Economic Analysis of Policy-based Operations: Key Dimensions · Economic Analysis of Policy-Based Operations: Key Dimensions 1 O VERVIEW macroeconomic policy framework. A macro-meso-micro

Economic Analysisof Policy-BasedOperations:Key Dimensions

Richard Bolt, Manabu Fujimura, Cindy Houser,Franklin De Guzman, Frederick Nixson, John Weiss

ii

Copyright Asian Development Bank 2004. All rights reserved.

This publication was published in June 2003 and edited in November 2004.

The views expressed in this book are those of the authors and do not necessarily reflect the views andpolicies of the Asian Development Bank (ADB), or its Board of Governors, or the governments theyrepresent.

ADB does not guarantee the accuracy of the data included in this publication and accepts noresponsibility for any consequences of their use.

Use of the term “country” does not imply any judgment by the authors or ADB as to the legal or otherstatus of any territorial entity.

ISBN 971-561-491-4

iiiEconomic Analysis of Policy-Based Operations: Key Dimensions

PREFACEThis volume is the result of an Asian Development Bank (ADB) Economics and ResearchDepartment (ERD) study: Good Practice Analysis in Support of ADB’s Policy-Based Lending. Itaddresses various aspects of analysis and preparation of policy operations. Issues coveredinclude understanding the macroeconomic context of sector policy reform operationsusing a macro-meso-micro perspective, sector analysis with emphasis on price policy andthe institutional context, analytical approaches for assessing the impact of policy changeincluding poverty impact assessment and political economy considerations of reformcosts, and design considerations for ADB policy-based loans.

Working papers prepared by the Economic Analysis and Operations Support Division(EREA) are frequently cross-referenced: Policy-Based Lending and Poverty Reduction: An Over-view of Processes, Assessment, and Options (Bolt and Fujimura 2002) and Toward a PoliticalEconomy Approach to Policy-Based Lending (Abonyi 2002). Both are part of the ERD workingpaper series. It is hoped that the works, combined, will contribute toward improving ADB’soperations through better practice in policy analysis.

This volume was prepared by Richard Bolt, Manabu Fujimura, Cindy Houser,Franklin De Guzman, John Weiss, and Frederick Nixson, with support from Xianbin Yao(Assistant Chief Economist, EREA, ERD) and David Edwards (Director, OperationsEvaluation Division 2). The following ADB staff reviewed and provided valuable com-ments on the draft paper: Douglas Porter, Stephen Curry, David Green, ChristopherEdmonds, Peter Choynowski, and Chia-Hsin Hu. The paper is intended primarily for useby ADB staff but may be of interest to others engaged in development assistance.

The opinions and views expressed in the paper are those of the authors and not of ADB.

PREFACE

vEconomic Analysis of Policy-Based Operations: Key Dimensions

CONTENTSPreface .......................................................................................................................................... iii

Contents .........................................................................................................................................v

Abbreviations .............................................................................................................................. viii

Glossary ........................................................................................................................................ ix

OVERVIEW ................................................................................................................................... 1

CHAPTER 1Introduction ........................................................................................................................... 5

CHAPTER 2The Macroeconomic Context of Sector Policy .............................................................. 9A. Introduction..................................................................................................................10B. Macro-Meso-Micro Framework ................................................................................10C. Importance of the Macroeconomic Context ...........................................................12D. Scope of the Macroeconomic Assessment ...............................................................14

1. Economic Growth Performance .........................................................................142. Macroeconomic Management Performance ......................................................153. Structural Policies .................................................................................................164. Outlook for Economic Performance .................................................................165. Macroeconomic Framework and Linkages ........................................................16

CHAPTER 3Dimensions of Sector Policy Analysis ...........................................................................17A. Introduction..................................................................................................................18B. Sector Analysis Concerns ...........................................................................................18C. Linkages Between Policy Reforms and Development ............................................19D. Scope of Sector Analysis ............................................................................................20E. Price Policy Context ....................................................................................................23F. Institutional Context ..................................................................................................25

1. Institution Building and Reforms .....................................................................252. Understanding Functions and Performance of Market-Related

Institutions ...........................................................................................................283. Managerial and Administrative Capacity of Institutions ...............................314. Technical Capacity of Institutions ....................................................................31

G. Reform Timing and Sequencing Issues ...................................................................311. Economic Considerations ...................................................................................312. Political Economy Dimensions ..........................................................................33

CONTENTS

vi

CONTENTS

CHAPTER 4Assessing the Effects of Policy Change ............................................................. 39A. Introduction..................................................................................................................40B. Assessing the Effects of Policy Change ...................................................................41C. Approaches to Analysis ...............................................................................................42

1. Descriptive Statistical Analysis (Cases 2 and 3 in Appendix 3) .................452. Partial Equilibrium Analysis: Market and Price Analysis

(Cases 1 and 4 in Appendix 3) ..........................................................................453. Partial Equilibrium Analysis: Comparative Institutional Analysis

(Case 4 in Appendix 3) .......................................................................................474. Applied General Equilibrium Modeling (Case 5 in Appendix 3) ................47

CHAPTER 5Assessing the Poverty Impact of Policy Change ................................................ 51A. Introduction..................................................................................................................52B. Mechanisms for Poverty Reduction .........................................................................52C. Current Poverty Impact Assessment Practice at ADB ...........................................54D. Review of Current ADB Practice ...............................................................................55E. Modifying the PIA Matrix ..........................................................................................57

1. Key Refinements ...................................................................................................572. Channels of Effect ................................................................................................573. Timing and Other Considerations ....................................................................59

F. Operational Considerations .......................................................................................60

CHAPTER 6Assessing the Processes and Costs of Policy Change .................................................63A. Introduction..................................................................................................................64B. Understanding Reform Costs ...................................................................................64

1. Time Dimensions..................................................................................................642. Distribution Implications of Reforms ..............................................................663. Political Economy Considerations ....................................................................67

C. Fiscal Environment for Policy Reforms ..................................................................68

CHAPTER 7Incorporating Policy Change Assessments into Program Design ............................71A. Introduction..................................................................................................................72B. Integrating Policy Analysis and Operation Design ...............................................72C. Conditionsy as a Guide to Implementation ............................................................74D. Linking the Program Framework and Reform Monitoring ..................................75

TABLESTable 1: Links between Macro-Meso-Micro Levels .......................................................11Table 2: Factors Causing Economic Prices to Exceed Financial Prices ....................24

viiEconomic Analysis of Policy-Based Operations: Key Dimensions

CONTENTS

Table 3: Factors Causing Financial Prices to Exceed Economic Prices ....................24Table 4: Institutional Issues Considered in ADB Program Loans ............................29Table 5: Economic Timing and Sequencing Issues in Program Loan Reports .......35Table 6: Selection of Techniques for Policy Change Impact Analysis ......................44Table 7: Approaches to Assessment of Policy Operations ..........................................50Table 8: ADB’s Current Poverty Impact Assessment Matrix .......................................55Table 9: Modified Poverty Impact Assessment Matrix ................................................58

FIGURESFigure 1: Effects of Programs and Projects on Production Possibility ...................20Figure 2: Stages in ADB Operations Giving Rise to Policy Issues ...........................21Figure 3: Different Reform Scenarios to Meet a Policy Goal ....................................41Figure 4: Alternative Real GDP Scenarios .....................................................................43Figure 5: Alternative Rural Poverty Scenarios ...............................................................43Figure 6: Relationship between Policy Reform and the Medium-Term

Expenditure Framework ...................................................................................69Figure 7: Integration of the PIA Matrix, Program Logical Framework,

and Policy Matrix ..............................................................................................73Figure 8: How Reforms Confer Benefits: Programs Logical Framework, and

Economic Analysis Links .................................................................................75

BOXESBox 1: Policy-Related Economic and Institutional Problems Identifiable in

Sector Analysis .......................................................................................................22Box 2: Examples of Conflicting Policy Alternatives in the Agriculture Sector ........23Box 3: Examples of Price Policy Trade-Offs in the Agriculture Sector ....................23Box 4: Factors to Consider in Assessing Managerial and Administrative

Absorptive Capacity ...............................................................................................32Box 5: Factors to Consider in Assessing Technical Absorptive Capacity .................33Box 6: Examples of Sequencing Considerations in Financial Sector Reforms .......34Box 7: Factors to Consider in Assessing Capacity to Adapt and Willingness

and Incentive to Change .......................................................................................37

APPENDIXES AND CASE STUDIESAppendix 1: Program Loan Reports Reviewed..............................................................77Appendix 2: ADB Experience on the Macroeconomic Context Section

of Program Loan Reports ...........................................................................79Appendix 3: Case Illustrations for Application of Analytical Techniques ...............85Appendix 4: Adjustment Costs in ADB Policy Operations—An Inventory

of Approaches Used .................................................................................. 133

BIBLIOGRAPHY AND REFERENCES ............................................................ 149

viii

ABBREVIATIONSADB Asian Development BankADTA advisory technical assistanceAIC average incremental costBAS Bureau of Agricultural StatisticsCGE computable general equilibriumDMC developing member countryERD Economics and Research DepartmentEREA Economic Analysis and Operations Support DivisionERIMU education reform implementation monitoring unitESDP Education Sector Development ProgramESW economic and sector workFDI foreign direct investmentGSDP Grains Sector Development Program (Philippines)IFI international financial institutionIMF International Monetary FundkWh kilowatt-hourM&E monitoring and evaluationMTFF medium-term fiscal frameworkMWh megawatt-hourNFA National Food AuthorityNPC National Power CorporationO&M operation and maintenanceOED Operations Evaluation DepartmentPAM policy analysis matrixPBL policy-based lendingPIA poverty impact assessmentPPSRP Philippines Power Sector Restructuring ProgramREC rural electrification cooperativeRRP report and recommendation of the PresidentSAM social accounting matrixSDP sector development programSME small- and medium-sized enterpriseSOE state-owned enterpriseSSE senior secondary educationT&D transmission and distributionTA technical assistance

In this report, “$” refers to US dollars, unless otherwise stated.

ABBREVIATIONS

ixEconomic Analysis of Policy-Based Operations: Key Dimensions

GLOSSARY

Advisory Technical Assistance (ADTA).Advisory support that may include training,which is usually extended in a sector- oreconomy-wide context for institutionbuilding. It usually assists a country inestablishing and strengthening institutions;in preparing national and/or sectordevelopment plans and programs; and incarrying out sector, policy, and issues-oriented studies.

Economic Internal Rate of Return(EIRR). Rate of return achieved on allproject resource costs measured ineconomic prices; for a project to beacceptable, the EIRR should be greater thanthe economic opportunity cost of capital.

Economic and Sector Work (ESW). Astand-alone effort that deals with broad andspecific economic, sector, and thematic,issues. Economic work covers mainlymacroeconomic issues, country strategystudies, or special studies that cut acrosssectors. Sector work covers issues that aretechnical, financial, economic, institutional,social, or managerial in nature. ESW is setin a medium- to long-term context and aimsto contribute to: (i) deeper understandingof economic and sector issues, and (ii) thefashioning of appropriate strategies andinstruments to strengthen economicmanagement and sector capacities. ESW isalso referred to as sector analysis work.

Financial Internal Rate of Return (FIRR).Rate of return achieved on all project costs,where all costs are measured in financialprices and benefits represent the financial

GLOSSARYrevenues that would accrue to the mainproject participants.

Program (or Project) Performance AuditReport (PPAR). Study that evaluates theeffectiveness of the program (or project)in achieving its intended objectives. Itincludes an analytical commentary and anaudit of the adequacy and integrity of theprevious program (project) completionreport. It focuses on specific issues meritingclose attention and analyzes the causes ofany deviations.

Project Preparatory Technical Assistance(PPTA). Aid to carry out feasibility studiesor detailed engineering studies for aproposed project, or to develop a pipelineof projects suitable for financing, or toconduct a sector review to identify sectorissues to be addressed by a project orprogram, or a master plan.

Report and Recommendation of thePresident (RRP). Program or project loanproposal submitted to the ADB Board ofDirectors and providing adequateinformation for it to make a decision onthe proposal.

Technical Assistance (TA). Financial aidto facilitate the flow and efficientutilization of development finance todeveloping member countries and toenhance their development capacity. It isalso used to foster regional cooperationthrough assistance in the preparation ofregional studies and conduct of seminars,conferences, etc.

1Economic Analysis of Policy-Based Operations: Key Dimensions

OV

ERV

IEW

macroeconomic policy framework. A macro-meso-micro perspective can help incomprehending the forward and feedbackeffects of sector policy changes.

An appreciation of the macroeconomiccontext and the mesoeconomic channelsthrough which sector policy changes arecarried out helps identify key influences onsector policy. The macroeconomic context,such as growth, inflation, unemployment,and the balance of payments, and keymacroeconomic variables, such as interestrates, money supply, and foreign exchangerates can greatly influence the outcome ofsector policy reforms. For example,vulnerability to external change and crisescan trigger system-wide policy changes andaffect the outcomes of structural and sectorreforms. Similarly, sector reforms can havefeedback effects on the macroeconomy.

Although macroeconomic assessment iscountry specific, a consistent approach toassessment is needed to understand:

• economic growth performance;• macroeconomic policy and management;• key structural policies;• economic outlook; and• macroeconomic framework including

risks or vulnerabilities.

Policy analysts need to carry out a country-specific assessment and analyze howmacroeconomic influences, workingthrough markets and institutions at theeconomy’s mesoeconomic level, affectdecisions and incentives by consumers andproducers at the microeconomic level.

OVERVIEW

Introduction

Policy-based lending to support policyreform is a necessary developmentinstrument. Support for sector policyreform is the main focus of policy-basedoperations of the Asian Development Bank(ADB), addressing economy- or sector-wide causes of structural constraints andunder-performance.

Two key characteristics of policy changearise. First, policy changes alter theunderlying framework influencingstakeholders’ behavior. Their response tochanges must be taken into account inpolicy analysis. Second, removal of a policy-related constraint gives rise to anotherlimitation that becomes the next “bindingconstraint.” In this sense, policy reform isa process of removing sequential bindingconstraints.

The features of policy reform needcomprehensive analysis. Policy change isalso a dynamic process, and to be effectivebuilt-in flexibility in design andimplementation is needed involvingstakeholders. The evolving nature ofreforms requires that ex-ante analysis besupplemented with monitoring andevaluation of the policy change process.

The Macroeconomic Context of SectorPolicy

Policy analysts need to understand themacroeconomic context for policy-basedoperations to ensure the consistency ofsector policy changes within the

OVERVIEW

2

Dimensions of Sector Policy Analysis

Sector analysis underpins ADB’s policyoperations and is needed to understandpolicy, institutions, and investment issuesthat affect sector performance. Sectoranalysis helps to assess:

• sector role and features;• markets, prices, and the incentives

structure;• government strategy and policies,

reform agenda• key public institutions, performance,

problems and underlying causes;• market and institutional performance,

problems and underlying causes;• solutions, reform and investment

needs;• the rationale for public sector and ADB

involvement, and alternativeapproaches;

• timing and sequencing of reforms anddevelopment activity; and

• political economy considerations.

Analysis of market structure, conduct andperformance including price policy issuesare important, given their impact oneconomic and financial incentives and onthe behavior of affected stakeholders, publicfinances, and trade.

Institutions include public serviceproviders, market institutions, and the“rules of the game.” Analysis of such issueswill help to identify constraints facingconsumers and users, producers andproviders, how they affect incentives andresponse, and opportunities forimprovement.

Analysis of institutional structure, conductand performance provides insights into

OVERVIEW

factors that affect policy implementationand impact. This includes understandingthe implementation capacity of public andprivate institutions to effect policy reform.Reforms involving institutional capacitybuilding usually require a medium- to long-term view, and the time and resourcerequirements should be factored into theanalysis.

The political economy context affectspolicy formulation, timing, reformacceptance, and implementation. Fullawareness is needed of the politicalenvironment, a country’s policy priorities,trade-offs involved, and the stance ofaffected stakeholders. This includesidentifying the views of those who will gainfrom the reforms, and of the vested interestswho will lose from reforms and who preferto maintain the status quo.

Assessing the Effects of Policy Change

Policy change can be analyzed ex ante indifferent ways with various levels of rigor.Analysis may be limited by practicalconsiderations such as data availability, costof collection, and human resource capacity.

Key considerations and analytical optionsare as follows:

• The starting point is to understand the“without program” situation. Sectorstudies can provide much of theframework and data, and can potentiallyemploy a range of macro, meso, and microanalyses, depending on the type of policychanges and strength of likely feedbackeffects.

• Advanced assessments of policy optionsand the effects of the “with program”


OV

ERV

IEW

situation based on deductive reasoningmay be checked by empirical verification.Where data are limited and collectionimpractical, valid and verifiableassumptions need to be clearly presentedto allow monitoring and evaluationduring implementation.

• Policy analyses and simulations, wherepractical, can identify short- andmedium-term effects of policy change.General equilibrium analysis is useful ifthere are significant feedback effects thatneed to be identified and assessed. Partialequilibrium analysis is useful wherefeedback effects are low. Whenequilibrium analysis is not possible,descriptive statistics and provisionalcalculations should be used tounderstand sector parameters andpossible responses.

Assessing the Poverty Impact of PolicyChange

An analytical framework used by ADB forthe poverty impact of policy change, thepoverty impact assessment matrix, identifiesthe main channels of effect through whichthe poor are affected, assesses if the impactis direct or indirect, and outlines thedistribution implications between poor andother affected groups. The poverty impactassessment framework is best used as adesign tool as part of analytical work.

Assessing the Processes and Costs of PolicyChange

Understanding is needed on the processesof cost and benefit realization of policychange as a complement to quantitativeanalysis of its effects. This will helpestablish whether reforms are feasible from

OVERVIEW

the stakeholders’ viewpoint, including howthey will be affected, whether they canadjust to change, and how they will adjust.

Assessing the policy change processinvolves four key dimensions of adjustment:

• Time: Policy change involvesrestructuring that often occurs over themedium to longer term. Policy changeeffects need to be considered in dynamic,not static, terms.

• Distribution: The distribution effects ofreforms need systematic assessment toidentify who the gainers and losers willbe as the adjustment process unfolds.

• Political economy: Given the time anddistribution effects, the groups affectedcan be identified and their viewscanvassed on reform support—oropposition and why. The strength ofvested interests compared with that ofreformers should be assessed. The needfor possible compensation for groupsnegatively affected during the adjustmentprocess can be assessed and built intothe program’s design and costs.

• Development implications: Often,reform measures will have direct linkagesto development efforts, such asinstitutional reforms and related capacity-building needs, and complementaryinvestments. The link between reformsand complementary development effortsshould be established to increase thelikelihood of sustaining the reform effort.

The government will have to shoulder anyfiscal impact of reforms. Reforms shouldbe considered in the context of theprevailing fiscal situation and their impact

4

on the future framework through analysisof national or local medium-term fiscalframeworks.

Incorporating Policy Change Assessmentsinto Program Design

The final, essential step for policy analystsinvolves incorporating the results ofanalysis into program design.

The analysis helps inform decision makersand affected stakeholders of options andtheir possible outcomes. A basic requirementof successful reform implementation is thatthese groups reach a shared understandingof the nature, costs, and dimensions ofpolicy changes. Summary matrixes used inADB loan documents help highlight keyanalytical and design points.

OVERVIEW

The policy matrix, forexample, can beconsidered as a road map for governmentcommitment on specific policy actions toensure that the reform implementation ison the right track. The emphasis of thepolicy measures in the matrix should be onmilestones and key actions. Clear linkagesbetween the poverty impact assessmentframework, program logical framework, andpolicy matrix help ensure consistencybetween the analysis and design.

Monitoring and evaluation during and afterreform implementation should supplementpractical policy analysis as part of programdesign. This will also help in overcomingthe inherent risks and uncertainties of up-front analysis and in putting policy reformsinto effect.

CHAPTER

INTRODUCTION

6

CHAPTER 1

The current approach of the AsianDevelopment Bank (ADB) to programlending1 was introduced in 1987. Theapproach adopted signified a shift fromthe financing of imported inputs aimed atincreasing capacity utilization to anemphasis on sector policy reforms tofacilitate growth and development. Overthe period 1987–2002, ADB supported136 programs loans, accounting for 21%of the cumulative total of all loanapprovals. Throughout, ADB carried outperiodic reviews of program lendingperformance.2 The reviews show thatpolicy-based lending complements projectfinancing and is a necessary developmentinstrument. However, analyses in supportof ADB policy-based operations indicatescope for further improvement in theirrelevance and effectiveness.

In 2001, ADB’s Operations EvaluationDepartment (OED) completed a specialevaluation study of program lending,identifying several key aspects of programdesign and implementation needed toachieve successful outcomes (ADB 2001).These include: first, the need for sufficient

sector diagnosis to understand thecontextual concerns of policy changes;second, improved ex-ante assessment oflikely impacts of policy reforms andrelated analysis of the adjustment andrelated costs; and third, the requirementfor institutional capacity building, before,during, and after the implementationphase of policy reform.

Policy changes relate to market andinstitutional reforms and address theeconomy- or sector-wide causes ofstructural constraints and under-performance. This has two key implications.First, since these reforms alter theunderlying framework that governs thebehavior of different stakeholders,stakeholders’ responses to policy changesmust be taken into account in policyanalysis. Second, once a policy-relatedconstraint is removed, another limitationbecomes the next “binding constraint.” Inthis sense, policy reform involves a processof removing sequential bindingconstraints. So, the various features ofpolicy reform present a challenge anddemand careful and comprehensiveanalysis.

