same rate as between 1990/91 and 1998/99. Under this assumption, the expected private
enrollment rate o f each quintile in 2007/08 i s computed. Fourth, the annual growth rate o f school-age
population i s assumed to be 1.025, which i s the average annual growth rate o f population between 1991
and 2001. Given these assumptions, this exercise should be considered indicative, which does not take
into account structural shifts, like a change in the demand conditions for schooling or a major change in
the pattern of private schooling - any of which factors can lead to very different scenarios in terms o f
what the public sector has to achieve.

14.      Under these assumptions, Table 3 shows that Pakistan needs to increase the overall public
expenditure for primary education by 73 percent in real terms (or Rs. 17.9 billion at 1999/00 prices) by
the fiscal year 2007/08, in order to achieve 80 percent o f primary school enrollment rate for the poorest
quintile. This implies that Palustan's government will need to spend 0.96 percent o f GDP for primary
education (assuming an annual growth rate o f GDP o f 4.5 percent) in 2007/08, while they spent 0.8
percent of GDP for primary education in 1998/99. If the share o f primary education in total public
expenditure on education i s fixed at the current level (47 percent), Palustan's government will need to
increase total education expenditure from 1.68 percent o f GDP in 1998/99 to 2 percent o f GDP by
2007/08. A similar GER target for the second poorest quintile will necessitate a 39 percent increase in
public primary education spending. Combining these two groups          - i.e., in order to achieve a primary
GER o f 80 percent for the poorest 40 percent of the population - public primary education spending
would need to be 55 percent higher inreal terms than the current level, amounting to 0.86 percent o f GDP
in 2007/08. If the share of primary education in total public expenditure on education is fixed at the
1998-99 level, totalpublic expenditure on education has to increase to 1.8 percent o f GDP by 2007/08.

15.      The gender gap remains serious inthese projectionstoo. T o achieve the same goal for girls inthe
poorest income group, Pakistan's governments need to double the overall public expenditure on primary
education in 2007/08 - amounting to 1.1 percent o f real GDP. Achieving the same goal for girls in the
bottom 2 quintiles combined- the poorest 40 percent o f the population-will require a spending increase
o f 85 percent inreal terms by 2007/08, which will amount to 1.02 percent o f GDP. Thus, if the share of
primary education intotal education expenditure i s fixed at the 1998-99 level, totalpublic expenditure on
education has to increase from 1.68 percent o f GDP in 1998/99 to 2.1 percent o f GDP in2007/08.

16.      Achieving the same targets for boys would require far less. Inparticular, Pakistan's governments
will achieve the goal for boys in the bottom 40 percent o f the population by increasing real public

                                                                                                                  Annex C
                                                                                                                Page 5 of 5

spending on primary education by only 23 percent, an allocation that will amount to 0.68 percent o f the
country's GDP, which i s less than the current share o fprimary education in GDP (0.8 percent of GDP).


     Table 3. PublicExpenditurefor PrimaryEducationin2007/08 to achieve 80 YOprimaryGER for a
                                                       givenquintile

                                                                         Public ExDenditure
                                                  Primary           Primary
                                     Ratioa       education        Education     All Education         All Education
                                                  1998/9gb          2007108'         9819gd              2007/08e


    Thepoorest quintile     All       1.73      24.3 (0.80%)     42.2 (0.96%)     50.8 (1.68%)          87.9 (2.0%)
                            Girls    2.03       24.3 (0.80%)     49.3 (1.12%) 50.8 (1.68%)              103.1(2.3%)
    The secondpoorest All             1.39      24.3 (0.80%)     33.8 (0.77%)     50.8 (1.68%)          70.6 (1.6%)
          quintile          Girls     1.68      24.3 (0.80%)     40.9 (0.93%) 50.8 (1.68%)              85.3 (1.9%)