Understanding of the nature anddynamics of policy change andinstitutional reform has evolved over time.The literature indicates the different stagesin this evolving process in terms of the“generations” of reforms (Brinkerhoff andCrosby 2002). Broadly, first generationreforms include macroeconomicstabilization and parametric policy changessuch as price liberalization. Structuralpolicy and underlying market andinstitutional changes are the secondgeneration. The approach to policy-basedlending is also changing. Issues such as civil

1 Program lending incorporates ADB’s presentlending modalities of program loans, sectordevelopment program loans, and cluster programloans. The term “program lending” refers to ADB’ssupport for developing member countries to carryout a program of policy reforms and institutionalchanges. However, program lending modalitieshave evolved to include projects with policy andknowledge product components that often addresspolicy issues. This volume uses the term “policyoperations” and “policy-based operations” forpolicy-related lending and nonlending activities.

2 ADB. WP 2-90: A Review of the Bank’s ProgramLending Policy, March 1990; R143-96: Review ofthe Bank’s Program Lending Policies, July 1996;and R210-99: Review of ADB’s Program LendingPolicies, November 1999.


INTRODUCTION

CH

AP

TER

1

3 The background analysis to the paper draws on avariety of materials: (i) an in-depth examination of10 program loan reports or reports andrecommendations of the President (RRPs) (seeAppendix 1); (ii) examination of the approaches toadjustment cost estimation from 30 other RRPs;(iii) the special evaluation report by OED (ADB 2001);and (iv) ADB policy reviews on program lending. Italso draws on discussions with ADB staff andconsultants. In addition, reference is made to therelevant work from other international agencies,especially the International Monetary Fund (IMF)and the World Bank.

participation and country ownership arebecoming essential elements of policy-basedoperations. This evolving understandingreinforces the need for more systematicpolicy and institutional analysis.

Recognition that policy reforms ofteninvolve a dynamic process indicates theneed for a flexible built-in mechanism indesign and implementation, withinvolvement of affected stakeholders. Italso highlights the fact that the principlesand good practices of sustainable policyreform are still evolving and have a stronglearning content. So, from the policyanalysis standpoint, the ex-ante analysisneeds supplementing with effectivemonitoring and evaluation during andafter the policy changes.

The present study provides a systematicexposition of key aspects of the economicanalysis of policy operations forpractitioners. It continues the operations-related research of ADB’s Economics andResearch Department (ERD) to addressthe above policy analysis concerns as theyrelate to operations (see Bolt and Fujimura2002, Abonyi 2002). The studyemphasizes the relevance and feasibility ofpolicy changes (Abonyi 2002). Relevance,in this context, refers to theresponsiveness of policy changes tospecific characteristics of developmentissues in a country or sector setting.Feasibility refers to designing a policychange to be consistent with the specificcontext and prevailing circumstances,including its likelihood of contributing toimproved country performance. It drawson policy operation practices within ADBand elsewhere,3 and contains case studiesthat help illustrate ways to conduct policyanalysis. The focus is on issues common

to all sectors, rather than on individualsectors.

Chapter 2 begins by providing a macro-meso-micro perspective and emphasizes theimportance of understanding themacroeconomic context for policy-basedoperations. This is to ensure theconsistency of sector policy changes withinthe overall macroeconomic policyframework. Chapter 3 addresses the sectordiagnosis that underpins policy operationsand stresses the comprehensive sectordiagnosis on issues relating to policy,institutions, and investment as the basis forpreparing program lending. The linkagebetween reform and development isdiscussed. Key aspects of market andinstitutional reforms are elaborated upon.Issues of timing and sequencing of policychanges and the political economy ofreforms are discussed.

Chapter 4 outlines analytical approachesand tools for assessing policy reform andhighlights how they can be appropriatelyapplied with regard to issues such asfeedback effects, data availability, andhuman resource capacity. Chapter 5 focuseson the poverty impact and distributionimplications of policy changes, anddiscusses current refinements in ADB

8

CHAPTER 1

practice. Chapter 6 brings intertemporal anddistribution aspects together to helpunderstand the costs of reform andpolitical economy considerations.

Chapter 7 addresses the implications ofanalytical dimensions for program loandesign, including the need to buildinternal coherence between ex-ante impactanalyses, the program logical framework,and the program policy matrix; theimportance of monitoring and evaluation;and issues in loan release conditions.

The study emphasizes practical andsystematic ex-ante analysis of policy reformin terms of context, process, and outcomes,rather than the implementation of policyoperations. It also recognizes the existenceof methodological and data limitations forsuch ex-ante analysis. This recognitionstems from the uncertainty prevailing in thepolicy reform process. The range ofanalyses presented should be carried outnot just in the context of program loanprocessing, but also as part of ADB’s regularcountry and sector policy and institutionstudies, country strategies, countryeconomic reports, and other economic andsector work.

CHAPTER

THEMACROECONOMICCONTEXT OFSECTOR POLICY

l Macro-Meso-Micro Framework

l Importance of the Macroeconomic Context

l Scope of the Macroeconomic Assessment

10

CHAPTER 2

A. Introduction

Policy reforms supported by ADB, withthe exceptions of “crisis” loans, operateprimarily at the mesoeconomic level ofthe economy. This is where the incentivestructures faced by economic agents atthe microeconomic level are determinedby market mechanisms and relatedinstitutions. The functioning of marketsis governed by an intricate web ofinstitutional arrangements that supporteconomic transactions. Understandingthe interrelationship between the macro-,meso-, and microeconomic levels isimportant for three reasons: First,macroeconomic policy decisions—especially those made in the face ofexogenous shocks and that are channeledthrough the mesoeconomic level tohouseholds and firms—can alter thecosts and benefits of a reform program.Second, macroeconomic crisis is oftentriggered, or at the least, exacerbated bymeso- and micro-level problems thathighlight the need for sector reforms.Third, structural and micro-level reformscan have macroeconomic feedback effects,such as the effect on relative prices andunderperforming institutions, which canaffect the costs and benefits of thereform program.

This chapter briefly sets out a macro-meso-micro framework for organizingthe discussion, reviews the importanceof the macroeconomic context in policyanalysis, and identifies the scope ofmacroeconomic assessment.

B. Macro-Meso-Micro Framework4

A useful framework for visualizing sectorpolicy work is one that shows how the

mesoeconomic level of the economyprovides a link between the macro- andmicroeconomic levels. A stylized pictureof the process is illustrated in Table 1.This framework shows that theparticular characteristics of aneconomy’s performance are determinedat the following levels:• macroeconomic, through the fiscal,

monetary, and exchange rate policiesthat are relatively flexible tools for short-run aggregate demand management;

• mesoeconomic, through theinstitutions where policy measuresmay entail a longer gestation period,and through more easily alterablemarket regulations, taxes, andsubsidies; and

• microeconomic, which are onlyindirectly influenced by the government,especially in market economies.

Analysis of the macroeconomic leveldeals with the highest level ofaggregation among markets, includinggoods, services, assets, and labor. Theinteractions determine the equilibriumlevels of several aggregate variables,including price level, real output, realinterest rate, real wage rate, employment,and exchange rate.

At the other extreme, the mostdisaggregated analysis focuses on themicroeconomic level of economic agents.At this level, the household is thefundamental agent that, in maximizingits own welfare, incurs costs of andgarners benefits from given economicactivity. Households decide betweenwork and leisure, savings and

4 The framework outlined here is a modified versionof that presented in Demery et al. (1993),pp. 4–8.


THE MACROECONOMIC CONTEXT OF SECTOR POLICY

CH

AP

TER 2

Economic growthInflationUnemploymentExchange rate Relative prices

Resource and incomedistributionMarket structure and conduct

Household welfareWork/leisureFirm’s profitsExpectationsInvestment decision

Mostaggregation

Leastaggregation

Focus

Individual/household- Utility function

Firm - Production function - Profit function

Sector Markets:Goods & Services• Agriculture• Industry• Private and public servicesAssets• Financial assets• Land and property• Durable goodsLabor• Unskilled• Skilled

Level ofAggrega-tion

TABLE 1: Links between Macro-Meso-Micro Levels

Household and businessresource situation andconstraints

Broad-based taxes andtransfers

Economy-wide resourcesituation and constraints

Fiscal, monetary, andexchange rate policies

Governance and standards

Customs, rules, and behavioralnorms

Enabling legal and regulatoryenvironment, rules

Institutions, services, andinfrastructure

Sector-specific and targetedtaxes and subsidies

Technology, comparativeadvantage

ParametersandPolicies

1. Reduce fiscaldeficit throughtax increases

Example ofFeedbackEffects

3. Reduced consumptionof thesecommodities

2. Higher commodityprices

6. Changes inlevel ofaggregateoutput

5. Lower levels ofsector outputsdelivered to themarket

4. Cutback in supply byfirms in face ofdecreased demand

MicroeconomicMesoeconomicMacroeconomic

Aggregate demand(consumption plusinvestment plus netexports)

Aggregate supply(function of resourcesand technology)

Source: Macroeconomics and Finance Research Division/EREA staff, Asian Development Bank.

consumption, and respond to changesin prices and expectations in accordancewith budget constraints. For their part,firms are essentially composed ofindividuals who produce goods and

services. Key decisions made by firmsdepend largely on changes in prices andexpectations, including investmentdecisions on physical capital andtechnology.

12

CHAPTER 2

Between these two extremes lies themesoeconomic level where prices aredetermined and agents exchange goods,services, assets, and labor. The level ofaggregation can be thought of as flexiblealong this spectrum. By successivelydisaggregating markets according tocharacteristic, region, and time horizon,one moves down along the spectrumtoward a progressively more micro-economic level of analysis.

An economy is not a static phenomenonbut a constant flow of activitiescontinually buffeted by exogenousevents. Many events are too small to havea measurable impact on the widereconomy and can be studied within apartial equilibrium framework. However,systemic changes can be thought of asoccurring at the macroeconomic level ofanalysis, resulting in changes inaggregate economic activity. These thenfeed back through the whole economy,requiring consideration through a moregeneral equilibrium framework.

Conversely, sector- and meso-levelreforms can have feedback effects fromthe micro level to the macro level,through market mechanisms andinstitutions. Table 1 shows an attemptto achieve fiscal balance through taxincreases that may be passed on toconsumers in the form of highercommodity prices. Price signals, in turn,may reduce consumption of thesecommodities. A decrease in demand foraffected goods may force firms to cutback supply, which then feeds back tothe macro economy through lower levelsof sector outputs delivered to themarket. The extent of the changes inaggregate output depends on how

effectively these changes are transmittedor filtered out by market andinstitutional factors at the meso-economic level. Analysis of theeffectiveness of such mesoeconomicfilters helps in understanding the macroand micro impacts of reforms, andrequires close study of themesoeconomic conditions thatdetermine how macro-level policies aretranslated into micro-level incentives(Zezza and Llambi 2001).

C. Importance of the MacroeconomicContext

Since most of ADB’s policy operationstake place at the mesoeconomic level, thepossibility of major changes inmacroeconomic conditions that can alteran operation’s benefits and costs shouldbe considered. For example, a change inthe strength and character of medium-term economic growth or in keymacroeconomic prices such as theexchange rate can have implications forthe viability of a project that supportsexport growth, or for the valuation of astate-owned enterprise (SOE) to beprivatized under a policy operation.Furthermore, as international trade ingoods, services, assets, and labor assumesmore importance in the domesticeconomy, economic interdependence cancreate new opportunities for economicgrowth as well as new challenges.Exogenous shocks, or the need to adjustto them, may impose higher coststhrough, for example, rising rates ofinterest or higher costs of importedinputs and domestic prices for reasonssuch as currency devaluation or increasingcommodity prices. Such instability mayincrease uncertainty, choking off



CH

AP

TER 2

investment and slowing the reformprocess. For example, a fall in the rate ofgrowth of world trade may adverselyimpact on export prospects with negativefeedbacks for agriculture sector reforms.Overall, various outcomes are possible,which underscores the need to under-stand the macroeconomic context andhow it affects the meso and micro levels.

While changes in macroeconomicconditions are difficult to predict, therisks to the policy operation should beassessed if there is evidence of anunsustainable trend, particularly inmacroeconomic policy, or if theeconomy is thought to be vulnerable toexternal shocks. So, it is necessary tounderstand past macroeconomic trendsand to be able to realistically assessfuture prospects. Recent examples inADB operations (Appendix 1), wherethe macroeconomic context has matteredgreatly to the outcome of reformprograms, include the following:• rapid monetary expansion triggered

high inflation that eroded balancesheets of commercial banks,subsequently undermining financialsector reforms (LAO: SecondFinancial Sector Program);

• stabilization and fiscal restraintmeasures reduced expenditures andreforms in the social sector, that, inturn, hampered transition to a newpension system (KAZ: PensionReform Program);

• the regional financial crisis reducedeconomic growth and financial inflows,delaying the implementation of SOEreforms and increasing SOE welfarelosses (VIE: State-Owned EnterpriseReform and Corporate GovernanceProgram); and

• the global economic slowdown in2000–2001 reduced the pace offinancial and corporate sector reformsin Asia’s crisis-affected economieswhile increasing nonperforming loans(INO: Financial Governance Reforms).

Other considerations emphasize theimportance of understanding themacroeconomic context for sector-levelpolicy operations. For example, a severemacroeconomic crisis or major changesin the political system can underscorethe need for structural changes at themesoeconomic level. Where several majormacroeconomic and structural policychanges are under way, it can be difficultto predict the economy’s response to afurther policy change. Assessment of thecosts and benefits of a given policymeasure in order to gauge theappropriate shape or sequence of policychanges would be needed. Here, it isimportant to consider the degree towhich the policy regime, and thestructural characteristics of theeconomy, are in a state of change, interms of both speed and direction.

Even for a straightforward investmentproject, cost-benefit analysis can bedifficult if the project is undertakenduring a period of great economic stressor rapid structural transformation. Theanalytical challenge can be more difficultstill in policy operations aimed atsystemic changes that will alter aggregateeconomic activity. These may alter themesoeconomic structure and, in turn,affect the macroeconomic phenomena thatarise from exogenous shocks, with theresult that observed market behavior andthe path of economic development maychange. The economic analysis of such

14

CHAPTER 2

policy operations demands analytical rigorbecause the time horizon may be long(especially for benefits). Also, the analystmust identify and quantify the linkagesthrough which these effects occur. Forexample, relative price changes can occurthat trigger dynamic shifts in resourceallocation, spreading costs and benefitsbeyond the originating sector and affectingaggregate levels of prices and output.

Similarly, institutional reforms canimpact on, for example, revenuecollection arrangements and the contentand approach to expenditures. Theeffects of decentralization are a case inpoint. Depending on the nature and sizeof the impact, this can change the overallfiscal balance, with possible implicationsfor the macroeconomy.

So, an understanding of the macro-economic context is essential forsector-level policy operations. Thishelps in designing the appropriatepackage of policy changes, byidentifying the channels through whichthe impacts of changes arise, and byassessing the sensitivity of the cost-benefit analysis to these changes.

D.Scope of the MacroeconomicAssessment

To the extent possible, the identificationof major macro-meso-micro economiclinkages should be done as a part of themacroeconomic assessment.

Aspects that need to be covered indiscussing the macroeconomic contextinclude the following “checklist” items:5• review of the economic growth

performance, including key events ortrends in the markets for goods,

services, assets, and labor;• assessment of the macroeconomic

management performance, includingfiscal, monetary, and exchange rateindicators;

• review of the major structural policiesor characteristics governing marketbehavior, with coverage tailored to theparticular policy reform context, butincluding trade, financial, investment,commercial, and labor policies;

• provision of an outlook for economicperformance over the relevant programperiod; and

• assessment of the overall macro-economic framework, includingexplicit identification of macro-economic factors, linkages, andassumptions.

Specific issues and questions withinthese aspects are as follows.

1. Economic Growth Performance

• What is the country’s level ofeconomic development?

• What is the character of recenteconomic performance (real GDPgrowth)?

• Is growth steady, concentrated, orvariable across sectors?

• What are the sources and uses ofsavings?Answers to these questions canprovide insights into the strengthof an economy. For example, thecharacter of investment can give anindication of the health of the

5 Appendix 2 provides some key lessons as to howdiscussion of the macroeconomic context and itsintegration into the analysis of a policy operationcan be improved, as well as providing the basicmacroeconomic indicators used to support theassessment.



CH

AP

TER 2

private sector. If a vibrant privatesector is important to a policyoperation, the analysis should takea closer look at the sector’sperformance. The analysis shouldalso compare investment as a shareof GDP relative to other countrieswith similar levels of per capita GDPand look at its breakdown into publicand private investment. Additionally,the analysis should look at theproportion of investment thatdomestic savings finance and theextent and magnitude of foreigninvestment. The analysis should alsocheck whether investment isconcentrated in one sector ordiversified.

2. Macroeconomic ManagementPerformance

• Does the country enjoy macro-economic stability?Such stability is always aprerequisite for successful reformprograms. For example, high andincreasing inflation is usually awarning of instability, and is oftenpreceded by a rapidly depreciatingexchange rate and a rapidly growingmoney supply. Indications ofinstability suggest that a closer lookat the health of the financial systemis warranted.

• How likely is a future episode ofmacroeconomic instability?The presence of an independent andcredible monetary authority, with aclear monetary goal and a monetarytarget appropriate to that goal,reduces the risk of future instability.Further, a weak or underdeveloped

banking sector may limit the abilityof the central bank in economicpolicy making.

• Is government finance sustainable,or are large deficits crowding outprivate investment, causing inflationand building up public debt?Large deficits may trigger fiscalausterity measures, unless they arecovered by official developmentassistance. If fiscal targets areambitious or there is extensive off-budget spending, then it isimportant to know what changes infiscal policy, such as cuts inexpenditures or additionalborrowing, will be made.

• How does government revenuecompare, as a share of GDP, to thatin other countries with similar percapita incomes?Very low levels of revenue, oftendominated by trade taxes, mayindicate that, with tradeliberalization, structural changes inthe taxation system will changebusiness conditions and relativeprices.

• Are there any indications ofinstability in the balance ofpayments?A high current account deficit isoften indicative of growth. However,a deficit that is rapidly rising as ashare of GDP and is financed byshort-term capital inflows may be asign of future risks.

• What is the exchange rate regime?Although a fixed exchange rate isprone to becoming misaligned, afloating exchange rate is oftenvolatile, leading to pressures forcentral bank intervention.

16

CHAPTER 2

3. Structural Policies

• What are the major elements of tradepolicy?Factors to consider include thecountry’s affiliation to internationaltrade bodies and its commitment totradeand capital account liberalization.

• What are the major elements ofinvestment policy?The rules governing foreigninvestment can affect theattractiveness of foreign investmentrelative to domestic investment.

• Are there other structural policiesor characteristics relevant to thepolicy operation?For example, labor law and thefunctioning of the labor market areoften important aspects to considerwhere investment climate andcompetitiveness are the focus ofreform.

4. Outlook for EconomicPerformance

• What is the development outlookfor the country?This should include a discussion ofmedium-term prospects. Such adiscussion can help inform progresswith the reform agenda, includingprojections for indicators that willbe used for monitoring reformprogram performance.

5. Macroeconomic Frameworkand Linkages

• What macroeconomic feedbackmechanisms, if any, will need to be

considered in undertaking theprogram impact analysis?These include the microeconomicelements most likely to be affectedby the macroeconomic frameworkthrough identified channels ofeffect, and vice versa. Additionallinkages may have to be identified,and after an initial review, mayrequire an adjustment to the scope,sequencing, or thrust of theproposed set of policy reforms.

• What mesoeconomic transmissionmechanisms and channels of effectare instrumental in supportingmacroeconomic to microeconomiceffects, and vice versa?

• What are the key assumptions aboutthe macroeconomic context andstructural characteristics of theeconomy that underpin the policyoperation?

• What types of systemic risks ormacroeconomic vulnerabilities willneed to be investigated in thecontext of a sensitivity analysis?

The above checklist is the minimumnecessary assessment for each policyoperation. Ideally, the review should becarried out by a country economist.Similarly, coordination with otherdonors that routinely carry out countryand sector reviews will increase theunderstanding of critical economiclinkages. Where significant feedbackeffects and mesoeconomic filters areidentified, the resources and capacity fordeveloping independent judgment mustbe provided to the program preparationteam.

CHAPTER

DIMENSIONSOF SECTOR POLICYANALYSIS

l Sector Analysis Concerns

l Linkages Between Policy Reforms andDevelopment

l Scope of Sector Analysis

l Price Policy Context

l Institutional Context

l Reform Timing and Sequencing Issues

18

CHAPTER 3

A. Introduction

The need for comprehensive sectoranalysis has long been recognized inADB. This chapter looks at issues relatedto the sector, policy, and institutionalenvironment in the context of policy-based operations. The linkages betweenpolicy reforms and development are alsoreviewed. Two fundamental sector issuesare then discussed—price policy andinstitutional arrangements. The chapterconcludes with an overview of reformtiming and sequencing issues, includingan introduction to the politicaleconomy dimensions of reform. Theselast points pave the way for placing thepolicy diagnosis in an analyticalframework, to be discussed in thesucceeding chapter.

B. Sector Analysis Concerns

The analytical framework for sectorprogram assistance in ADB’s 1987review of program lending policies calledfor, among other things, assessments ofissues such as: sector-specific policyconstraints and their effects onproducer incentives; domestic andborder input and output prices; thedegree of sector reliance on marketforces; sector adaptability to changingcircumstances; and the adequacy ofinstitutional support and improvementneeds (ADB 1987). Written in the mid-1980s, the review was primarilyconcerned with aligning priceparameters, fostering marketcompetition, and ascertaining thedynamic environment. ADB’s policycontinues to emphasize the medium-term development orientation of policychanges, rather than short-term macro-

stabilization measures. The reviewprovided for a sector program focus—policy framework, investment program,and underlying institutional readiness—that closely relates to the policy–investment–institution trilogy of thesector-lending modality. It also set outthe basic analytical issues involved inadjustment, further discussed inChapter 6.

A decade later, the 1996 review of ADB’sprogram lending policies (ADB 1996)stressed the need for programs to bebased on a comprehensive sector analysisand policy dialogue covering sectorinvestment plans, institutionaldevelopment needs, and social andenvironmental aspects. Moreover, someobservations from the special evaluationcarried out by the Operations EvaluationDepartment (OED) (ADB 2001)underscored the need for a thoroughand comprehensive sector analysis.

Recognizing the importance of theinterconnectedness of policy changes,program loans are generally consideredas “slices of reform” implemented in adynamic environment. As indicated byCollier (2001), for example, oncemacroeconomic misalignments havebeen addressed, growth is still usuallyconstrained by many other factors. Thischaracteristic emphasizes the importanceboth of understanding clearly thatreforms are related to the rest of theeconomy or sector, and of identifyingthe most important binding constraints,which, in turn, implies appropriateselection of reform priorities. It alsoraises two questions about the programdesign: First, are the proposed policyand institutional reforms representative


DIMENSIONS OF SECTOR POLICY ANALYSIS

CH

AP

TER 3

of the binding constraints, given thecircumstances? Second, what alternativepolicy and institutional reform measuresare being considered that take intoaccount the interlinkage of the subjectreform area with the rest of the policyand institutional structures of aneconomy? Effectively addressing thesequestions reinforces the long-standingemphasis on a careful and comprehensivesector analysis.

C. Linkages between Policy Reforms andDevelopment

Reforms are ultimately concerned withmore efficient and/or equitable use ofresources, goods, and services. Bothpolicy and institutional reforms havedevelopmental, welfare-enhancementcharacteristics, which may involve thedismantling and reforming of existingunderperforming institutions while, atthe same time, developing, throughinvestment-type activity, newinstitutional arrangements for futuredevelopment. Ali (1990) makes thepoint that it is difficult to understand,or even have, good investment projectsin a bad policy and institutionalenvironment. The persistence of such asituation may reduce real returns toprojects, implying that a high rate ofreturn to policy reforms may begenerated by correcting distortions tothe policy environment. Ali suggests thatinvestment projects and policy reformsare closely linked and that practitionerscan employ similar analyticalapproaches.6 Kanbur (1990) notes thatmany investment projects includeconditions on policy changes in thesector being invested in (usuallyincluded as “assurances” in ADB project

documents). Appropriate project-policycombinations can lead to an overallwelfare improvement if these arecomplementary.

The relationship between policy reformand investment is shown in Figure 1using a stylized production possibilitycurve. The total reform and developmentexpectations are represented by amovement from point A on theprereform/development productionpossibility curve, to point D on thepostreform/development productionpossibility curve. This can bedecomposed into four elements.• First, reforms that improve the

efficiency of existing institutions andresource use result in movement fromA to B, with A representing a state ofsuboptimal performance, to thepresent production possibility frontier.

• Second, a movement from B to C canbe the result of, for example, pricepolicy changes that will lead to amovement along the productionpossibility frontier.

• Third, a movement from C to D canbe the result of investments ininstitutions or production.

• Fourth, a movement from A to C, andthen to D, represents an idealized statethat reflects a sequencing of reformsand development.

6 Ali (1990) also observes that price policy reformcan be viewed as a means of narrowing pricedistortions and the wedge between demand andsupply. In this sense, the purpose of (price) policyreform is to increase efficiency through resourcereallocation, which involves movements along theproduction possibility frontier, whereas a publicinvestment project could be seen as adding to thesupply of capital and moving the curve itself.

20

CHAPTER 3

D.Scope of Sector Analysis

Program proposals can result from theanalysis carried out as part of countrystudies, sector analyses, and dialoguewith the government during countryprogramming. They can also result fromproject analyses highlighting policyissues that are fundamental both toproject sustainability and replication,and to program and project monitoringand evaluation (M&E). Figure 2exemplifies the various stages in ADBoperations that can give rise to policyissues. While project analysis,implementation, and M&E can all makevaluable contributions to the

identification of policy issues, sectoranalysis is the logical starting point forconducting systematic analysis of thesector and of related development andpolicy issues.

A significant focus of sector analysis inADB is to identify and understandconstraints and their causes. Whileresource limitations and allocativechoices are at the heart of an economicproblem, policy-related failures may bea cause of the problem. Examples areprovided in Box 1.

To identify and understand the rangeof market, nonmarket, and institutional

PPF = production possibility frontier.

FIGURE 1: Effects of Programs and Projects on Production Possibility



CH

AP

TER 3

DM

C =

dev

elop

ing

mem

ber c

ount

ry.

Sou

rce:

ER

EA, A

sian

Dev

elop

men

t Ban

k.

FIG

UR

E2:

Sta

ges

in A

DB

Op

erat

ion

s G

ivin

g R

ise

to P

olic

y Is

sues

Monitoring and Evaluation

22

CHAPTER 3

BOX 1: Policy-Related Economic and Institutional ProblemsIdentifiable in Sector Analysis

• Differences between the market and opportunity costs of a resource due to government pricemanagement (e.g., minimum wages).

• Barriers to market entry due to government-imposed subsidies, tariffs, quotas, nontariff barriers, andpublic enterprise dominance.

• Investment disincentives due to poorly defined and unsecured property rights.• Failure to correct negative externalities due to lack of defined standards and their enforcement

(e.g., pollution).• Limited access to and supply of resources available for development (e.g., underdeveloped capital

markets).• Uncompetitive markets and inefficient provision of goods and services (e.g., lack of monopoly

regulation).• Failure of private markets to provide a good or service.• Failure of markets to provide information (e.g., prices or product safety).• High transaction costs among economic agents and markets.

Source: EREA staff, Asian Development Bank.

issues that affect sector performance, asystematic analysis is required. Arelevant scope for sector analysis ofissues to support program lending isdescribed in the 1996 review of ADB’sprogram lending policies (ADB 1996).Elements of the discussion are includedin the following summary of theelements of a sector analysis:• Macroeconomic Context and Country

Management. This includes (asdescribed in Chapter 2) economicprojections, macroeconomic assess-ment, national development strategiesand plans; public sector resourcemanagement; overall social conditions;and country poverty assessmentincluding income distribution andpatterns of poverty.

• Sector Description. This includesassessment of the sector’s role,features and resources, changes andtrends, the demand for sector goodsand services, the overall supply ofsector goods and services includinginput and output system efficiency,

government policies, reform agendaand key institutions.

• Markets, Prices, and the IncentiveStructure. This relates to a specificindustry’s conduct, performance andstructure; policy-caused pricedifferentials such as those from taxes,subsidies, and monopolisticpractices; market underperformance orfailures; influences on transactioncosts, profitability, andcompetitiveness; response to theglobal context and internationaltrade; and assessment of privatesector needs and constraints.

• Institutional Performance. Thisconcerns the role, conduct, andperformance of government and publicinstitutions, including: theinstitutional framework; public goodsand service provision and capacity;fiscal and public expenditure incidenceand impact; the state of the enablingand regulatory environment; andinstitutional and nonmarket failures,such as bureaucratic malfunctions andcorruption.



CH

AP

TER 3

BOX 3: Examples of Price Policy Trade-Offs in the Agriculture Sector

• Price controls to lower consumer prices can reduce farmer incomes (depending onelasticities).

• Raising producer prices can raise the consumer cost of living.• Input subsidies may increase production but distort resource allocation and weigh in on

fiscal resources.• Food subsidies may lower consumer prices but weigh in on fiscal resources.• National food self-sufficiency through government traders (monopolies) can result in

allocative inefficiencies and welfare losses, and undermine sector growth.• Production extensification and intensification can have negative environmental externalities.

Based on: P. Streeten. 1987. What Price Food? Macmillan Press.

A distinction needs to be made betweentwo major types of market-relatedpolicies—price policies that canfundamentally affect enterprise returnsas well as economic efficiency, andinstitution-related policies that affect thefacilitation and regulatory role ofgovernment, and the alignment ofinstitutions in relation to users. Thenext two subsections highlight specificissues in price policies and ininstitutional policies that should beassessed at the sector level.

E. Price Policy Context

Differing policy objectives can haveprofound effects on the allocativeefficiency of resources and distribution

that can lead to changes in overallwelfare. So, policy implications ofreforms are of interest to a wide rangeof domestic and internationalstakeholders as well as developmentpractitioners. The economic soundnessand impact on stakeholders of existingand alternative policy environments needto be assessed in relation to each other.Box 2 provides examples of conflictingpolicy alternatives in the agriculturesector. The examples of policyalternatives provided in Box 2 are oftenachieved through the use of differentprice policies. Box 3 provides examplesof how the use of price policy inagriculture can affect differentstakeholders. Further, it shows thatpolicy alternatives are often mutually

BOX 2: Examples of Conflicting Policy Alternatives in the Agriculture Sector

• Promoting resource allocation efficiency in agriculture (and the economy) to raiseproductivity and competitiveness.

• Accelerating economic growth through expansion and support for agriculture.• Reducing malnutrition and increasing rural incomes by emphasizing expanded small farmer

production and employment creation for the landless.• Improving food security through price and supply stabilization and extensification.• Promoting self-sufficiency and price control to maintain political stability.

Source: Adapted from P. Streeten. 1987. What Price Food? Macmillan Press.

24

CHAPTER 3

exclusive, implying efficiency anddistribution trade-offs.7

Project sector assessments andeconomic analysis routinely identifyprice differentials resulting from marketstructure and policy choices, as reflectedin the divergences between financial andeconomic prices. In many cases, the pricedifferential and its implications may betolerable within the context of economicperformance. However, when thesedifferentials are so large that they

undermine efficient allocation ofresources—as when production issubsidized but prices are controlled—then policy changes may need to beconsidered.

Tables 2 and 3 summarize examples ofthe causes of variation in financial andeconomic values in the context ofagriculture and rural development. Thesetables indicate that the extent ofdifferences between values is usually areflection of the policy environment and

7 This example is discussed in further detail togetherwith approaches to analysis in Chapter 5 and Case1 in Appendix 3.

TABLE 2: Factors Causing Economic Prices to Exceed Financial Prices

Taxes on output

Examples from Rural Development andAgricultural ProjectsFactors Causing Difference

Subsidies on inputs

Foreign exchangepremiums

Positive externalities

Price ceilings

• Export duties and taxes• Controlled food prices for consumers (depresses producer price)• Fertilizer prices to farmers• Nonmarket-based credit

• Overvalued foreign exchange depresses output earnings andraises input costs (in domestic currency)

• Users of water at a tariff that does not cover all capital, andoperation and maintenance costs

• Agricultural project production encourages growth in agro-processing

Source: EREA, Asian Development Bank.

Subsidies on output

Examples from Rural Development andAgricultural ProjectsFactors Causing Difference

Minimum wages

Negative externalities

• Produce bought at high fixed price by state marketing board• Import duties on agrochemicals and equipment

• Undervalued foreign exchange increases output earnings anddepress input costs (in domestic currency)

• Financial price of unskilled/surplus labor is higher than itseconomic price

TABLE 3: Factors Causing Financial Prices to Exceed Economic Prices

Foreign exchange discounts

• Environmental damage


Taxes on inputs



CH

AP

TER 3

the choices being made by policy makers.In Table 2 where, for example, economicprices are higher than financial prices offood products due to a tax, these canreduce the incentive to produce andsuggest a policy regime favoring theconsumers. Conversely, a subsidy oninputs will cause their economic pricesto exceed their financial prices, reflectinga policy bias toward producers (Table 3).

As more economies adopt market-basedsystems and freer trade principles, thedivergence between financial andeconomic prices has generallydiminished. However, there can still beconsiderable variations between financialand economic prices across sectors,especially for nontradable goods andservices within an economy, reflectingdivergent sector policies and priorities.In most project cases, a degree ofdistortion exists, due to taxes andtransfers, but at an acceptable level. Inother cases, excessive biases in revenuecollection or wasteful use of transferscreate significant welfare inefficiencies.Where this is an issue for projects, thegovernment may agree, as part of aproject, to a limited set of new policies,such as improved tariff collection effortsor the reduction of a subsidy, to ensuresustainability of the project.

In general, projects assume a givenenvironment of market or nonmarketimperfections and distortions, whereaspolicy-based loans are intended tochange or remove these distortions. Thismeans that understanding the extent andcauses of financial and economic pricevariations is still needed in addressingthe policy issues related to productivesectors. In cases where the sector

assessment and project analyses identifypolicies that are disruptive to marketprocesses—and, for example,significantly affect the financial oreconomic viability of a project—thenpolicy reforms may be warranted toreverse losses.

The general implication is that reasonedand plausible analysis is needed toidentify the positive and negativeimpacts of price policy on financial andeconomic performance of productionand services, including the effects ondifferent stakeholders. Onceperformance-affecting distortions havebeen identified, the extent to whichstakeholders understand the trade-offsshould also be assessed, since theirinfluence can dramatically affect themomentum for reform.

F. Institutional Context

1. Institution Building and Reforms

Until the early 1980s, developmentefforts focused largely on filling thehuman and financial resource gaps.Throughout the 1980s and into the1990s, the IMF and the World Bankemphasized “getting prices right”through price-focused structuraladjustment. Such reform measures arestill needed to address, for example,specific price constraints where they arebinding. But, the nexus betweeneconomic management, institutions, anddevelopment is receiving greater focus.

The changing focus stems from thecrucial role that institutions play in thefunctioning of markets, especially in thetransmission of market signals, and in

26

CHAPTER 3

8 The World Development Report 2002 provides acomprehensive description of the facets ofinstitutions that need to be understood, includingfirms (from farmers, to the governance of firms, tofinancial systems), government (politicalinstitutions, the judicial system, managingcompetition, and the regulation of infrastructure),and society (norms and networks, and the media).

providing appropriate incentives toeconomic agents. More importantly,institutions allow major elements in thepolicy reforms to be sustained over themedium term. While policies can have asignificant influence on institutions,institutions, in turn, can also affect theform and adoption of policies. Stiglitz(1998) suggests that development isabout transformation of society, andinstitutional changes are part of thistransformation. So, countries will haveto build capacity as well as reform andreplace the traditional institutions thatwill be superseded in the developmentprocess. In this regard, three elementsneed to be recognized: (i) the historicaland social context of institutionalchanges; (ii) that changes take place asa process of evolution and adaptation;and (iii) that such changes involve broadparticipation and consensus building.

Viewed holistically, an institution is acollectively shared, self-sustaining systemof beliefs about the “rules of the game”in a society, the enforcementmechanisms, and the organizations thatsupport market transactions and publicgoods delivery. These rules, mechanisms,and organizations are important as theydetermine the incentives for political,social, and economic exchanges,including such aspects as property rights,regulations against fraud andanticompetitive behavior, the rule of law,and the judiciary (Rodrik 1999). Whensuch basic rules do not functionefficiently, market opportunities arelimited by institutional constraints suchas: high transaction costs arising fromasymmetric information; poorly definedand weakly enforced property rights; andbarriers to market entry for new

participants. A vital aspect of the reformprocess is therefore the realigning ofinstitutions, including policy changesgeared toward improving institutionalarrangements that determine the contextin which policies are made.Improvements in institutions that lowertransaction costs, manage risk, andenhance competition are also key(World Bank 2002a).

The World Bank’s World DevelopmentReport 20028 highlights the importanceof institutions by asking why somemarkets are rewarding for some groupsand not for others, and why some areinclusive and well integrated while othersare localized and segmented (WorldBank 2002a). The reasons for theseperformance variations are explained interms of the rules, mechanisms, andorganizations. The effectiveness ofinstitutions determines the extent towhich people, from the rich to the poor,have the opportunity and incentive to beinvolved in gainful market activity (WorldBank 2002a). Stiglitz (2000) notes that“in assessing institutional arrangements,one has to evaluate not only theinstitutions that exist, but those that areabsent as well.” These points underscorethe need for institutional responses forproviding appropriate social insurance,enhancing industrial relations,overcoming labor-market rigidity,expanding access to information and



CH

AP

TER 3

education, and streamlining titleregistration and business licensing.Stiglitz (2000) further adds “from theperspective of policymakers, the mostimportant issues are how to encouragethe creation of good institutions andhow to design institutions that canchange as the environment changes. Bothquestions demand careful thoughtabout organizational design andsequencing of reforms.”

An implication of institutional policychange is the complexity of reforms thatmust be understood within the contextof individual economies, including theinstitutional environment andarrangements9 (Klein 1999). Policychange is also a process that takes placeover time, with the cumulative effectemerging from the complementary andmutually supporting development ofother institutions. So, the timedimension of institutional change needsto be better appreciated, as the time lagsbetween initiating changes and theirultimate impact can be significant,especially at the lower layers of thehierarchy of institutions. The longer-term consequences of such institutionalunderpinnings should not beunderestimated.

In addition, changes in one set ofpolicies or institutions typically lead tothe need to overhaul related institutions.For example, the problems of corporate

debt overhang and insolvency have ledto measures for financial liquidationand bankruptcy proceedings thatsubsequently required changes inexisting laws and the judicial system.Given their context specificity and localinter-connectedness, the idea ofinternational best practice in institutionshas limited relevance, with few cases of“one size fits all” in institutional policyand design, especially in the context ofinstitutional structures that lead toeconomic growth and poverty reduction.

Strategically, institutions need to bedeveloped with local involvement, usinglocal knowledge, hands-on experience,and even experimentation (Rodrik1999). Aoki (2000) notes that a widevariety of arrangements, both formaland informal, can be considered as“institutions” for as long as economicagents take them to be relevant, fromstatutory laws and regulations tounwritten rules and codes of behaviorbased on social ties and trust. Similarly,a formal regulatory body is only aneffective institution if it can enforcesanctions on these agents. Thissuggests that only institutionalarrangements that are mutuallyconsistent or reinforcing may be viableand sustainable in an economy. So, theoutcome of policy advice is determinedby the interactions of the strategicexpectations of stakeholder groups.

With regard to the capacity-buildingaspect of program design, institutionaldevelopment can only emerge whenstakeholders are willing to bring thisabout; when they coordinate theirchoices; and when they share a consensuson the reform measures to be adopted.

9 Institutional environment refers to the background/institutional constraints, or rules of the game thatguide agents’ behavior. Institutional arrangementsrefer to the governance structures designed tomediate particular economic relationships, whichcan include business firms, long-term contracts,public bureaucracies, and nonprofit organizations.

28

CHAPTER 3

Consequently, policy change that dealswith or affects institutions involves arelatively complex process thatfrequently evolves in conjunction withdevelopment efforts.

Bringing together some of the abovestrands, the study of institutions, as partof sector analysis, should focus onfostering institutional arrangements andinnovations that address causes ofconstraints and failures. The ultimateaim is to induce coordination andcooperative solutions in implementingpolicy changes. In turn, this will helpmitigate conflict and realize mutual gains.Second generation reforms, for example,will require the design of new sets ofinstitutions and institutionalarrangements that should ensureappropriate incentives to economic agentsand organizational structures that are“incentive compatible.” With the aim ofaligning governance structures withopportunities and incentives, reformprograms may also be designed tominimize institutional constraints,coordination failures at various levels ofgovernment, transaction costs, andasymmetries in wealth/assets andinformation. Thus, the interaction betweenfirst and second generation reforms needsto be clearly understood while the timingand sequencing of respective reformsshould be carefully considered (seeChapter 3.E). In this context, policyanalysts and makers need both to adopta long-term view of development and toundertake detailed diagnosis that will helpin building local institutional capacity.

Throughout the 1990s, ADB policyoperations increasingly addressedinstitutional reforms. The need to

address institutional development andcapacity limitations in reform situationsis explicitly acknowledged in coreprogram loan reports, as summarized inTable 4.10A theme that runs through eachof these loans is the absorptive capacityof existing institutions to deal witheconomic and social development.However, absorptive capacity is a broadterm and various perspectives need to beconsidered if capacity-building initiativesare to be appropriate. In this context,absorptive capacity refers to the policy-making environment; the managerial andadministrative capacity of institutionsresponsible for translating policy intothe delivery of programs and services;and the technical capacity ofinstitutions. With current practices inmind, the following subsectionssummarize generic points in assessingthe institutional framework, as well asissues in the managerial, technical, andabsorptive capacity of institutions.

2. Understanding Functions andPerformance of Market-RelatedInstitutions11

In its discussion of market institutions,the World Development Report 2002 providesa useful basic framework forunderstanding the functions andperformance that could become the basisfor effective institutions. Three aspects

1 0 For example, Abonyi (2002) identifies variabilityin the assessments of certain reform programs(e.g., VIE: State-Owned Enterprise Reform andCorporate Governance Program), not least fromthe point of view of reaching a mutualunderstanding between the government and theinternational financial institution as to theapproach to and content of the reforms.

1 1 This subsection draws heavily on World Bank(2002a) (p. 8).