      The bottomtwo         All       1.55      24.3 (0.80%)     37.6 (0.86%)     50.8 (1.68%)          78.7 (1.8%)
    quintiles combined BOYS           1.23      24.3 (0.80%)     29.8 (0.68%) 50.8 (1.68%)              62.5 (1.4%)
                            Girls     1.85      24.3 (0.80%)     45.0 (1.02%)     50.8 (1.68%)          94.0 (2.1%)
  Note: All public expenditure figures are billionRs at 1999/00prices. Parenthesesrefer to the share ofrealGDP (at
        1999/00prices) for the correspondingyear.
  a. Ratiorefers to the ratio ofthe public expenditurefor primaryeducation in 1998/99FY to that in2007108FY.
  b. Public Expenditurefor primary educationin 1998/99.
  c. Public Expenditurefor primaryeducationin2007/08 to achieve 80 YOgross enrollment rate for agivengroup.
  d. Public Expenditurefor all educationin 1998/99.
  e. Public Expenditurefor all educationin 2007108 to achieve 80 % primary GER for aparticulargroup assumingthe share
        o f primary educationintotal educationalexpenditure i s the same as in 1998-99.


17.     This analysis reveals that the real challenge ofPalustan's governments is how to increase primary
GER for children (especially girls) o f the poorest income group. Achieving such an increase in such a
short time i s likely to be a tall order for any government, especially given the trends of social spending in
Palustan inrecent years evident fiom graphs presentedinthe previous section.

18.     The analysis done so far to gauge the impact o f public spending on education outcomes like
enrollment assumes that the current conditions - rather, conditions prevailing in 1998-99 - will continue
to prevail. Given that public spending has been highly inefficient inPalustan till date, which has stymied
improvements in outcomes, it i s thus natural that the current projections show that very large - and
therefore very hard to achieve - increases in public expenditure would be needed to reach the targets in
outcomes. On the other hand, it can be argued that the same targets can be reached with less increase in
the amount o f public spending, ifthe eficiency o f public spending were to increase -particularly interms
of targeting such spending to the lower income groups.


STATISTICAL ANNEXES


               G
               s
               Lo
               d     m
                     0




'2 2
0 -

2o %
   w   .-v1
w .o   .-E
      e

.-
 2
.-S
E
.=6    d
       .-04
P      'cl
       .-
       m'

4


            i,
            i
           (II
           4        .-0
                    m

            .-a0    5
            D
            .-      (/I


       m    2
.-m
.r
5
F

4


I


4

4




4M
P
r


5M
P
E


-001 N



      O o 0 l


0000 N0 0 0




o o tN
     -0 0 0
      m
      -*"
o o Nw0 0 0




o o lN
     n0 0 0




N o mN0 0 0




m o tN
     -m o o
      N
      t-




oor- NE O 0
      m



m o NmZ O O

                 4

N O N*-
      w o o
      m


m o z E O 0
      d



o o w-

0 0 2


0 0 2



O O NN


o o w
  w 0  zocr&o
  m 0"     t
     d      In
            N


       - 0 g o
            2

o m o -m O N 0
  t-
  2         m
       0"
       - 1 00-



0 -at-
    N  N O LN O
              O

             N.
             vi


       N z w -
             o
             ""
             r-

       m o c o
             z
       W N rr -- 0
         w
             ""
             2

       -m omo 0
             B

o m m
  N b
  tmt
  -


O tN- O
      N
  9 r-"
  - N




o o w
  m w a


0010 O N 0
    10    lo




- 0 - o m 0
N         r-
          N




0 0 0   0 - 0
          d
          ""
          N


0 0 0 o m 0
          m
          o\




0 0 0 m m o
          lo




0 0 0   N o r -
          N
          lo10




0 0 0 O N t -
        r-10




o o rm-



0 0 2
     d




O O Nr-
     m



o o mW
     ';f



o o z
     m



o o m
     10




O O Nw
     N




0 0 %

     c)

m o


N
010




m o




b 0
0


-
m o




Z 0



d N
N




m o
m

-
t - 0




2 2


m m
N m



-N mW

N 0 0 - b ~ m 0 m 0 g ~ O m 0 0 0 0 0 0 0 0 0
  r-  N m - m
      m m      - w -        ""  r - m
                                w t-               I


m 0 0 t - m ~ m m ~ O r - " 0 m w 0 0 0 0 0 0 0 0
  b   m i 3 - Z ; "
      m w               c o rN,
                             -  G %                I