CH

AP

TER 3

Sources: Asian Development Bank, reports and recommendations of the President of above loans (see Appendix 1).

TABLE 4: Institutional Issues Considered in ADB Program Loans

Weak commercial bankingskills and inadequatefinancial systems

Policy development, market-relatedinstitutions, and public support serviceefficiency improvement (management andtechnical):

Banking supervision, regulation standards andprocedures development, skills developmentthrough long-term advisory support, andon-the-job training.

MON: FinancialSector Program

Strengthening institutionaland regulatory framework forstate-owned enterprisereforms

Policy development and coordination:Individual steering committees integrated toidentify priorities and sequencing of reforms,monitor progress, mobilize support.

VIE: State-OwnedEnterprise Reform andCorporate GovernanceProgram

Development of financialmarket infrastructure tofacilitate private sectorparticipation includingcommercializing andmodernizing banks, securedlending, and capital marketdevelopment

VIE: Financial SectorProgram

Market-related institutions and public supportservice development (management andtechnical):

Development of the legal and regulatoryframework including passage of legislation,implementation of asset valuation, collateralsecurity and registration systems, accountingand audit system establishment to internationalstandards, and information disclosure.

Facilitation of structuralreforms, mitigation ofshort-term impact ofstructural reform, socialservice efficiencyimprovement, anddecentralization

Facilitation of structural adjustment reformsand public service development (management):

Start of decentralization with specificassistance in mitigation for displaced workersand education support for low-incomefamilies, planning and management systemsand capacity, monitoring systems, and labormarket information flows.

THA: Social SectorProgram

Fiscal management, andfinancial institutionstrengthening in terms ofinformation disclosure andregulation

Nonmarket and market failure containment,policy development for market-relatedinstitutions, and public support service(management and technical):

Management and technical training in centralbank and nonbank financial servicessupervision, skills development in riskassessment and new financial services, andaccounting and performance audit skillsdevelopment.

INO: FinancialGovernance Reforms

Pension system efficiency andsustainability of market andnonmarket systems

Policy development, market and nonmarketinstitutional development to address serviceefficiency and sustainability:

Reorganization and establishment of publicinstitutions, system expansion, improvementof management and administration capacity,and technical capacity building.

KAZ: Pension ReformProgram

Implementation of tradeliberalization, tradefacilitation restructuring, andprivatization

Support for trade liberalization policy,development of market-related institutions, andnonmarket support services:

Modernization of the management andtechnical aspects of customs administration,and efficiency improvement of the exportpromotion board.

PAK: Trade, ExportPromotion andIndustry Program

Loan Type and Means to Address CapacityLimitations

Example of CapacityIssues Addressed

30

CHAPTER 3

are highlighted: the channels ofinformation that exist to inform currentand potential new participants aboutmarket conditions and goods; the extentto which property rights and contractsare defined and enforced as the basis fordetermining access to resources andincome; and the extent to whichinstitutions increase or decreasecompetition in markets.

a. Channels of information

Good information flows help formaland informal businesses identifyproduction and trading partners, as wellas compare and identify possible returnsin terms of size, risk, trustworthiness,and creditworthiness of potentialpartners such as suppliers, producers,buyers, borrowers, and lenders. Each ofthese partners, as part of the institutionalsetup, can play a role in the collection,production, analysis, verification, anddissemination of information andknowledge. Public or privateinstitutions, such as banks, can performthis role for specific purposes. Theextent to which the various participantshave access to sufficient informationdetermines the likelihood ofparticipation. Development-needsassessment is based on an analysis of:the extent to which formal and informalinstitutions collect, provide, anddisseminate information as a matter ofpolicy; the costs of collectinginformation; and related capacitylimitations.

b. Property rights and contracts

Understanding and clearly definingrights to assets and income and beingable to protect agreed rights is essential

for market development. The rights ofthe private sector in relation to the state,and foreign participants in relation todomestic participants, as well as theability of institutions to handle disputesand enforce contracts, have a crucialbearing on market and participantconfidence. Not only must corporateand SME investors and innovators beassured of their rights, but also the poor,risk-averse investors who stand to losewhat little they have. The focus ofassessment is likely to be: constitutionalprovisions; the judicial system; othersocial networks and mechanisms forrights determination, adjudication,registration, and settlement; and thecosts of enforcement.

c. Competition in markets

The value of competition throughequal opportunity is the incentive forparticipants to do better and channelresources to their best use, comparedwith systems that depend on social andpolitical connections. On the one hand,this requires institutions that facilitatecompetition, by providing key marketentry information and rules ofoperation and by minimizingtransaction costs. On the other hand,institutions play a key role inregulation that ensures fair practicefrom the point of view of allparticipants. Problems arise in over-and under-regulation of markets,resulting in exclusion to the benefit ofa few participants in the former, andexploitation and other negativeexternalities in the latter. Assessmentson the appropriate level of marketregulation will help determine theextent to which public and private



CH

AP

TER 3

regulatory institutions help or hindercompetition.

Collectively, the above aspectsdetermine critical policy-related mattersincluding the distribution of assets,incomes, opportunities, and incentivesfor market participants, as well as theefficiency of market transactions. Theperformance of these institution-dependent economic variablesdetermines the pace of improvement ofproductivity, economic growth, andpoverty reduction.

3. Managerial and AdministrativeCapacity of Institutions

Assuming that the politicalcommitment exists to pursue a reformpath, the next areas to be assessed arethe nature, role, and abilities of theinstitutions responsible forimplementing policy changes. Whenpolicy reforms address enterpriseunderperformance problems, such aspoorly managed public enterprises, infavor of emerging market-basedopportunities, reforms can move themanagement of activity to moreappropriate public , or private ,institutions. When reforms addressthe delivery of public goods andservices, assessment of the capacity ofinstitutions to deliver in the contextof programs will be as important asproject institutional assessments, ifnot more so, because of their oftensector-wide effects. Box 4 outlines keyelements of managerial andadministrative absorptive capacity thatrequire an understanding for thepurposes of policy operations.12

4. Technical Capacity of Institutions

Closely related to managerial andadministrative capacity is the issue oftechnical capacity to implement reforms.Again, where reforms are proposingrealignment of responsibilities betweeninstitutions, whether public or private,appropriate technical procedures and aminimum level of technical capacitymust exist or be developed forinstitutions and reforms to be credible.For example, a realignment of publicservices may require substantialinvestment in technical training orreorientation from direct serviceprovision to a support and regulatoryfunction. Decentralization oftenrequires not only massive fiscaldecentralization, but also extensive andcostly investments both in local skillstraining for administration of funds andthe technical aspects of planning, andservice delivery that were previouslyhandled by central government staff.Such capacity building needs to beintegrated with the adjustment effort.Box 5 summarizes technical capacityconsiderations for policy operations.

G. Reform Timing and SequencingIssues

1. Economic Considerations

Economic theory suggests that an agenttakes one action that leads to the highestmarginal returns in the circumstances.This theoretical proposition shouldapply to all decision-making

1 2 A comprehensive ADB source on issues andassessment of public administration isSchiavo-Campo and Sundaram (2001).

32

CHAPTER 3

BOX 4: Factors to Consider in Assessing Managerialand Administrative Absorptive Capacity

• Which are the core public institutions involved in the implementation of policies, specificsteps (e.g., decisions/activities), and related policy conditionalities?

• What are the key capacity-related assumptions about implementation capacity for reforms,steps, and related conditionalities?

• How do decisions/activities relate to each other including required linkages, coordination,and cooperation among institutions?

• Do senior and mid-level management understand and support the reform?• Is there an acceptable level of administrative integrity and accountability (has due diligence

been conducted)?• What is the extent of user confidence in target institutions?• What capacity gaps exist and how can they be reduced, such as change of conditions, or

strengthened capacity?– Is the organizational structure appropriate and what, if any, are the requirements forchange?– Is there clarity in management and administrative roles and responsibilities?– Is management of public service personnel, including motivation, incentive, and rewardsystems, conducive to reforms and what is the likely responsiveness?– What are the prospects for institutional adaptability?– Have human resource development needs been assessed and determined?

• What are the implications for public expenditure, including distribution of fiscalresponsibilities; interdepartment realignment/reallocations; short-term adjustment costs;medium-term savings; budget execution; changes/introduction of cost-recovery systems; andreporting, accounting, and auditing systems?

Source:Based on G. Abonyi. 2002. Toward a Political Economy Approach to Policy-Based Lending. ERD WorkingPaper Series No. 14. ADB, Manila.

processes—choosing among manyfeasible actions the one that makes thelargest impact, or alternatively, choosingto tackle the binding constraint.Sequencing policy changes should alsofollow this first order of principles:What is the most binding policy orinstitutional constraint? Oncedetermined, it should be removed first.Similar choice and sequencing processesin the fields of project management andengineering include such techniques ascritical path analysis (ADB 1986).

The analytical principle that shouldgovern the timing and sequencing ofchanges is always the concept ofremoving the binding constraint at themacro, meso, or micro level (Chapter2.B). Indeed, as Collier and Gunning

(1999) observe, many structuralconstraints have to be addressed alongsidemacro stabilization measures. In thissense, stabilization and structuralchanges cannot be viewed in isolationor in a static way. Rather, they shouldbe viewed as an interlinked and dynamicprocess. Easing an interlinked sequenceof bottlenecks, therefore, is a dynamicprocess. Nevertheless, priority actionsand measures should be established.

For example, in response to external andinternal imbalance, the intended effectof macroeconomic stabilization is torestore balance and stability asfundamental to economic growth. At thesector and microeconomic levels, loss ofdomestic and international marketcompetitiveness—due to chronic



CH

AP

TER 3

BOX 5: Factors to Consider in Assessing Technical Absorptive Capacity

• Existence of appropriate technical systems, procedures, and standards.• Technical capacity of institutions and staff to carry out realigned tasks.• Sufficiency of budgets to carry out realigned tasks.• Adequacy of technical support for institutions and staff assuming new responsibilities.• Adequacy of human resource development provisions for technical aspects.• User confidence and satisfaction in technical aspects of service delivery.


inefficiencies in factor use and lowproductivity, falling export prices, orrising imported raw material prices—may require more specific structuralreform action as a subset of firstgeneration reforms.

Similarly, Agenor and Montiel (1999)indicate that the sequential order, timing,and optimal pace of specificmacroeconomic and structural reformsand their interrelationship are importantfor conceptual and practical reasons. Interms of, for example, the financialsector, they suggest that fiscal imbalanceand control should be addressed as aprerequisite to undertaking the fullrange of financial sector reforms, suchas liberalization of the domesticfinancial system and removal of capitalflow restrictions, in order to avoid therisk of inflation and debt overhangwhich may, in turn, result in capitalflight. Further examples of sequencingconsiderations for financial sectorreforms are provided in Box 6.

In the context of ADB operations,sector-level reforms usually assume thatfirst generation reforms have alreadybeen implemented. If they have not been,macroeconomic constraints will likely

influence the success or failure of policyoperations. For this reason, it isnecessary that ADB program loanreports include a realistic and pragmaticassessment of the extent to which earlierreforms have been effectivelyimplemented. Table 5 shows examplesfrom three of the core program loanreports as to how timing and sequencingissues arose and were addressed.

2. Political Economy Dimensions

The timing and sequencing of policychange should not, though, be viewedas a mechanistic process. The timing andsequencing considerations need to gobeyond the technical aspects of “optimalpolicy design,” and several otherconsiderations need to be borne in mind.These include the related magnitude ofeconomic and social dislocations—theadjustment costs and the politicalfeasibility of a policy change.

Collier (2001) stresses the need for adiagnostic procedure that sufficientlyanalyzes the range of possible bindingconstraints on growth and the propersequencing of priorities. Furthermore,he observes that policies depend on thebalance between political constituencies,

34

CHAPTER 3

with many stakeholders and electorateshaving insufficient information toinfluence and control their leaders. Fourareas are identified as bottlenecks: lackof information, lack of a way to turninformation into knowledge, lack ofcapacity for analysis, and lack of capacityfor policy design. Assuming thatinternational financing institutions havethe capacity for analysis and can assistwith design, it is essential that they:encourage standards of good practice interms of gathering and sharinginformation; help in analysis, design,and dissemination of policy options;and encourage lender-borrowercollaboration, based on partnership asopposed to coercion. The practices areconsistent with issues such as inclusion,empowerment, and civil participation asthey relate to policy-based operations.

In considering the political economydimensions of adjustment, even at thesector analysis stage, it is necessary forthe policy analyst to: (i) sift through

various policy alternatives and paths tosearch for a “least-cost” solution, as wellas design an acceptable package andsequence of reforms; (ii) test thepolitical reaction as a means of assessingthe demand for changes from both activeand latent constituencies, and improveunderstanding of the nature of policyreforms; and (iii) identify the likelyadjustment process and related costs asa way to assess the different dimensionsof reforms (see Chapter 6). The lastpoint is also made by Collier (2001)who indicates that in countries withinadequate policies, there can be a largelatent constituency for policy reform.

In discussing the feasibility of policyreforms, Sadoulet and de Janvry (1995)suggest that “A policy that does notpass the test of political feasibility is autopian proposition.” Made in thecontext of introducing quantitativedevelopment policy analysis, thishighlights the need to understand thepolitical as well as the economic and

Box 6: Examples of Sequencing Considerations in Financial Sector Reforms

• A sound fiscal position is needed to deal with potential bad loan situations and to ensurethat franchise banks maintain their value.

• Establishing and deepening the domestic money and securities markets are necessary todevelop the domestic banking system and to ensure that financial opening does not lead toexcessively high interest rates and overintermediation.

• Policies that safeguard property and promote investor confidence are necessary beforeremoving capital flow restrictions to avoid capital flight, especially where there is debtoverhang.

• Of particular importance is the liberalization of the enabling environment for foreign directinvestment (FDI), trade, and finance to raise investor confidence, and to ensure that thebenefits of FDI can be realized (know-how transfer and improved commercial openness).

• The liberalization of and increased competition in domestic financial systems (freeingdomestic interest rates, use of indirect instruments for monetary control, etc.), as protectedby adequate accounting-standards supervision and prudential regulation, should occur beforethe capital account is opened to avoid capital flight, a balance-of-payments crisis, andmisallocation of external funds to under-performing investments.Source:Pierre-Richard Agenor and Peter J. Montiel. 1999. Development Macroeconomics. Second Edition.

Princeton University Press.



CH

AP

TER 3

TABLE 5: Economic Timing and Sequencing Issues in Program Loan Reports

RRP Key Context/Issues Justification for Reform Timing/Sequencing

VIE: State- Viet Nam’s gross domestic Following the immediate stabilization measures,Owned product (GDP) growth dropped the Government recognized that sector reformsEnterprise by nearly half from 8.2% in 1997 would have to be significantly deepened forReform and to about 4.4% in 1998. This Viet Nam to effectively cope with theCorporate decline was partly caused by consequences of the regional crisis and toGovernance the regional financial crisis, reestablish sustainable growth. To sustain growthProgram declining foreign direct in the industry sector, specific reforms were

investment disbursements, intended to: (i) create an enabling environment forand a sharp drop in exports. foreign investment; (ii) support the development ofThe Government reacted by private enterprises; (iii) accelerate state-ownedcontrolling public expenditure enterprise reform by strengthening the institutionaland monetary growth, devaluing framework forsupporting corporatization andthe dong on two occasions, and commercialization of state-owned enterprises;introducing import bans on major (iv) introduce a system of improved corporateconsumer goods. governance; and (v) enhance labor mobility.

THA: Social After a GDP growth of 8.7% in Initially, the Government proceeded with itsSector 1995, the Asian financial crisis economic reform in five areas: (i) monetary andProgram resulted in zero GDP growth fiscal policies, including macroeconomic targets

in 1997 and an anticipated and management to restore macroeconomicnegative growth rate in 1998. stability; (ii) financial sector problems faced byThe crisis was primarily financial institutions; (iii) public administration,traceable to an exchange rate including reforms of state enterprises and the civilpolicy of tying the baht to a service; (iv) industrial restructuring to boostbasket of currencies heavily overall industrial competitiveness; and (v) newweighted by the US dollar, initiatives in the social sector. Social sector policiescombined with imprudent were seen as important in terms of their timing tolending policies and inadequate mitigate the negative social consequences of thesupervision of banks, especially crisis and the possible hardships resulting fromprivate financial companies. economic reform/restructuring measures.

INO: Financial Although economic performance Following the unfavorable reception of the draftGovernance was largely on track in the first 1998/99 budget, the Government entered into aReforms half of 1997, and despite the strengthened stabilization program with the

apparent strong macroeconomic International Monetary Fund in mid-January 1998fundamentals, the rupiah began with an emphasis on fiscal stabilization measures.to weaken in the second week of As the crisis had adversely affected the financialJuly 1997, following the floating sector, the program loan was formulated to restoreof the Thai baht. A number of confidence through financial governance reforms.factors such as the level of Reform measures consisted of: (i) improvingforeign debt, heavy investment financial governance practices, (ii) increasing thein the property sector, disclosure and transparency of financialweaknesses in supervision and information, and (iii) strengthening the legal andregulation of the banking system, regulatory framework of the financial sector.governance problems,uncertainty about politicalsuccession, and emergingsocial tension erodedconfidence and contributed toa large decline in the rupiah.

RRP = report and recommendation of the President.Sources: Asian Development Bank, program loan reports of above loans (see Appendix 1).

36

CHAPTER 3

social dimensions of reform. Abonyi(2002) also makes the point thateffective policy operations require anappreciation that policy reform is a“domestic game,” even when initiatedexternally. Thus, policy making andassociated processes of economicliberalization, such as the pace ofprivatization, need to be pragmatic andto closely consider influences thatconstrain macroeconomic and sectorpolicy choice, including politicalacceptability. Reforms and policies mustbe consistent with the broader policyenvironment.

Implementing institutional changes,such as decentralization, may take evenlonger, given entrenched vested interestslet alone the logistical challenges of sucha change. Again, the time lags betweeninitiation of change and implementationshould not be underestimated. Sectoranalysis can play a major role inassessing the readiness, interest, and timeinvolved in such changes and caninitially assess the program implications.

The above points stress that institutionsshould not be treated as “black boxes”that readily transform inputs intoexpected program outputs. A relevantdesign issue is the extent to whichpolicy conditions can be formulated ina way that is compatible with adoptionincentives. In this regard, anunderstanding of the political economyof the process of policy reform isessential (Abonyi 2002). According toStiglitz (1999), many attempts atreform, at both the macroeconomic andsector levels, are less than successfulbecause of their proponents’misunderstanding of the foundations of

a market economy and of the basics ofan institutional reform process.

It is therefore clear that policyoperations and the responsibility ofgovernments to implement sometimesfar-reaching reforms require, mostimportantly, careful assessment of thecapacity to adapt and the willingnessand incentive to change (Box 7). To thisend, sector analysis presents anopportunity to update theunderstanding of domestic institutionalarrangements and of the political andpolicy development process; assess thetime and resources needed to effectchange and develop capacity; andminimize efficiency losses.

Overall, the determination of what thepolicy and institutional constraints are,including the most binding constraints,should be the task of sector diagnosis.Carefully done, this should include notonly a descriptive review of what is goingon in a sector, but also an analysis ofwhy things are going on as they are andof what alternative measures can be takento improve the prevailing policy andinstitutional context. This requirespolicy makers and analysts to have agood understanding of the workings ofthe economy and the sector.Appreciating the distinction betweenprice and market reform issues on theone hand and institutional reform issueson the other, which second generationreforms typically address, can also becrucial in comprehending the context ofspecific types of reform and probableresponse, timing, and sequencing issues,as well as possible outcomes. Clearly, thetiming and sequencing of policy changesconstitute a complex and interlinked



CH

AP

TER 3

BOX 7: Factors to Consider in Assessing Capacity to Adapt,and Willingness and Incentive to Change

• What are the sector and institutional boundaries of the policies under consideration?• Who are the key stakeholders with an interest in and likely to be affected by policy changes?• How will political and development ideology, ongoing shifts and possible inertia, and their

sources affect alternatives?• What is the policy-making style of the client government (e.g., autocratic, democratic,

consensus-based), including an understanding of the legislative and consultation processes,and, importantly, their timing?

• To what extent is the policy area a perceived priority by different stakeholders?• Does a “reform champion” exist?• Which stakeholders perceive decreases/increases in net benefits as a consequence of policy

changes? What are the political implications and the way this is likely to be handled by thegovernment, including compromises, compensation, and trade-offs? Do stakeholders havethe power and means to influence the policy process (e.g., approval, block, implement) andthe incentive to do so? How can opposers be induced to support, or at least, not opposepolicy changes?

• What is the extent of technocrats’ and legislators’ supporting internal analyses of policyissues, and of their understanding?

• Is the timing of policy change appropriate in terms of the stability, reputation, and life of thegovernment (e.g., progressive and reform-minded, new, mature, with agenda conflicts,transparent, intransigent, existence of policy vacuums)?

• What is the overall extent of political will and commitment, and what are the keymotivating/deciding factors?Source:Based on G. Abonyi. 2002. Toward a Political Economy Approach to Policy-Based Lending. ERD Working

Paper Series No. 14. ADB, Manila.

system that has to be dealt with in adisciplined manner. That is, there aresome fundamental principles that need

to be observed, and they have to beanchored in a rigorous sector diagnosis.

CHAPTER

ASSESSING THEEFFECTS OF POLICYCHANGE

l Assessing the Effects of Policy Change

l Approaches to Analysis

40

CHAPTER 4

1 5 Kanbur (1990) and Ali (1990) show in a stylizedfashion how project economic analytics can beapplied to analyzing policy reforms, referring to pricepolicy reforms. In fact, in the literature of publiceconomics, any policy or project variable can bechosen as endogenous to see the change in welfarelevel with respect to a marginal change in the chosenvariable within the same analytical model.

A. Introduction

The starting point for any assessmentof policy change is the status quo orthe “without program” situation. Ideally,much of the framework and data for thisassessment should have been collectedas part of sector work and relatedanalysis. With the status quo as thereference point, conceptually, reformmeasures are similar to project activitiesin an investment project (Squire 1989).In addition, reform measures lead to anoutput, such as in a new law, in supportof desired outcomes. So, the underlyinginput-output relationship becomes theconceptual framework for identifyingand valuing the benefit and cost streamsof policy operations, as in projects.13

OED’s Special Evaluation Study onProgram Lending (ADB 2001,paragraph 77) notes: “Program designseems to begin with the proposedreforms, derived from internationalpractice, with sector analysis andexpected benefits as justification.” Thisfinding calls for efforts to refine andfurther develop the conceptual andanalytical framework for undertaking ex-ante economic analysis of policy andinstitutional reforms.

Kanbur (1990) notes, in reference toWorld Bank experience, that manycommentators have claimed that it isdifficult to carry out economic analysisfor policy operations.14 The same isoften heard with regard to ADB program

loans. However, such claims arequestionable because when proposing apolicy reform measure as part of theprogram design, an opinion has alreadybeen expressed—however implicitly—about the expected linkage between thereform measures and their benefits. Forexample, a road investment projectproposal to rehabilitate a length ofhighway implies an underlyingassessment of the relative priority andthe incremental benefits that such aninvestment should generate, includinga justification for the investment as ameans of allocating scarce resourcestoward some welfare objective.

Seen in this light, the problem is notabout whether it is possible to performeconomic analysis on policy reforms.It centers on how such analysis can bedone, or the issue of the analyticalframework and specific methods to beused. To make effective use of a definedanalytical approach involves setting upvalid and verifiable assumptions. In thecontext of project design, the economicanalysis entails developing valid andverifiable assumptions of certain inputand output relationships relevant to aproject. In the context of policy reform,formulating valid and verifiableassumptions is an even more criticalpart of the economic analysis.15 Thisendorses the need for solid sectoranalysis that provides the basis to

1 3 Chapter 7.D further discusses this relationship, andFigure 8 offers a schematic of it.

1 4 See also Ali (1990).


ASSESSING THE EFFECTS OF POLICY CHANGE

CH

AP

TER 4

validate the assumptions that underlieeffective reform. A further analyticalchallenge is to apply the framework inassessing institutional reforms.

B. Assessing the Effects of Policy Change

Once a policy problem is identifiedthrough diagnostic sector work, the nextstep is to determine alternativesolutions and predict their possibleoutcomes. Ideally, policy options shouldbe subject to simulations of possibleoutcomes in order to assess theirpossible impact on, for example, sectorperformance. Such simulations wouldalso assist governments andstakeholders in determining alternativepaths and approaches to improvingperformance and selecting from thesepaths. Given the possible mixes ofreforms and investments, several paths

can be taken. Figure 3 shows thisconceptually, with t0 representing thecurrent sector performance and tnperformance after implementation. Thetask of policy analysts is to assess andpresent viable options to governmentsand stakeholders to facilitate selectionof the least-cost and best sequenced mix.

Short-run negative consequences canarise as part of the policy reform process.For example, expenditure reductionprograms may constrain governmentexpenditure and the delivery of services,or resource immobility may hamperefficiency gains from price liberalizationin the short run. A priori, a range ofreform measures could have negativeshort-run consequences for certainstakeholders in the absence of explicitmitigatory interventions in the form ofcompensatory mechanisms and social


FIGURE 3: Different Reform Scenarios to Meet a Policy Goal

42

CHAPTER 4

safety nets. Despite the potentiallybeneficial longer-term consequences ofthese reform measures for even short-term“losers,” the short-run negativeconsequences cannot be neglected. Thiswas one of the early lessons of structuraladjustment lending in the 1980s and1990s. It retains its relevance today withrecent reforms that increasingly focuson institutions. Figures 4 and 5 depictprofiles in terms of without adjustment,with adjustment, and the “no shock”situation based on a five-sectorcomputable general equilibrium modelfor Ecuador in the late 1980s (de Janvry,Fargeix, and Sadoulet [1990] reportedin Kanbur [1990]).

To the extent possible, policy analysis,including simulations, should be carriedout to identify the short- and medium-run effects. However, ex-anteassessments of policy options are oftenbased on hypotheses drawn from a priorireasoning that should, in principle, besubjected to empirical verification, ratherthan be accepted as statements of fact.Given the nature of some reformmeasures, this can be challenging. Forexample, improving the “enablingenvironment” is often used as anargument for policy reform. Among theADB program loan reports reviewed(Appendix 1), a strong case is usuallymade for reforms and often an intuitive“with-without” comparison can beinstructive. For example, the Kazakhstanpensions system must be reformed toavoid collapse; the Federated States ofMicronesia (FSM) must rapidly developthe private sector to compensate for theshrinkage of the public sector asCompact of Free Association fundingis reduced; financially distressed SOEs

in Viet Nam need urgent restructuringand, where possible, privatization; andfinancial sector development ineconomies, such as Mongolia and VietNam is a necessary, even if not sufficient,condition for accelerated and sustainedeconomic growth and poverty reduction.Where a priori reasoning cannot beverified and intuition must be reliedupon to facilitate close monitoring, theanalytical limitations should beacknowledged, and the assumptions onprobable response and expectedoutcomes should be stated.

C. Approaches to Analysis

Sector-wide and loan-specific studiescan potentially employ a full range ofmacro-, meso-, and micro-analysistechniques that could shed light on thepossible effects of a given policy change.As described in Evans (1999), analysiscan range from economic modeling andfull quantitative analysis, to historicaland qualitative analysis, to cross-countrycomparisons and direct countryestimates, and the “checkable story”fallback, including rapid appraisaltechniques. The limitations of theseprogressively less rigorous techniquesare discussed in Evans (1999).16 Forexample, where analysis of feedbackeffects is desirable, general equilibriumanalysis is appropriate. Comparative staticanalysis, using partial equilibriumtechniques, allows analysts to see theeffects of, for example, a price change(including taxes and subsidies) or quotachange on supply and demand of acommodity. Examination of descriptivestatistics and ad hoc calculations provideinsights into available secondary data, butamount to a qualitative assessment.



CH

AP

TER 4

Source: de Janvry, Fargeix, and Sadoulet. 1990. The Political Economy of Stabilization Programs: Growth, Welfare, and Sustainability.Department of Economics, University of California, Berkeley. Reported in R. Kanbur. 1990. Projects versus Policy Reform. Paperpresented at the World Bank Annual Conference on Development Economics, Washington, DC.

FIGURE 4: Alternative Real GDP Scenarios

FIGURE 5: Alternative Rural Poverty ScenariosPercentage of RuralPoor inUrbanPopulation

Index of Real GDP

44

CHAPTER 4

The World Bank (2002b) provides auseful framework to help the analyst inselecting the appropriate approach andtool for analyzing the effects of changes.The framework considers two keydimensions: the importance of feedbackeffects in the reform measures, and dataavailability and analytical capacity(Table 6). In cases where the policyreform is likely to lead to high feedback,with significant multiplier effectstransmitted through a number ofchannels and markets lagged over time,the effects will need to be estimated.Examples of reforms with high feedbackeffects include monetary and fiscalpolicy changes that affect inflation,interest rates, balance of payments andreserves, fiscal deficits, trade andexchange rates (removal of tariffs andexchange rate policy), and financial

sector reforms. Reforms that are lesslikely to have feedback effects includenarrowly focused public sectorinstitutional reforms, such as civilservice reforms, land reforms,privatization of SOEs, and labor marketreforms.

The main messages emerging fromTable 6 are as follows:• Reliance on qualitative assessments

results in non-parametric analysis anddoes not permit order of magnitudeimpact assessment projections. Thegreater the quantitative and parametricanalysis, the greater the possibility ofidentifying the magnitude ofoutcomes and predictions.

• As a minimum, descriptive statisticalanalysis is required to provide anorder of magnitude assessment of the

1 6 A summary of Evans’ suggestions is also providedin Bolt and Fujimura (2002).

TABLE 6: Selection of Techniques for Policy Change Impact Analysis

• Descriptive statisticalanalysis and economic andsocial characterization

• Qualitative impactassessments with specifiedassumptions

• Use of secondary sources ofanalysis with clearassumptions

• Social accountingmatrixes—input/outputanalysis

• Computable generalequilibrium

• Multi-market analysis• Reduced form input/output

analysis and simulationmodels

• Survey-based householdmodels

• Full econometric demandand supply analysis

• Fiscal impact and publicexpenditure analysis

• Comparative institutionalanalysis/transaction costanalysis

• Qualitative impactassessments of inter-temporal and distributioneffects

• Descriptive statisticalanalysis

• Partial equilibrium analysisof prices, supply, anddemand parameters

• Budget analysis• Benefit incidence analysis• Partial equilibrium analysis

of price policy• Affordability and

willingness-to-pay analysis

Acc

om

mo

dat

ion

of

Fee

db

ack

Eff

ects

LowData and Analytical Requirement

HighMedium

Source: Adapted from “A User’s Guide to Poverty and Social Impact Analysis.” Draft. 2002. World Bank Poverty ReductionGroup and Social Development Department, 19 April 2002.

L

o

w

Hig

h



CH

AP

TER 4

economic and social situation as astarting point for assessing possibleimpacts.

• Where qualitative assessments anda priori reasoning are to be reliedupon, then the limits of the analysisand the underlying assumptionsshould be clearly stated.

• The initial descriptive work and thenature of the policy change willprovide guidance as to whether therewill likely be significant feedbackeffects that will guide furtheranalyses, especially the sufficiency forpartial equilibrium analysis or theneed for general equilibrium analysis.

• Increasingly rigorous analyticaltechniques require greater use of data,resources (analyst time and fundingfor field and office work), andcapacity.

The rest of this chapter outlines the basicfeatures of these analytical techniquesand limitations, and references to theirapplication in ADB policy operations.

1. Descriptive Statistical Analysis(Cases 2 and 3 in Appendix 3)

Where neither general nor partialequilibrium modeling is possible, thenreform impacts will have to be assessedwith simpler descriptive statisticalanalysis and ad hoc calculations andmethods for possible effects prediction.These could involve, for example,estimates of the price effects ofparticular interventions and estimates oftheir impacts on consumers andproducers on the basis of simpledemand and supply elasticities. Suchelasticities can be derived econo-metrically, but in simpler calculations,

they can be taken from other studies orsimply assumed.

Case 2 in Appendix 3 (PhilippinesPower Sector Restructuring Program)illustrates the approach to estimatingchanges in economic subsidies accruingto a particular consumer group such asthe poor. This approach is also usefulfor assessing the impact of social sectorprograms, as in health or education,which involve the introduction of, orincrease in, user charges, or theexpansion of private sector provision.Another useful exercise is a budgetimpact analysis. This has little economicmodeling behind it, but is easy to applyto a package of fiscal reform measuresthat are a usual feature of sector reforms.Case 3 in Appendix 3 provides anapplication to the Uzbekistan EducationSector Development Program.17

2. Partial Equilibrium Analysis:Market and Price Analysis (Cases 1and 4 in Appendix 3)

Partial equilibrium modeling is a moreformal analytical technique where onemarket or sector is looked at alone,without allowance for feedback to otherparts of the economy. “Static” analysiseither assumes low feedback effects orignores effects in other markets andindustries, so is appropriate in analyzingchanges brought about by price policychanges that have limited impacts oneconomy-wide aggregates. Applicationsinclude, for example, enterprise budgeteffects and competitiveness issues, aswell as situations of low data availability

1 7 The linkage between policy change and fiscal impactis further discussed in Chapter 6.

46

CHAPTER 4

and capacity. Case 1 in Appendix 3provides an application of a simplepartial equilibrium analysis to the FSM,which was found to be practical duringADB policy dialogue with governmentofficials on the appropriateness ofsubsidizing efforts to establish apoultry industry.

Another illustration is provided in thesupporting analysis for the PhilippinesGrains Sector Development Program(DAI Consultants 1998). This programhas several policy reform menus but theelement modeled covers the reform ofthe rice market and pricing policy of theNational Food Authority. Case 4 inAppendix 3 gives the structure of themodel and shows how it was used toassess the effect of a policy change.Using the same reform program, furtherextension of the rice market analysis toan agency cost analysis was undertakenby combining the organizational costsand market inefficiency costs. Partialequilibrium modeling of this type isuseful in situations in which key pricesrelevant to the assessment of reformimpact are best treated as endogenouslydetermined, rather than as exogenousparameters. For example, in the GrainsSector Development Program, the priceof paddy received by farmers and theprice of rice paid by consumers are criticalvariables. The model allows prices to bedetermined by domestic demand andsupply conditions, allowing anassessment of how they vary with theassumed impact of reforms.

However, in an ex-post examination ofthe costs and benefits of price policyreforms based on a case study in the

cashew nut sector in Mozambique(McMillan, Rodrik, and Horn Welch2002), caution is suggested on basingprice policy reform entirely on ananalysis of the likely response to supplyfrom price liberalization. This isespecially the case where “a priorigeneralizations” are relied upon as thebasis for assessing the expectedefficiency gains and what eventuallyproved to be overstated benefits. Itbecame apparent that consideration ofthe Government’s level of commitmentto the reforms and identification ofthe level of political opposition wereboth inadequate. There had been afailure to consider compensation forredundant processing laborers; tointroduce a “credibility enhancingmechanism” for the reforms; and togenerate public awareness. This failureled to low expectations by processorsand their laborers, traders, and farmersthat the reforms would hold.Processing laborers did not find newemployment because they anticipatedthat the Government would backtrackon the reforms, and traders captured agood part of the price benefitsintended for poor cashew farmers.Clearly then, in cases where keystakeholders’ expectations are notsufficiently managed, there is a highrisk that the assumed market responsewill not occur. The three main lessonsthat emerge from this case are:expectations management is needed; thereform was as much a political problemas a technical one; and the use ofconditions to push through reformswas ineffective given the lack of reformownership by the Government.



CH

AP

TER 4

1 8 For further discussion, see for example, http://www.encycogov.com, 2001, “Decomposing Cost intoTransaction Costs and Production Cost”; andKherallah et al. undated.

3. Partial Equilibrium Analysis:Comparative Institutional Analysis(Case 4 in Appendix 3)

A strand of the emerging institutionaleconomics, the transaction-costframework of analysis, is at a nascentstage but has been gaining interest as itoffers an additional perspective tomicroeconomic analysis. Transactioncosts are variously defined as the costsof making exchange or as indirectproduction expenses, as opposed to thecosts of production or transforminginputs into outputs or directproduction expenses.18

While the standard neoclassicaleconomic analysis rests primarily on theprice mechanism in allocating resources,transaction-cost analysis proceeds fromthe premise that the price mechanism isnot always efficient and that a differentmechanism functions within the firm.This logic dates back to Coase’s (1937)explanation that when a firm internalizesits costs, the internalization overridesthe price mechanism due to thetransaction costs incurred. This impliesthat the boundaries, or “frontiers” of afirm depend not only on the productivetechnology, but also on the “costs ofdoing” business (Klein 1999).

According to the transaction cost schoolof thought, institutions that evolve tolower these costs are key to theperformance of economies. Whentransaction costs are absent, the initial

assignment of property rights does notmatter from the point of view ofefficiency because rights can bevoluntarily adjusted and exchanged topromote increased production. Butwhen transaction costs are substantial,as is usually the case, the allocation ofproperty rights is essential.

A limitation still exists on the extent towhich comparative institutional analysiscan be appropriately quantified in viewof the difficulty in measuringtransaction costs and in assigningempirical proxies for key variables,especially uncertainty (Klein 1999). So,empirically, institutional analysisdepends mostly on qualitative casestudies. An illustration of comparativeanalysis for each type of governancestructure or institutional arrangement isgiven in the Agency Cost Approach inCase 4, Appendix 3.

4. Applied General EquilibriumModeling (Case 5 in Appendix 3)19

The most comprehensive andsophisticated way to assess povertyimpacts of policy change, including thecapture of feedback and intertemporaleconomic effects, involves using eithera social accounting matrix (SAM) or,in a more dynamic form, a computablegeneral equilibrium (CGE) model basedon a SAM. A CGE model is used toprovide an evaluation of the effects ofbroad policy changes includingexogenous shocks, economic policy

1 9 Case 5 in Appendix 3 is preceded by an expandeddiscussion of the principles of computable generalequilibrium analysis.

48

CHAPTER 4

changes, and changes in the domesticeconomic and social structure. A SAMmodel is a square matrix with columnsfor expenditure and rows coveringincome accounts. It combines input-output data with the national accountsto reflect the circular flow of income ata particular point in time. In thiscontext, its key use is as a means ofassessing the direct and indirect incomeeffects of a particular exogenousimpact—such as a policy change leadingto different expenditure patterns.20

Since SAMs can be constructed withdifferent groupings of households, theycan be used to assess how the poorerhouseholds are affected. SAMs aretypically static systems with fixedcoefficients and prices. However, theycan be extended to more dynamic modelsby the incorporation of behavioralequations relating to the major marketsin an economy. These, combined withkey parameters relating, for example, tovarious demand and supply elasticities,can be developed into a full macro model.Such models normally assumeoptimizing behavior, with flexible pricesclearing all markets. The advantage ofthese models is that they can bedesigned to incorporate features ofindividual economies and can be run fordifferent policy simulations. Suchmodels are conceptually the onlyrigorous means of assessing thecounterfactual—what would have takenplace in an economy without a particularpolicy reform. Versions of such modelshave been used in a number of countriesto assess the poverty impact of policy

change. In Asia, these countries includeIndonesia (Thorbecke 1991), andMalaysia (Demery and Demery 1991).

A combined use of CGE and SAM isconsidered a state-of-the-art methodologyfor ex-ante assessment of policychange.21 However, the practical relevanceof such methodology remains a subjectof considerable debate. Even relativelysophisticated models based on the datafrom a large and recent SAM must, ofnecessity, be a major simplification ofreality. Key parameters will often beassumed or simply taken from work onother economies. Also the structure,such as the “macro-closure rule” andfunctional forms of productionaggregation, used for the models canitself influence their results.

For example, de Maio et al. (1999) andthe reply by Sahn et al. (1999) debatethe usefulness and accuracy of CGEmodels in assessing the relationshipbetween poverty and adjustment inAfrica. In addition, such models cannormally cope with the impact ofdiscrete policy changes such as adevaluation, a fall in aggregategovernment expenditure, a change ininterest rates, or the removal of a

2 0 For recent application to policy analysis, see Iqbarand Siddiqui (1999) and Khan (1999).

2 1 Dervis et al. (1982) and Robinson (1989) providesurveys for CGE models that follow the assumptionof imperfect substitutability between imports andexports—the Armington assumption. A detailedintroduction to the structure of SAMs and CGE modelsis given by Sadoulet and de Janvry (1995), chapters10 to 12. Scarf and Shoven (1984) and Shovenand Whalley (1992) provide a survey of models ofa more pure trade nature (perfect substitutabilitybetween imports and exports of a homogeneouscharacter). They also provide the workings ofbenchmark equilibrium calibration and solutionalgorithms.



CH

AP

TER

4

particular subsidy. However, the effectsof more general reform measures, suchas strengthening of the financial sector,institutional change in the civil service,or a new road-building program, may bedifficult to express in exact quantitativeterms and hence, difficult to incorporatein the models.

Finally, and perhaps crucially in thecontext of ADB operations, accuratemodeling of this type is demanding indata, skills, and research resources. Aneffort by the World Bank—theIntegrated Macroeconomic Model forPoverty Analysis (IMMPA) Project—is an attempt to combine CGE and SAMfurther with a financial programmingmodel and household survey data inorder to measure distribution impactsof various macroeconomic policies(World Bank 2001a).

A short-cut approach to country-specificCGE modeling is to establish “generic”models for particular types ofeconomy—for example, one could be fora low-income labor-surplus economy asfound in South Asia or the poorer partsof East Asia and another for transitionaleconomies. Such models could be basedon “typical” data of an assumed nature

and used to assess the possibleconsequences of different interventionsfor poverty. Their results can be no morethan suggestive since the modelsthemselves are not based on anindividual economy, but they can givepolicy makers an insight into possiblepoverty outcomes under differentscenarios and assumptions. Thorbecke(2001) describes the construction ofsuch an archetypal model for an Africaneconomy, as well as the extension of hisearlier work in Indonesia, to model thepost-Asian financial crisis trend inpoverty and distribution. Further, hehighlights one of the main uncertaintiesin this type of work: how distributionwithin particular income groups changesin response to external economicshocks. Potentially useful as suchmodels are, they are best seen as part ofresearch initiatives that can provide thebackground to discussions ofoperational issues of policy operations.

Table 7 summarizes the differentapproaches to poverty impact assessmentof policy changes and their limitations,and gives examples of ADB-supportedprograms that either have been or couldhave been assessed using theseapproaches.