- N O m w N " b 0 ~ w 0 m ~ 0 0 0 0 0 o o m
  b   N b t - b r - N
      m r - - m         z ?     z w                I
                            3




b m - N?"In"" " w 0 w
        o o                  m 0 0 N 0 0 0 0 - -
P'N      N. w N                                   s
                                                  b



m m O m m 0 b b r - 0 0 r - 0 b m 0 0 0 0 -
r-    m am-- r - r -                            O P
             m m        m m
                        m t--    o m
                                 --               -
                                    10




                                 dY
                                 VI
                                 h       in

                                         0

0 0 0 0 0 0




0 0 0 0 0 0




0 0 0 0 0 0




0 0 0 0 0 0




0 0 0 0 0 0




0 0 - 0 0 0




0 0 - 0 0 0




0 0 2 0 0 0
    00




m o 0
    m 0 0 -

    0"      f
00002 0 0 - G     4
            8
            'c

0 0 0




0 0 0




0 0 -




0 0 -




0 0 -

4
P
::


r


4



3e


U



m - v i m v i - o o o o b m o t - c
m - t - 3 "       w        m o o
                              0" vi"
O"
3            N                * i o




tm - Om~ * m3.m ON m 0 0 0 m 0 0 C
         m o w    m
N        v.                v ivr0 2



         N   -




           .   .
          3  3




           .   .
          3  -




              c:
              i
              cri
             4          m
                        0

3
U

4




&M
P
E

        2 :
        V I V I O O O O ~
                     N




        0 0 0 0 0 0 ~ 0 0 0
                         m
        m o
        ~m               N



        m t - 0 0 N O m 0 0 0
        2 s           - r -



        d o o o - o y ?   0 0 0
        mO N
        N  O              O    I



        d N O O O O m d O O
        2 2               -   I
                          N




        ~m
        r - m 0 0 0 0 W r - 0 0
        N                 m
                          N




I I"

           4




           GM
           t
           P




m
c
k;    m
P     E


.-.-
2
E
VI

-c
b       VI
0       e

                                                                        L

                                                                        M
0 m ~ 0 N Q l d NmN w 0 0 ~ 0 ~ - w r - 0 tw- o Nmo o m o o o ooo md    2:
  o m  m
  m -      m * *N O
                           -".e
                             ems.         L o t -
                                              m   m  m
                                                     0      T I N "0

                                                                  0 0 0 0 0 0 d r-
                                                                                d




                                                                  0 0 - 0 0 0 4 1
                                                                                N
                                                                                N




                                                                  0 0 0 0 0 0 - iz


                                                                  O O O O O O W




                                                           o m rm
                                                             w -
                                                             2"*  0 0 - o o o m




                                                                  o o - o o w ~




                                                           O a m
                                                             Z"
                                                             O N  0 0 ~ 0 0 0 -
                                                                r-              m
                                                                                r-




                                                                  0 0 - o o o m m



                                                                  o o w o o o -m      4
                                                                                W




                                                           o o m
                                                             moo
                                                             Z" * o o ~ o - o m P


$ 1                                                               O O N O O O O
                                                                      .;r       r-
                                                                                m



     r - N o * o " w wNw mwDNo N o ~ o r - - 0 0 0 m o       z -
                                                             m m      m   md r  N-
4 m2.      r - m d         m   m        2 0 ' ;   m m r - 0o m - o o d o m o -
                                                         5

                                                                  o o m o m o -
                                                                      w    v l m
                                                                           m    m



                                                                  o o m o w o o
                                                                      Z    G    8


                     .-hl
                     .-                                      c:

                      c
                      e                                      2