50

CHAPTER 4

TABLE 7: Approaches to Assessment of Policy Operations

Approach Advantages Disadvantages Possible Application

Descriptive • Simple to apply • Requires specific • Power restructuring e.g.,statistical • Minor data and time parameters such Loan 1662-PHI.analysis and requirements as demand and • Health reform e.g., Loanad hoc supply elasticities 1568-MONcalculations • Captures only one • Education e.g., Uzbekistan

dimension Education SectorDevelopmentProgram

Partial • Fewer data required than • Cannot capture • Self-contained priceequilibrium computable general computable general changes such asmodeling: equilibrium modeling equilibrium modeling agriculture subsidyMarket and • More workable and easier feedback effects e.g., Loan 1739-PHIprice analysis to interpret results • Not useful for complex

qualitatively reforms

Partial • Considers the costs of • Limited in ex-ante • Transaction cost approachequilibrium institutional transactions analysis because of is relevant in analyzing themodeling: that can be key to the difficulty in rural economy where smallComparative understanding incentives estimating transaction farmers and traders facinginstitutional and efficiency costs high transaction costsanalysis improvements resulting in thin markets,

market failure in theprovision of credit, inputs,and services, andincomplete or imperfectland and labor marketse.g., Loan 1739-PHI

Applied • Can simulate • Considerable data and • Trade reform e.g., Loancomputable counterfactuals time requirements 1680-PAKgeneral • Useful for impact of price • Results are sensitive to • Financial sector reformequilibrium changes like devaluation assumptions and model e.g., Loan 1485-VIEmodeling and interest rate rises specifications


CHAPTER

ASSESSINGTHE POVERTYIMPACT OF POLICYCHANGE

l Mechanisms for Poverty Reduction

l Current Poverty Impact AssessmentPractice at ADB

l Review of Current ADB Practice

l Modifying the PIA Matrix

l Operational Considerations

52

CHAPTER 5

A. Introduction

The discussion so far emphasizes that,depending on the nature of the policyreform, the effects on poverty reductioninvolves interlinking macro-meso-microcausal mechanisms and, importantly,political economy factors. In terms ofthese factors, vested interests who preferto maintain the status quo can be keyobstacles to reform and to thesubsequent realization of possiblepoverty reducing effects. In fact, in adeteriorating economic situation, theopportunity cost of not reforming maybe a decline in growth rates and aworsening incidence of poverty.

Chapter 3 discussed the importance ofissues such as incentives and greateraccess to opportunity throughinstitutions and markets. Whether the“win-win” situation of growth andgreater market opportunities, forexample, is a feasible outcome for poorgroups, depends on the conditions anddevelopment path specific to theeconomy in question. Further, thepoverty reduction impact of a given rateof economic growth will vary with theform that growth takes, since the varioussector patterns of economic growth havedifferent poverty consequences.

B. Mechanisms for Poverty Reduction

Reform has various mechanismsthrough which it can reduce poverty.First is the effect of reform onemployment creation. If economicactivity responds positively to reform,the poor can find their wage employmentincreasing. It is this link that has led tothe focus on labor-intensive growth in

discussions of poverty reduction(World Bank 1990). However, in termsof employment creation, even successfulpolicy reform could lead to negativeshort-run employment effects. The mostobvious example is the curtailment ofimport-substitution activities as a resultof a more outward-looking export-oriented strategy. In the short run, thenumber of people below the poverty linemay even rise, not fall, and employmentreallocation, where needed, may taketime. So, the transitional poverty effectscannot be disregarded.

A second mechanism is macroeconomicstabilization. It is often argued that whileinflation is a tax on the whole of society,it falls most heavily on the poor, whosemonetary incomes are not just low, butrelatively inflexible. A significant sourceof income for the poorest is likely to becasual employment or the monetaryequivalent of government-providedservices, both of which may rise sluggishlyat a time of accelerating inflation. Hence,the implication is that if inflation can bestabilized through fiscal and monetarypolicies, this loss of real income can beavoided. Also, in principle, any real declinein government expenditures on services,as part of the fiscal retrenchment requiredfor macro stabilization, can have seriousshort-run consequences for the poor, orsome measures to introduce user chargesdo not make adequate exemptions for thepoor. However, in cases where the poorplace little or no value on these services,fiscal reform may negatively affect thebetter-off to a greater extent and it mayactually benefit the poor in cases wherepublic expenditure reduction or switchingmay realign resources more efficiently.


ASSESSING THE POVERTY IMPACT OF POLICY CHANGE

CH

AP

TER 5

A third and potentially complex area inwhich reform can interact with povertyis through relative price shifts. Chapter3.D described, for example, how pricepolicy changes can favor producers butnegatively affect consumers. In mostmacro adjustment or stabilizationprograms, two important relative priceshifts can be expected. One is a rise inprices of goods traded relative tonontraded, which occurs when realexchange rate depreciation is requiredto, for example, remove internal andexternal imbalances. The other is withinthe nontraded sector where a rise in pricesfor publicly supplied services may beneeded to reduce the fiscal burden ofservice delivery. How the poor will beaffected by such shifts will depend upontheir situation as net producers orconsumers of the goods and servicesconcerned. However, these relative priceshifts will occur at a time when realincomes are also changing and so the fullimpact on the poor will be determinedby the net effect, allowing for bothincome and price changes. Even in theshort run, this can be difficult to predict.

Finally, in recent years, reforms havefocused on various aspects ofinstitutional changes, since it isrecognized that market relations operatewith widely different levels of efficiencyin various institutional contexts. Suchinstitutional changes have beenwidespread in some economies, coveringenterprise restructuring and privatization,financial sector reform, civil servicedownsizing and reorganization, andvarious forms of change to social servicedelivery mechanisms and organization.These reforms are intended to strengthenthe growth impact of more conventional

macroeconomic adjustments—and to thedegree that they succeed, they shouldhave some positive impact on povertyreduction.22 However it is far from clearhow such changes will impact on the poorin the short run. For example, whilefinancial sector reform is widely seen asessential for sustained economic growth,it is not immediately obvious whetherand how soon the commercialization ofpublicly owned development banks willaffect the poor positively.

Overall, economies developing andreforming in a dynamic environmentoften face limits to identifying andtracing, ex ante, where the impactsassociated with an economy- or sector-wide policy loan start and end, and theirspecific poverty impacts. The same canbe said for projects, although usually toa lesser extent because of their greaterfocus and clearer specified boundary.The World Bank (2001c) considers thatexogenous influences on operations thataddress poverty—as well as the indirecteffects on poverty of operations thataddress the underlying (systemic)constraints to growth, social development,and improved governance—can be greaterthan the direct effects of targetedinterventions. Furthermore, as all ADBoperations are required to show theirdirect and indirect effects on poverty, itis essential to identify and understandexogenous and endogenous macro- andsector-wide influences.

In such a way, poverty assessment at thecountry and sector levels may often be

2 2 See Aron (2000) for a discussion of the evidenceon institutional change and economic growth.

54

CHAPTER 5

more appropriate than at the project orprogram level, especially where thecauses and effects of poverty can onlybe considered meaningfully at the macroand sector levels rather than at theproject or program level. In a similar waythat the costs and benefits of a projectare shared among different groups suchas consumers, producers, andgovernments (ADB 1987), reforms andpolicy changes may have distribution, aswell as efficiency, implications that needto be understood as part of thedecision-making process. So, the analystmust not only comprehend the net effectof change, but also understand whogains and who loses from reforms.

C. Current Poverty Impact AssessmentPractice at ADB

Recently, distribution analysis of policyreforms at ADB has focusedsubstantially on poverty impactimplications. Operational work atinternational aid agencies can apply themost appropriate approaches amongthose discussed before in Chapter 4,depending on the resources and skillsavailable. For current practice at ADB, auniform minimum requirement forpolicy operations preparation is apoverty impact assessment (PIA) matrix(discussed below). Rather than aquantitative tool, such as thosediscussed in the previous chapter, it is asimple but pragmatic way of establishingthe economic logic that assesses theimpact of proposed policy changes. ADBhas applied the PIA matrix to its policyoperations since 1995 (ADB 1995).The matrix was reexamined in asubsequent study and given a generallyfavorable review (Nelson 1998).

The logic of the matrix is that the welfareof a poor household is affected throughfour key channels: (i) the labor earningsof household members in the workforce;(ii) prices that they face in selling andpurchasing goods; (iii) access to andthe return on the nonlabor assets of thehousehold; and (iv) their net receiptsof public and private transfers. A similarapproach is discussed in a more formalway in Behrman (1993). Theimportance of these channels variesbetween different poor groups and anyframework for poverty impact assessmentmust be sufficiently broad to cover arange of possible scenarios.

The PIA matrix involves two dimensionsor axes: one relates to the channelsmentioned in the previous paragraph,and the other to the timing and degreeof impact. The channels, specified onthe rows, are labor market, prices, accessof the poor to non-labor assets, and nettransfers. The impacts, on the columns,are subdivided into direct effects, indirecteffects, macro effects, and impact on thenonpoor. Table 8 provides the basiclayout of ADB’s current PIA matrix.

This framework was never intended tobe more than a means of organizingthoughts on how particular policychanges might impact on the poor.Nevertheless, its usefulness is settingout clearly the underlying assumptionsrequired to achieve the envisagedoutcomes. In this sense, the nature ofthe matrix is similar to the programlogical framework that is now widelyused by funding agencies. The advantageof the program logical framework is thatit sets out objectives a project is toachieve and the assumptions that must



CH

AP

TER 5

hold for these objectives to be met. Therealism of the critical assumptions mustthen be tested, and if necessary,measures applied to ensure that, as faras practicable, reality matches theseassumptions. An analogous procedureis adopted in the PIA matrix. However,the longer the chain of assumptionsrequired, the less likely the envisagedpoverty impact will be attributable tospecific reform measures.

D. Review of Current ADB Practice

Review of ADB’s core program loanreports revealed several areas of concernabout the way the PIA matrix has beenapplied. The use of the matrix in thesereports is largely qualitative. The cellsin the matrix are either completed witha narrative or more simply with a positiveor negative sign to indicate direction ofimpact on the income of the poor, orleft blank where inapplicable. Several ofthe program loan reports adopt the same

judgment on short-run impacts ofinterventions on the poor. For example,the Pakistan Trade, Export Promotionand Industry Program loan reportrecognized negative short-runemployment effects for the poor fromtariff reform and privatization, as wellas higher prices for the poor as a resultof privatization, although in the longrun, the report argues that these will beoffset by improved allocative efficiencyand rising exports. Several programloan reports also acknowledge thathigher user charges due to market-oriented reform programs causenegative effects on poor consumers,such as the Mongolia Health SectorDevelopment Program. In suchinstances, however, the reports arguethat longer-term improvements inefficiency will benefit the poor as servicepurchasers. Similarly, liberalization ofagricultural trade and the breakup ofstate farms are recognized as causing aloss of agricultural jobs among the poor

TABLE 8: ADB’s Current Poverty Impact Assessment Matrix

Channels\Impacts

Labor Market

Prices

Access for Poor

Transfers

Net Effects

Narrative

Source: Asian Development Bank. 1995. Staff instruction on Poverty Impact Assessment. Strategy and Policy Office. Manila.

Type of Effects

Direct Indirect Macro Nonpoor

56

CHAPTER 5

in the short run, but higher long-runproducer prices are assumed to boosttheir incomes.

For reasons discussed above, there arecases where policy changes will impactnegatively on the poor in the short run.How the poverty situation improves overtime will depend on a number of factors:principally, the effectiveness of safetynets; the supply response of varioussectors to the new set of price and otherincentives created by reform; the labor-intensity of expanding sectors; and thegeneral fiscal position. Even where theintervention in question is assumed tobe growth enhancing in the longer term,it is far too simplistic to assert that italways has net positive impacts overall,especially when there are large tangibleshort-run costs. Furthermore, inprinciple, the short-run effects havehigher values than the medium- andlong-run effects due to discounting,thereby complicating the intertemporalaggregation of the net benefitsassociated with reform programs. Forexample, privatization or tradeliberalization measures may be growthenhancing in the longer term, but howfar the poor share in this growth andhow long they will have to wait to receiveany of the benefits is unclear. In anumber of program loan reports,assertions about potential long-runeffects are made in place of detailedanalysis. Modeling in either the full orpartial versions (discussed in Chapter4.C) can provide some of the answersto test such assertions.

Another limitation with the current useof the PIA matrix is the over-simplisticuse of a poor/nonpoor distinction. If a

poverty focus is to be a major elementin the analysis, it is preferable todisaggregate the “poor” in some way,because different “poor” groups may beaffected in various ways by particularinterventions. For example, thecircumstances and problems of the urbanand rural poor are very different. Withinthe rural group, there are usuallysmallholders with land and those whoare landless. Further, within thelandholding group, there are those whogrow cash crops, export crops, andimport-competing crops. Thesedistinctions need to be considered whencompleting the matrix to ensure that thecontext is clear.

The key point is that it would be veryrare for “poor” groups to be affecteduniformly by a policy change.Privatization may cause job losses tothose with formal urban employment, butmay have little impact on the rural poor(or may benefit them as consumers).Financial sector reform that leads togreater availability of credit, at interestrates below those in the informal creditsector, may benefit poor rural borrowerswith some land to offer as collateral, butnot borrowers who have no land to offer.Similarly, agriculture sector reform thatliberalizes marketing, and as aconsequence, raises prices to farmers nearworld market levels, may benefit ruralpoor farmers of export crops but mayhurt the urban poor and farmers ofnontraded crops. Also, since there can bewide variations in standards of socialsector provision among a country’sregions—the poor in one region mayreceive a different level of health oreducation from the poor in anotherregion—there would be no standard



CH

AP

TER 5

way of decomposing the poor as acategory. Depending on the depth ofthe socioeconomic assessment on, forexample, the segment of the poor thatis likely to be affected most directly,such information can augment theexplanation in the PIA matrix.

A further limitation of the current useof the PIA matrix is the ambiguity ofthe result in terms of poverty targets.Even where targets are set out in theprogram logical framework, these are fortotal population served rather than thepoor explicitly. A recent exception is theUzbekistan Education SectorDevelopment Program, which has anexplicit poverty loan component and,therefore, targets regions and schoolsin poor areas. Part of the reason for thisis that many policy-based loans aredesigned primarily to promote improvedefficiency of the sector as a whole ratherthan targeting the poverty reduction ofspecific groups. Other reasons are thatthe complex causal mechanisms throughwhich reform measures impact on thepoor and the need for data-intensive ex-ante empirical analysis make theestimation costly and uncertain.Consequently, at best, a priori reasoningis relied upon.

E. Modifying the PIA Matrix

1. Key Refinements

As a simple framework, the current PIAmatrix can be modified in various waysto articulate the likely or possiblepoverty impacts of policy changes.23 Amodified PIA matrix is laid out in Table9. The first focus should be on thechannels through which the poor are

affected and the distributionconsiderations, bearing in mind thatspecific reform measures will unlikelywork through all channels and therelevant main and specific channelsshould be identified. The second focusshould be on the intertemporalconsiderations of the impacts on thepoor and other stakeholder groups ofthe intervention. The third focus shouldbe on the degree to which the reformimpact is direct or indirect. The fourthfocus should be the mitigation andenhancement measures to reduce thenegative effects on the poor or enhancetheir inclusion as appropriate. Thesemeasures, as with the reformsthemselves, are a design rather than ananalytical issue. These impacts can bedescribed qualitatively, but it ispreferable for these to be quantified, forexample, in terms of losses/gains in jobsand incomes. Any assumptions,especially when the assessment is madequalitatively, that are critical for thesemechanisms to work or for the analysisto be valid should be listed in the PIAmatrix.

2. Channels of Effect (Rows)

Channels of effect will be differentdepending on the nature of the reformmeasure, and should be identified forindividual reform measures. Severalpoints are worth clarifying concerningways in which reforms may work throughrespective channels of effect. First, inrelation to labor markets and wages (row

2 3 “Likely” when empirical evidence is used, or“possible” in the case of a priori reasoning.

58

CHAPTER 5

1), it is often helpful to distinguishbetween formal employment and jobs inthe informal sector, since policy changecan affect these differently. Much of thecriticism of public sector reformprograms is that they can replace the firsttype of employment with the second. Tomore fully assess whether the effects arepositive or negative, a fuller account ofthe structure of the labor market needsto be considered, including theexistence of labor market rigidities andfragmentation and new and transientemployment opportunities that mayarise as a result of structural change.Such an account includes assessing theflexibility and absorptive capacity of thelabor market as a whole and wage effects

on different segments of the labormarket.

Second, it is necessary to distinguishbetween relative price effects of goodsand services that exert different impactson the poor, depending on whether thepoor are market suppliers or consumers.Lifting controls on commodity prices,for example, may benefit poor farmerswho are producers, but may hurt thepoor who are net consumers of thegoods concerned. In addition, it isimportant to analyze the impacts on thepoor of aggregate price changes thatfollow policy reforms. A refinement thathelps remove the potential ambiguity inprice changes is to decompose the price

Net Impact

Information Basis or Crucial Assumptions

TABLE 9: Modified Poverty Impact Assessment Matrix

Possible Channels ofEffect

Effect on the Poor Effects onOther

Stakeholders

Mitigation orEnhancement

MeasuresExample ofspecific channels

DirectShort Run

IndirectShort Run

IndirectMedium Run

Accessto LaborMarketsandWagesAccesstoMarketsandPrices

Accessto Assets

ServiceAccess

DirectTransfers

FormalInformal

As output consumerAs output supplierAs input consumerAs input supplier

PhysicalFinancialSocialHumanNatural

NonmarketedPublic services

Private transferPublic transfer



CH

AP

TER 5

mechanism category into threesubgroups: (i) relative price changes forsuppliers and producers; (ii) relativeprice changes for consumers; and(iii) relative price changes betweensectors such as agriculture, industryand services. Prices can include expectedpolicy reform-induced macro-levelvariables such as interest rates. A furtheraspect involves institution-relatedissues of market access andperformance, as discussed in Chapter3.E, that the policy reform may address.Under this heading fall, for example,reforms that affect market access andequal opportunity throughinformation dissemination andpractices that, in turn, influence marketentry, transaction costs, and regulationsaffecting competition. However,depending on the nature of the policyreform, quantification of the likelyeffects on market volume and pricesarising from market institution changescan be difficult, and qualitativeassessments may need to be used.

Third, assessment of policy reform-induced changes in access by the poorto assets is needed, as lack of assets isan important dimension of poverty. Agrouping of assets, in terms of physical,financial, human, social, and naturaldimensions, is suggested (World Bank2002b). To assess this channel of effect,an adequate account is needed ofdifferent types of assets and how issuessuch as property rights and contractenforcement affect access and control. Aswith market institutions, quantificationof asset ownership aspects is likely to bedifficult. Qualitative analysis ofmechanisms that improve (or diminish)asset access is acceptable in most cases.

Fourth, the channel in terms of accessof the poor to publicly providedservices, such as schools, hospitals, andclinics should cover access tononmarketed public services for whichonly a nominal or zero charge is made.This is because, if the poor are to beprovided with public services oncommercial terms, then these will beequivalent to any other market-intermediated commodities and shouldlogically be covered under the access tomarkets and prices row.

Fifth, the mechanism of direct transferswill largely occur through public orprivate transfers that arise due to thepolicy reforms in question. Examplesinclude tax reforms, social securityreforms, and workfare programs wherethe poor are offered work opportunitieson publicly organized infrastructureand related schemes. Subsidizedsystems of supply will be covered bythe previous categories of price effectsand publicly supplied services. However,these more indirect transfer impacts of apolicy are more difficult to predict as theyrequire knowledge of endogenous publicfinance mechanisms or household-levelbehavioral responses, for determiningpublic and private transfers, respectively.

3. Timing and Other Considerations(Columns)

The columns of ADB’s current PIAmatrix (Table 8), can lead to ambiguityin distinguishing between direct,indirect, and macro effects on the poor.For example, removing a fertilizersubsidy will have a direct effect in raisingfarm costs and thus, indirectly, inlowering wage employment among hired

60

CHAPTER 5

farm laborers. On the other hand, if theremoval of the subsidy stimulates higherproduction, there will be positiveindirect employment effects. Anyemployment consequences under themacro effect column would arise fromthe way in which the subsidy had beenfinanced and the way in which the savedfunds are utilized. This direct, indirect,macro distinction makes sense whenthere is a clearly defined market involved.However, once one considers morebroad-based policy interventions and,especially, various institutional reforms,the distinction between an indirect andmacro effect becomes blurred, sooffering less justification formaintaining separate categories.

In general, the distinction betweenshort- and medium-run effects on thepoor needs to be clearer where inter-temporal trade-off considerations areimportant. While direct effects can allbe considered short run, indirect effectscan be viewed in both short and mediumterms. Institutional reforms, because ofthe time needed for implementation, areoften medium term in impact and canhave both direct and indirect effects.Regarding long-run effects, since somany other factors will be at work,isolating long-run policy impacts isunpredictable and speculative. A short-to medium-run focus seems to be mostpractical in the PIA matrix.

Most policy changes will affect thoseabove the poverty line as well as thosebelow it. The use of a column for otherstakeholders is intended to capturebroader impacts. However, this broadcategorization implies homogeneity

across the two groups, which is clearlynot the case. Identification of otherstakeholders, such as consumers,producers, or residents of particularregions, is recommended.

Furthermore, as some policy reformswill result in gainers and losers, the issueof how to handle negative effects shouldbe explicitly addressed to the extentpossible. This may includecompensation for redundant workers, oradding investments that facilitateadjustment, such as worker retraining.Measures that enhance inclusion andaccess by the poor can also help increasethe positive impact of reforms on thepoor. Modifying the current PIA to addthe column for corresponding mitigationor enhancement measures allows formore explicit design considerations tomitigate the negative potential impactson the poor or to include enhancementmeasures, such as ensuring that the poorhave access to services.

F. Operational Considerations

The refinements, as recommended, areintended to reinforce the existinganalytical requirement for systematicassessment of poverty impacts of policychanges. The modifications are largelybased on already applied innovations inthe use of the current PIA matrix and,as such, have already been adopted to alarge degree. The intention is to promotemore consistent and wider applicationof innovative practice.

To conclude this chapter, three practicalconsiderations in the application of thePIA matrix are outlined.



CH

AP

TER 5

• Internal design consistency betweenthe program logical framework,program policy matrix, and PIA matrixis necessary. The PIA matrix isnaturally linked to the program policymatrix as currently practiced, but thelinkage of these two matrixes with theprogram logical framework needs to beclear. The program policy matrix canbe considered as a road map forgovernment commitment to specificpolicy actions in order to ensure thatthe implementation of the reform ison the right track. The PIA matrix isconsidered as a further elaboration ofimpacts of selected policy actionsfocusing on the poor. To ensure clearlinkage, the three matrixes should bepresented as a package of logical andsupporting analysis in a coherent way(see Chapter 7 for a furtherelaboration).

• To ensure that the PIA matrix isunderpinned by reasonable analysis,sufficient time should be given forpreparation of the matrix, andcommencing early in the analytical

work. Thinking through the possiblereform effects of using the PIA canhelp determine the scope of theanalysis that underpins the reforms ingeneral. The government andstakeholders must understand thePIA. To ensure this understandingand input by stakeholders in aspectssuch as mitigation and enhancement,the PIA matrix should be used as aloan design tool rather than anend-of-design reporting exercise.

• Where there are clear limitations in ex-ante analysis of individual programoperations, economic sector work canplay a crucial role in providing ananalysis of wider scope and greaterdepth, and in highlighting keyindicators for monitoring duringimplementation (see Chapter 3).

Used in a pragmatic way the PIA matrixis a useful framework to help assess eachpolicy measure’s channel of effect,analyze the intertemporal anddistribution implications, and identifyareas for further analysis or monitoring.

CHAPTER

ASSESSINGTHE PROCESSESAND COSTSOF POLICY CHANGE

l Understanding Reform Costs

l Fiscal Environment for Policy Reforms

64

CHAPTER 6

A. Introduction

Chapter 3 explained that, while pricereforms are relatively discrete and canbe rapidly implemented through a well-functioning market system, constraintsin transmission mechanisms andsystems will affect benefit realization.For example, institutional realignmentand reforms may take time to implement,affecting the benefits. Furthermore, onereform or program of reforms, or even aset of interrelated reforms, may havelittle impact on people’s lives until theeconomy, related institutions, and thesystem of governance confer broadcomplementary changes. In this sense,procedures and expected outcomes forprogram loans should allow for acomplex picture of reforms that may beexpected to generate benefits over themedium term.

Chapters 4 and 5 discussed how resultsof analyses carried out through sectorwork, including supporting quantitativeand qualitative analysis, providediagnosis and simulation of theproblems and effects of a given policychange. These analyses provide insightsinto the costs and benefits of policyreform, estimates of net welfare gains,as well as the distribution andintertemporal dimensions of a policychange.

On the assumption that policyproblems have been analyzed, andoptions and expected outcomesidentified, this chapter turns to issuesrelated to the process of realizingdesired policy outcomes.

B. Understanding Reform Costs

1. Time Dimensions

Reforms are often triggered in responseto a crisis in economic or sectorperformance, or to declining and chronicunderperformance of key institutions interms of their efficiency andeffectiveness. The reforms, in somecases, can also be seen as a shock—or atleast a change—to the status quoarising from a change to the structureof the economy or the nature ofinstitutions. Once a reform path isembarked upon, those affected oftenfocus on the reforms themselves asopposed to the underlying cause thatinitiated the need for change. In part,this is because adjustment is a processthat takes time, due to possibleunderlying structural rigidities and toresponse lags. Gainers may becomeimpatient with expected improvements,while losers face immediate welfaredecline and await reabsorption into achanging economy or access toalternative services. These situationsoccur because adjustment resulting fromreforms is not always an instantaneousevent due to issues such as asset fixityor missing complementary measures.

During the transition process, a dip canoccur in aggregate output. In terms oftheir flow, the nature of these costscould be seen as similar to that of thecapital costs in projects (althoughdifferent in nature), in that they tendto be incurred up front. For example,the profile of net returns for policyreforms is similar to investmentprojects with negative net flows in the


ASSESSING THE PROCESSES AND COSTS OF POLICY CHANGE

CH

AP

TER 6

early years, becoming positive oncebenefits are being realized in full. Aswith projects, whether the reform willgenerate a positive net present value atan appropriate discount rate dependson the extent of net benefits (Kanbur1990, Ali 1990).

Reforms are often initiated when theeconomy or sector is already in a weaksituation, and may have to beimplemented in the face of, for example,limited fiscal resources to help theadjustment process. This may mean thatthe short-run rigidity in factor marketsis exacerbated and a short-run declinein national output and income occursbefore an overall increase materializes.There may, therefore, be a need toalleviate the short-term costs ofadjustment or accelerate adjustmentthrough budget outlays during thisperiod of redistribution of income (Ali1990).

The full extent of reforms, especiallythe institutional type, may extend wellbeyond the usual administrative life ofa program loan (typically 3 years). Anexpectation by stakeholders of short-term tangible impacts may result in themlimiting their efforts at the margin—especially if politically unacceptablelosses of longer-term benefits or apossible policy reversal are expected(McMillan, Rodrik, and Horn Welch2002). There may be a case for a moreextended implementation time horizonfor institutional programs with tranchesbeing released when the institutionsconcerned are “ready” for the next phaseas opposed to being driven by two orthree closely spaced and difficult toabsorb quick-disbursing tranche releases.

However, even when the administrativeimplementation period of policy-basedloans is short and disbursements arequick, their full effects are often not feltuntil well after the program loan isclosed. This is a feature thatdistinguishes them from projects. Thedynamic nature of policy change effectsalso points to the need to use, to theextent possible, analytical methods thatprovide insights on the changes overtime, rather than methods that simplysuggest the static position.

The above points help in understandingwhat constitutes the costs ofadjustment. First, there are the costs ofthe process of adjustment includingstructural and institutional changes,attendant resource reallocation andmarket realignments, dislocations, andreabsorption into new structures, forlosers as well as gainers. Understandingthis process and its implications isessential for policy makers, businesses,trading partners, and employees, amongothers, to assess the possible impactsand responses to policy changes. Manyof these changes take place over time asadjustments occur and cannot be fullyanticipated. As such, some costs,including the costs of uncertainty, areintangible while the benefits are oftennot immediately realized. Nevertheless,policy analysis can help assess thesechanges and identify, to the extentpossible, the range of economic costsof the equilibriating process, bothtangible and intangible, to the economyand to stakeholders.

Second, there are the real financial costsof the equilibriating or institutionalchange, such as compensatory payments,

66

CHAPTER 6

loss of revenue, new recurrent costs, andother changes in public expenditure.Hence, while a price or institutionalanalysis may reveal a clear net welfaregain from policy reform, the tangible andintangible costs and processes ofadjustment require close examinationand clear articulation to help decisionmakers and stakeholders understand thelikely process.

2. Distribution Implications ofReforms

Closely related to the lagged effects andintertemporal issues of policy change arethe issues of resource reallocation thatcan occur as a result of reforms, theimplications for intra- and intertemporaldistribution of income, andcompensation for losers—assumingthat there is a net welfare gain for society(Ali 1990). This also implies that anyaggregate effects will have to bedisaggregated in terms of theirincidences to different stakeholders.Here, it is important to note the twotypes of distribution effects of reforms:• the intended effects that, if the reforms

are successful, should be enduring inthe prevailing environment; and

• the unintended effects that occur aspart of the adjustment process, butthat, over time, should be overcomeand corrected as resource reallocationproceeds.

Where distribution effects arise,identifying who gains and loses fromreform options is important, given theeffects that distribution can have on theselection and support for policyoptions. In the case of price policychanges, this could affect, for example,

consumers and producers differently. Inthe case of fiscal policy reforms involving,for example, changes in revenues andexpenditures, these could lead to lossesor gains to governments, taxpayers, andbeneficiaries of services.

Understanding distribution effects onconsumers, producers, and governments,all of whom are potential gainers orlosers, can help in:• handling policy conflict and trade-offs

between efficiency and distribution;• designing and implementing reforms

to mitigate or offset any possiblenegative effects on particular groups;and

• raising understanding amongstakeholders—both gainers andespecially losers—in the event of aredistribution of resources andincome.

Public debate about the distributionimpacts of policy change often focuseson the short-term or direct effects ofreforms on individuals and households,as well as on businesses and resourceowners (Productivity Commission2001, p. 23). The indirect effects areoften less evident, despite theirimportance in improving the aggregatefuture situation and the tendency forlosers to also become gainers asadjustments occur. This point isstressed with policy changes because,unlike projects (except for those thatare extremely large relative to the size ofthe economy), the effects of policychanges can be economy- and sector-wide.

The impact of adjustments resultingfrom reforms has been of concern for



CH

AP

TER 6

well over a decade, especially in terms ofthe unintended effects on the poor (aframework for which was elaborated onin Chapter 5). As described, wherereforms clearly have negative effects oncertain groups, mitigation orcompensation for losers may be needed.This can be approached by, for example,taxing winners to compensate losersthrough public funds and resources.However, constraints on relatedincreased short-term taxation mayrequire budget outlays beyond thecapacity of existing fiscal resources andnecessitate the mobilization of realresources to effect compensation, thusincurring an economic cost ofadjustment. Consequently, program loanfinancing can help meet financingrequirements if it is shown to be feasible.Alternatively, mitigation may beconsidered as a separate operation.

3. Political Economy Considerations

The discussion above shows howthe intertemporal and distributioneffects highlight political economyconsiderations of policy change.Clearly, reallocation of resources,redistribution of income, price changes,changes in access to and the cost ofservices, and changes in transfers allaffect the status quo. In turn, this maygenerate resistance to reform measuresamong vested interests.

In principle, governments can resolvechoices between conflicting policyobjectives by weighing prioritiesaccorded to each one of them. However,in practice, governments may be lessthan precise or consistent in their

objectives. This may leave the resolutionof trade-offs problematic. Critically,influential vested interests can be keyobstacles to any kind of trade-offprocess.

As discussed in Chapter 3.F, Sadouletand de Janvry (1995) emphasize thedistinction between political feasibilityon the one hand and efficiency andwelfare on the other, stressing the pointthat any policy reform should firstsatisfy the constraints of politicalfeasibility. It then becomes important,in advance of the passage of the reforms,to identify and understand whoconstitutes the various groups affectedby and opposed to the reforms. Theneed, magnitude, and practicalconsiderations for possible mitigationor compensation require carefulassessment.

For example, while a relatively discretegroup such as retrenched SOE workerscould be provided for in terms oftemporary loss of earnings andretraining, it is virtually impossible tocompensate all poor urban householdswho may be negatively affected bydevaluation. Additional measures areonly likely to be appropriate where thesocial safety net and other generallyavailable measures are ineffective or donot fully address relevant impediments.The range of possible additionalmeasures includes direct compensation,and other specific adjustment assistancesuch as reskilling, workforce programs,and targeted social science assistanceprograms(Productivity Commission2001).

68

CHAPTER 6

It is also generally easier to identifyshort-run costs than long-term benefits.But even in the short run, some costsare uncertain and, in part, will be afunction of the timing and sequencingof the policy reforms, and of the internaland external economic and financialresources made available as part of thereform program. Likewise, economicbenefits cannot be guaranteed. Politicaleconomy factors and unforeseenexogenous shocks may throw policyreforms off course. Program outcomesmay be very different from thoseanticipated. The size of more certainshort-run costs, compared withuncertain medium- and long-termbenefits, may also make stakeholdersunwilling to embrace the reformprogram. However, this uncertaintyshould not be viewed as a constraint inassessing program costs, benefits, andoutcomes.

The outcomes of reforms will be furtherinfluenced by the structuralcharacteristics of the economyconcerned, especially with respect to theexisting distribution of productiveassets and the severity of themacroeconomic situation which, amongother factors, led to the adoption of thereform package in the first place. Ifeconomy-wide stabilization andstructural adjustment programs are stillbeing implemented or have in some sensefailed, at the same time as sector reformsare being implemented, the absorptioncosts of the stabilization and structuraladjustment reforms will representadditions to the costs imposed by sectorreforms. Furthermore, as economy- orsector-wide policy changes do notusually occur in isolation, it may be

difficult to identify, in advance, theresponse from other influences(Productivity Commission 2001).Kasper (1999)24 notes the problemsassociated with ex-ante attribution ofadjustment costs to specific policymeasures. These constraints support theuse of a fiscal framework analysis.

C. Fiscal Environment for PolicyReforms

Policy changes affecting publicinstitutions, spending, and revenuerequire assessment of their fiscalimpact. Figure 6 provides a stylizedrepresentation of an institutionalreform, such as civil service reform. Toachieve the anticipated improvedperformance, a period of adjustment isrequired, which will incur costs.Importantly, unless the reformedinstitution receives sufficient recurrentspending to consolidate and sustain thechanges, performance is likely to declineeven in its reformed state. As such,adjustment costs may be seen as aninvestment in a new policy “asset,”followed by the need for appropriaterecurrent spending to maintain the newpolicy asset. The key difference betweenrecurrent cost implications arising fromreforms, compared with a project, is thatthe former can often be sector-wide.

Viewing policy reform as the creationof an asset requiring maintenancespending implies that understanding isneeded of the net fiscal impact ofreforms in relation to the overall fiscal

2 4 P. 145, as cited in Productivity Commission 2001,p. 16.



CH

AP

TER 6

situation. This can be done throughassessment of a country’s presentmedium-term fiscal framework (MTFF).The MTFF provides the basis forassessing whether government has thenecessary fiscal space to carry out bothits overall reform program and specificreform measures. The followingquestions, especially for reforms withfiscal implications, can be asked inrelation to the MTFF:• What is the existing net fiscal

situation, including both revenues andexpenditures?

• Will new sources of revenue begenerated, how much and when?

• Will increases in recurrent costsarising from reforms be passed on toservice users and has their willingnessto pay been assessed?

• Does the policy reform add to publicexpenditure and, if so, in what ways?

• Will the reforms facilitate therealization of budget savings and whendo they appear?

• What are the real financial costs overtime of effecting the policy change,including consolidation or recurrentcosts?

• What is the difference between thefinancial costs of reform and thegovernment’s net fiscal position? Thisprovides an indication of the gap infinancial resources needed toimplement and sustain the reform.

The MTFF should cover the “withoutreform” stage, through the “withreform” adjustment process, up to theconsolidation stage.

Gradual budget realignment in responseto institutional change and capacitybuilding is addressed in a number of

70

CHAPTER 6

ADB program loans, such as the FSMPrivate Sector Development Program. Inaddition to the adjustment costs or“asset creation” stage, the steps used toassess the recurrent budget implicationsof reform-driven institutional changesrelated to realigning and strengtheningservices include:• assessment of the “without program”

resources on an agency-by-agencybasis and of overall government;

• assessment of the resources neededwith program staff and nonstaffbudget requirements for reformed andstrengthened agencies;

• identification of the capacity to meetincremental recurrent costs by localand national governments andagencies;

• assessment of executive and legislativebranch support for realigned budgets;

• assessment of budget cycles and theperiod needed to hire, train, andintroduce the delivery of realignedservices; and

• preparation of a multiyear budget andanalysis of possible external supportneeds during the transition.

The assessment of the fiscal impact ofthe policy reform complements theeconomic assessment in the same waythat the financial and economicevaluations of a project complementeach other. The financial analysisprovides pertinent information on fiscalimpact and affordability by thegovernment in a fiscal context. Thefinancial assessment of the adjustmentprocess and of the recurrent spendingneeded to sustain the policy reform

provides valuable insights into aspectssuch as efficiency gains anddistribution.25

Assessing the dimensions of a reformprocess can assist policy analysts inunderstanding how the reform will workand in identifying the nature of costsand benefits. Understanding theintertemporal dimensions provides anidea of the timing of the costs andbenefits. Assessment of the distributionaspects is necessary to identify thegainers and losers, possible costsinvolved in facilitating a realignment ofresources, and how to manage levels ofpublic expenditure. Attempting to“soften” transitional impacts can involveboth real financial and economicresource costs. Where there are politicaleconomy implications from reforms, thecosts of managing and sustainingreforms need to be weighed. Whileprecise quantification and valuation ofthe economic costs and benefits of areform process may be difficult, anunderstanding of the nature of costs andbenefits, and how, when, and on whomthey will impact, will help analysts andpolicy makers develop a judgment as towhether the reform is feasible.

2 5 Appendix 4 provides an inventory of how theidentifiable cost aspects of reforms were assessedin recent ADB program loan reports.

CHAPTER

INCORPORATINGPOLICY CHANGEASSESSMENTS INTOPROGRAM DESIGN

l Integrating Policy Analysis and ProgramDesign

l Conditions as a Guide to Implementation

l Linking the Program Framework andReform Monitoring

72

CHAPTER 7

A. Introduction

The preceding chapters have examined anumber of analytical aspects thatunderpin the preparation of policy-basedoperations. Macroeconomic assessmentand sector diagnosis are aimed atidentifying policy- and institution-relatedbinding constraints that hinder thedevelopment of a sector. At the same time,the outcomes and potential impacts ofthe policy reforms need to be assessedwith appropriate analytical tools. Thereform process is analyzed to betterunderstand its dimensions and dynamicsas well as the fiscal implications offacilitating the process. Particularattention is paid to assessing thepotential impact of policy changes on thepoor and, in turn, to consider appropriatemitigating measures to overcome negativeimpacts. The findings from this analysisshould inform and influence the designof any new policy-based operation. Thisconcluding chapter discussesintegration of analytical results into theoperation’s design, and implications forthe design of loan conditions andmonitoring.

B. Integrating Policy Analysis andProgram Design

As part of preparing the groundwork forADB’s policy-based operations, three keymatrixes are used to highlight essentialanalytical results and design features: thePIA matrix; the program logicalframework; and the policy matrix.26 Acommon thread is that they can all be

used as effective design tools, ratherthan reporting frameworks.

The three matrixes represent differentperspectives on a policy-based operationand so their internal logic should bepresented in a coherent and well-integrated way. First, the cause and effectmechanisms of policy changes identifiedin the PIA matrix, which affect both thepoor and other stakeholders, highlightthe assumptions that underlie theprogram logical framework.27 Theconnection between mitigation andenhancement measures in the PIA matrixand the program logical frameworkshould also be clearly linked. Second, asthe outputs of the program logicalframework typically represent theculmination of a series of activities andsteps that results in the completion ofpolicy action, there should be a clearconnection between the major outputsin the program logical framework and thepolicy matrix. In effect, each policycondition in a policy matrix should belinked to the corresponding cell in theprogram logical framework. Figure 7illustrates these links.28

Putting the three matrixes together willpresent essential features of the cause andeffect analysis, economic logic and rationale,reform processes, and expected impacts of

2 7 For a full description of applying logical frame-work to ADB operations, see “Using the LogicalFramework for Sector Analysis and Project De-sign: A User’s Guide” by C.D. Saldanha and J.F.Whittle, Manila, 1998.

2 8 A thorough attempt was made in the UzbekistanEducation Sector Development Program to effectthese linkages. Case 3 in Appendix 3 illustrates thispoint by further extending the presentation in theprogram loan report.

2 6 These three matrixes are discussed in more de-tail in Chapter 5.


INCORPORATING POLICY CHANGE ASSESSMENTS INTO PROGRAM DESIGN

CH

AP

TER 7

FIGURE 7: Integration of the PIA Matrix, Program Logical Framework, and Policy Matrix

1 - Effect analysis summary, assumptions, and risks.2 - Actions to mitigate negative impacts of reform measures or enhance inclusion of poor groups.3 - Policy actions and outputs.

Policy Matrix

Policy Area Tranche 1 Tranche 2 Tranche 3and Measures Policy Actions Policy Actions Policy Actions

Policy Area 1 consistent with program framework purposePolicy Measure

Policy Area 2Policy Measure

3

3

Poverty Impact Assessment Matrix

Channel of Effect

Access to LaborMarkets, WagesAccess to Marketsand Prices

Access to AssetsAccess to PublicServicesAccess toTransfers

Net Impact

Effects on the Poor Effects on Mitigation orDirect Indirect Indirect Other Enhancement

Short Run Short Run Medium Run Stakeholders Measures

21

Design Summary

Program Logical Framework

Performance Monitoring AssumptionsTargets Mechanisms and Risks

Impact: medium- to long-term impact ofthis and related programs

Outcome: the reason why this program isbeing done and the expected end-of-program change

Outputs: the specific deliverables of thisprogram

Activities: the main tasks to accomplishoutputs and effect policy measures 2

31

74

CHAPTER 7

the policy operation. The resulting clearyet simple picture is fundamental toproviding an informed and common basefor public discourse. This will helpstakeholders understand the specificmeasures that will help them during thereform process. Their understanding isa prerequisite for successful reformimplementation.

C. Conditions as a Guide toImplementation

An integrated perspective of policy reformoperations provides new insights into theappropriate function of policyconditions, that is, it serves as a guide toimplementation. Once the main policychanges have been determined, conditionsin policy reform embody the key reformmilestones by setting out reform stepsagreed to by the government and fundingagency, including preconditions, or triggeractions for loan release compliance, andsubsequent tranche releases.

In this sense, conditions can take on anadministrative function for loanimplementation. However, conditionsand a rigid implementation time frameshould not be used excessively, for twomain reasons. First, they can detract fromthe broader issues of the reform processand the possible need for midcoursecorrections where evidence supportsmodifications. Second, where capacity forreform implementation is limited or notfully understood, they can raiseexpectations on immediate realization ofreform benefits among intendedbeneficiaries and other stakeholders,leading to later disappointment if resultsfall short of expectations. These, in turn,can weaken commitment to the reform and

can lead to policy reversals and programcancellation.

To avoid such risks, a focus on results-based conditions and outcomes ispreferable to an extensive list of detailedcompliance conditions. It is alsopreferable to take a sequenced approachto implementation. Better considerationof complex intersector and institutionalsituations and of the overall developingmedium-term fiscal situation would alsobe a feature of such an approach.

Given the wide range of reformcircumstances facing governments,selection of the right program loanmodality is a crucial design decision.29

The integrated perspective highlightedabove provides further insightsregarding the choice of operationsmodality. ADB’s program lendingpolicies and modalities have evolved overtime to accommodate emerging andspecial needs of governments indeveloping Asia. The present range ofprogram loan modalities offers thenecessary flexibility to adapt programdesign to individual circumstances, fromsupport for IMF stabilization efforts orrelated operations, to sector-wide andsubsector structural adjustmentoperations that may involve complexsequencing of reforms and developmentmeasures. When used to the ful lpotential offered by ADB’s program loanpolicy and available modalities, theiroperational flexibility can accommodatea medium-term perspective on policyreforms and the sequencing of reform

2 9 For a full discussion of ADB’s program lendingpolicies and modalities, see ADB, 1987, 1996,1999.


INCORPORATING POLICY CHANGE ASSESSMENTS INTO PROGRAM DESIGN

CH

AP

TER 7

FIGURE 8: How Reforms Confer Benefits:Program Logical Framework and Economic Analysis Links


Define reform targets

Target Risks and assumptionsexogenous to theoperation’s influence

Goal

Objective

Policy reform outputs

Reform process and activities

Reform analysis and monitoring over time

Outcomes

Benefits

Outputs

Outputs to benefitslinkage from policychange

Identification andwhere possiblevaluation of benefits;intertemporal anddistribution analysis

steps necessary for complex, unfoldingoperations, as well as dovetailing with amedium-term fiscal framework to ensurereform measures are “on budget.”

D. Linking the Program Framework andReform Monitoring

Figure 7 highlights the usefulness of theprogram logical framework as a meansof reflecting key cause-effect issuesthat have arisen from, for example,sector analysis , reform-relatedactivities, outputs, objectives, andimpacts. Applied effectively, theprogram logical framework shouldreflect the program’s economic logicand rationale, and its performancetargets, as well as assumptions andrisks—surrounding factors that are

outside the influence of the operation.In this way, the program logicalframework provides the monitoringand evaluation logic, including whatshould be monitored duringimplementation and how to monitorit. The importance of the logicallinkages that emerge from the reformsanalysis and the need to monitor howreform measures help realize intendedbenefits and outcomes over time aredepicted in Figure 8.

A detailed discussion on monitoring andevaluation is beyond the scope of thispaper30 but a few, final, observationshighlight the importance ofcomplementing ex-ante analysis withenhanced monitoring and evaluation. Inaddition to monitoring identified

3 0 ADB’s monitoring and evaluation efforts areguided by its Operations Evaluation Department.The Project Performance Monitoring System isthe primary monitoring and evaluation system usedfor programs.

76

CHAPTER 7

exogenous factors that might affect areform, close monitoring, analysisupdates, and regular reassessment ofrequirements based on progress becomeessential reform support activities.

Specific reasons for ensuring anadequate monitoring effort as theoperation unfolds include thefollowing:• It can provide improved

understanding where the up-frontanalysis is limited by data and relatedanalysis on problems and response, aswell as by feedback effect uncertainties.

• This improved understanding,especially where behavioral responsescould be variable or less predictable,will help accommodate and justifyflexibility in implementation andconditions.

• On-site monitoring with real-timeresults improves the management ofthe reform process.

• Monitoring the key indicators in theprogram logical framework can provideearly indications of possible problemsand the basis for modifications.

Such a monitoring effort requires strongcapacity to be in place to be effective.Countries with weak database andstatistical systems would need capacitybuilding and sufficient funding.Monitoring and evaluation at the publicsector reform program level, for example,requires expanded operational capacityand stability in staffing and managerialguidance. However, a more pressingpolicy challenge is to redress the lowpriority often accorded to monitoringand evaluation.

Integrating analytical results into anoperation’s design provides the basis formonitoring and, as necessary, furtheroperation-related research to analyzekey policy variables, impact effects, andthe need for midcourse corrections.Analytical and monitoring dimensionsthat are well integrated into the designprocess and sufficiently well understoodand accepted by stakeholders willimprove the prospects for a relevant andfeasible policy-based operation and fora favorable outcome.

APPENDIX

PROGRAM LOANREPORTS REVIEWED

78

APPENDIX 1

Loan Date of Board RRP DocumentNumber Approval Number

1485 VIE: Financial Sector Reform 19 Nov 1996 RRP: VIE 291381509 MON: Financial Sector Program 19 Dec 1996 RRP: MON 282001568 MON: Health Sector Development Program 04 Nov 1997 RRP: MON 284511589 KAZ: Pension Reform Program 16 Dec 1997 RRP: KAZ 310911611 THA: Social Sector Program 12 Mar 1998 RRP: THA 316061618 INO: Financial Governance Reforms (Sector

Development Program) 25 Jun 1998 RRP: INO 316601680 PAK: Trade, Export Promotion and Industry 31 Mar 1999 RRP: PAK 234271734 VIE: State-Owned Enterprise Reform and

Corporate Governance Program 21 Dec 1999 RRP: VIE 300581800 SRI: Private Sector Development Program

(Subprogram I) 12 Dec 2000 RRP: SRI 313821873 FSM: Private Sector Development Program 12 Dec 2001 RRP: FSM 333141961 UZB: Education Sector Development Program 06 Dec 2002 RRP: UZB 34160997/8 PNG: Agriculture Program Loan 12 Dec 1989 RRP: PNG 223551406 KAZ: Agriculture Sector Program 23 Nov 1995 RRP: KAZ 284491445 CAM: Agriculture Sector Program 20 Jun 1996 RRP: CAM 271541458 LAO: Second Financial Sector Program 12 Sep 1996 RRP: LAO 265631466 COO: Economic Restructuring Program 26 Sep 1996 RRP: COO 303461506 IND: Gujarat Public Sector Resource

Management Program 18 Dec 1996 RRP: IND 294581513 RMI: Public Sector Reform Program 30 Jan 1997 RRP: RMI 296581520 FSM: Public Sector Reform Program 29 Apr 1997 RRP: FSM 296571608 SAM: Financial Sector Program 19 Feb 1998 RRP: SAM 320501618 INO: Financial Governance Reforms 25 Jun 1998 RRP: INO 316601622 INO: Social Protection Sector Development

Program 09 Jul 1998 RRP: INO 322551624 VAN: Comprehensive Reform Program 16 Jul 1998 RRP: VAN 314851627/8 SOL: Public Sector Reform Program 27 Aug 1998 RRP: SOL 321571661 NAU: Fiscal and Financial Reform Program 16 Dec 1998 RRP: NAU 321041662 PHI: Power Sector Restructuring Program 16 Dec 1998 RRP: PHI 312161673 INO: Power Sector Restructuring Program 23 Mar 1999 RRP: INO 316041675 INO: Health and Nutrition Sector Development

Program 25 Mar 1999 RRP: INO 325161693 TUV: Island Development Program 13 Jul 1999 RRP: TUV 315381738 INO: Industrial Competitiveness and Small and

Medium Enterprise (SME) Development Program 16 Mar 2000 RRP: INO 316441739 PHI: Grains Sector Development Program 24 Apr 2000 RRP: PHI 300871762 BHU: Health Care Reform Program 21 Sep 2000 RRP: BHU 330711779 KAZ: Farm Restructuring Sector Development Program 14 Nov 2000 RRP: KAZ 301061803/4 IND: Gujarat Power Sector Development Program 13 Dec 2000 RRP: IND 296941807 PAK: Energy Sector Restructuring Program 14 Dec 2000 RRP: PAK 321461821 MON: Agriculture Sector Development Program 21 Dec 2000 RRP: MON 312121859 CAM: Financial Sector Program (Subprogram I) 15 Nov 2001 RRP: CAM 32431

Note: The first 10 RRPs in the table are the original RRPs referred to as the core program loan reports in the main text.

Title

APPENDIX

ADB EXPERIENCEON THEMACROECONOMICCONTEXT SECTIONOF PROGRAM LOANREPORTS

80

APPENDIX 2

This appendix reviews the discussions onthe macroeconomic context section thatemerge from an examination of the 10 coreprogram loan reports. Primarily, thequality of assessment of themacroeconomic context in the programloan reports is uneven, and is sometimescompounded by the lack of identificationof variables that may result in criticalfeedback effects.

A. Key Points on How to Improve theDiscussion of the MacroeconomicContext

1. Presentation of the Macro-economic Context. There is variabilityamong the program loan reports interms of the data presented, economicprojections, macroeconomicassessment, and the relevance of themacroeconomic context. Inconsistencyarises due to lack of a clear applicationof the macroeconomic assessment indeveloping the reform program, as wellas uncertainty on how to effectivelyand efficiently undertake suchassessment. The ultimate goal shouldbe to identify the critical macro-meso-micro linkages that will affect the costs,benefits, and risks to the reformprogram. Recommended coverageshould include: (i) a review ofeconomic growth performance,(ii) assessment of macroeconomicmanagement performance,(iii) discussion of key structural policies,(iv) outlook for economic performanceover the relevant program period, and(v) a summary of the macroeconomiclinkages and assumptions to beincluded in the macroeconomicframework.

2. Data Presented. A discussion ofthe macroeconomic context shouldinclude a minimum set of keyindicators. There is inconsistencyacross the program loan reports interms of data presented even thoughADB’s economic reports and countrystrategy and program documentsprovide an accessible and standardizedset of economic indicators. However,additional details may need to beprovided by the analyst. Tablespresented in the program loan reportsshould include precise definitions ofvariables reported. It is also importantto consider the quality of the data beingpresented and their appropriateness tothe analysis. For example, aggregateinvestment data for many least-developed countries is often imprecise.Sector decompositions of investmentshould be interpreted with care incountries with high levels of publicsector or SOE investment, or programsthat would channel investments tofavored sectors.

3. Economic Projections. Theprogram loan reports usually include asection on medium-term prospects andprojections. However, these sectionsoften do not distinguish between ADBstaff projections and governmentmacroeconomic targets anddevelopment goals. ADB projectionsfor the next 2 years for key indicatorsare readily available in the annual AsianDevelopment Outlook and should be usedas appropriate. However, because ADBdoes not generally provide long-termprojections, it is acceptable to usegovernment targets and developmentgoals where the program horizon makes


ADB EXPERIENCE ON THE MACROECONOMIC CONTEXT SECTION OF PROGRAM LOAN REPORTS

AP

PEN

DIX 2

long-term projections necessary. Incases like these, the program loanreport should clearly identifygovernment targets and projections andcritically assess the reasonableness ofthe data.

4. Macroeconomic Assessment. Theprogram loan reports generally do notexhibit a consistent and systematicapproach to progressive discussions ofpast trends, current performance, andprospects. Moreover, there is a tendencyto support broad conclusions withlimited information. This can be avoidedby adhering to a standard format (inwhich discussion of prospects is keptseparate), and by having a thoroughfamiliarity with macroeconomics and theunique characteristics of a particulareconomy. Where judgments are to bemade regarding recent economicperformance and reform programs, anumber of criteria should be used,including, as appropriate, trends ingrowth of output and incomes,employment, income inequality, andpoverty.

5. Relevance of the MacroeconomicContext. While the description ofthe macroeconomic context isadequate in some cases, its relevanceand linkage to the subsequent sectorand program description are, often,unclear. The description can be proforma and superfluous to the programloan report, and may miss keymacroeconomic influences on thesector and the subsequent program.To avoid the tendency toward amechanical review, a final section ofthe macroeconomic context sectionof the RRP should provide an overall

assessment of those macroeconomicfactors that are most critical to policyoperations. These should form thebas is of the macroeconomicframework.

B. Basic Considerations for ImprovingMacroeconomic Assessment forPolicy Operations

As discussed in Chapter 2, a sufficientunderstanding of the macroeconomicenvironment is needed to gain insightsinto the context and possible feedbackeffects of sector reform measures. Thebasic minimum for the assessment,concerning the review of keymacroeconomic indicators, usuallypresented in a series of 5 or more yearsplus projections as appropriate,provides a minimal snapshot of themacroeconomic environment. In theADB context, analysts should be carefulin properly defining and accuratelyreporting this data set (2-yearprojections for this standard set). Thedata set may need to be expanded toprovide a more complete picture of themacroeconomic context, depending onthe specifics of the policy reform exerciseand after these key indicators have beenused for preliminary perusal. (Keymacroeconomic indicators for Cambodiaare provided in Table A2.1 as anillustration.)

Several other points are emphasized.First, it is not necessary for the analystto undertake a detailed assessment ofthe macroeconomic performance of theeconomy. A country economist at ADBis an important resource person for apolicy operation, whose expertise shouldbe used. Second, the analysts should be

82

APPENDIX 2

TABLE A2.1: Key Economic Indicators

Item 1997 1998 1999 2000 2001a

Output GrowthGDP per capita ($, current) 281 247 264 261 259GDP Growth (percent, in constant prices) 4.3 2.1 6.9 7.7 6.3

Agriculture 5.5 3.0 0.0 (0.3) 3.9Industry 21.3 7.3 13.2 34.6 15.5Services (2.6) 0.7 7.1 5.8 2.9

Savings and Investment(current and market prices) ( % o f G D P )Gross Domestic Investment (% of GDP) 14.3 11.3 15.9 13.5 17.9Gross National Savings (% of GDP) 9.0 9.5 11.8 6.9 9.6

Money and Inflation ( a n n u a l % c h a n g e )Consumer Price Index 9.1 12.6 0.0 0.5 (0.5)Money Supply (M2) 16.6 15.7 17.3 26.9 20.4

Government Finance ( % o f G D P )Revenue and Grants 9.0 8.3 10.6 11.0 11.4Expenditure and On-lending 13.0 13.8 14.5 16.3 17.5Overall Fiscal Surplus/(Deficit) (4.0) (5.5) (3.9) (5.3) (6.0)

Balance of PaymentsMerchandise Trade Balance (% of GDP) (7.1) (5.8) (8.3) (7.8) (6.6)Current Account Balance (% of GDP)b (8.2) (6.9) (7.8) (7.6) (6.4)Merchandise Export ($) Growth (annual % change)c 81.0 13.0 17.9 53.2 9.9Merchandise Import ($) Growth (annual % change)c 5.8 1.6 27.0 37.1 5.2

External Payments IndicatorsGross Official Reserves (including gold, $ million) 262 390 422 485 548(in weeks of current year’s imports of goods) 10.6 16.2 16.1 14.2 15.3External Debt Service(% of export of goods and services) 1.2 2.1 1.6 4.0 4.0Total External Debt (% of GDP) 62.9 71.3 67.4 65.5 64.2

Memorandum Items:GDP (current prices, KR billion) 9,778 11,364 12,587 12,932 13,365Exchange Rate (KR/$, average) 2,991 3,774 3,814 3,859 3,924Population (million) 11.6 12.2 12.5 12.8 13.1

( ) Negative.GDP = gross domestic product.a Provisional.b Excluding official transfers.c Some imports are reexported to neighboring countries, e.g., Viet Nam.Data sources: Ministry of Economy and Finance, National Institute of Statistics, National Bank of Cambodia, International Monetary

Fund, and ADB staff estimates.Source: Cambodia Country Strategy and Program Update (2003–2005).


ADB EXPERIENCE ON THE MACROECONOMIC CONTEXT SECTION OF PROGRAM LOAN REPORTS

AP

PEN

DIX 2

cautious about importing into the policyframework the analysis conducted byanother institution. For example, theIMF reports, which are producedspecifically for IMF operations, maynot cover aspects of the economycrucial to the policy, may containtargets for key indicators rather thanprojections, or may base someprojections on assumptions that are not

made explicit. Thus, externally producedreports should be treated only assources of information. Third, theanalysts should focus on answeringkey questions about the macroeconomiccontext in the assessment, and not on apro forma review. Fourth, the depth ofthe discussion in each section of theanalysis will depend on its importanceto the analysis of the reform measures.

APPENDIX

CASEILLUSTRATIONSFOR APPLICATIONOF ANALYTICALTECHNIQUES

86

APPENDIX 3

A. CASE 1. Agricultural and FoodPolicy Decisions in the FederatedStates of Micronesia: Applicationsof Partial Equilibrium Analysis

1. Policy Issues Addressed

The issues in agriculture and food facingdeveloping countries are varied andcomplex. Their nature and relativeimportance depend not only on thephysical circumstances affecting agri-cultural production, but also on a complexinteraction of social and economic factorsthat may provide incentives or discriminateagainst agricultural production andmarkets. Factors that are under the owner-ship and control of farmers, such as land,labor, and capital resources, are termedendogenous. Factors that are beyond thecontrol of farmers, such as market demand,and product and input prices, are termedexogenous. It is the exogenous factorsthat policymakers may wish to influence,giving rise to distortions in productionsystems and markets. The type ofexogenous factor interventions dependson the main objectives of agriculturalpolicy, which may include the following(Streeten 1987):• improving the efficiency of resource

allocation in agriculture to raise foodproduction and its productivity;

• accelerating economic growth throughthe balanced expansion and support foragriculture with respect to other sectors;

• eliminating malnutrition byemphasizing expanded small farmerproduction, employment creation forthe landless, and the reduction of rural-urban differentials and migration;

• improving food security through priceand supply stabilization; and

• maintaining political stability.

Unfortunately, the interventions requiredto meet these policy objectives arefrequently contradictory within theagriculture sector itself and, moresignificantly, between agriculture andother sectors, such as the industrial andurban-based sectors. A policy thatsubsidizes the prices of inputs foragricultural production to increase supplycan result in production inefficiencies andresource allocation distortions, as well asbeing a burden to government budgets.Similarly, lowering food prices toconsumers can cause a fall in productionand market distortions. Raising domesticfood prices to producers increases thecost of living for consumers and reducesthe competitiveness of local producers.In the interests of promoting foodsecurity or self-sufficiency, governmentproduction projects and marketingagencies may be introduced to ensure thatfood supplies are available. But theseefforts frequently become inefficient andundermine the development of markets.Pursuing or proposing policy changeswithout understanding even basic priceand market implications often leads torecommendations that are either ignoredin favor of politically preferred options,or adopted without an understanding oftheir negative implications.

2. Technique Description andApplication

a. Policy Analysis Matrix

Given the array of potential policyinterventions that can be used toinfluence production, prices, andmarkets, it is necessary that adequateanalysis of the effects of interventionsis carried out to avoid undesirable

CASE 1


CASE ILLUSTRATIONS FOR APPLICATION OF ANALYTICAL TECHNIQUES:

AP

PEN

DIX 3

2 Adapted from Agricultural Policy Analysis, JonathanKydd, Wye College External Programme, Universi-ty of London, 1998.

impacts, or at least to understand thetrade-offs of a policy decision. This isespecially true where data on supply anddemand are needed to assess the impactof policies that results in improvedproduction. The effect of subsidies,tariffs, or quotas should, likewise, beanalyzed. However, the lack of availabledata and time constraint frequently limitthe extent of analysis possible. In suchsituations, simplified methods may needto be relied upon.

A useful methodology, based onsimplified partial equilibrium analysis,is the policy analysis matrix (PAM).1The PAM technique was developed to,at the least, provide key indicators onthe market distortions arising frompolicy interventions on verticalcommodity systems (i.e., the productionof a commodity from the farm level allthe way up to its final market). ThePAM is based on principles of partialenterprise budgeting using the basicbusiness relationship: Revenue – Costs= Profit. Use of this principle has thepractical advantage that a great deal ofdata are available (e.g., from farm

management or farm household surveys)or the data can be easily generated, oftenas a matter of course in typicalagricultural sector and project work. Thetechnique also allows policy to beanalyzed directly in the context of easilyrecognizable enterprise profitability asa major concern to farmers. The PAM’skey application is in the static analysisof distortions arising from governmentpolicy intervention as opposed todistortions from endogenous factors.

The means by which the PAM assessesthe divergence between market andefficiency prices is to value revenues fromenterprise sales, costs of production, andprofit in market or financial prices andin economic prices. Table A3.1 lays outthese elements in a form recognizable asa basic farm budget.2

The budget in the first column representsthe costs, revenues, and profits of anenterprise at market prices. The secondcolumn is the same budget but usingeconomic prices. The third column is thedivergence or the difference between therespective items in the first and second

1 Developed by Scott Pearson and first published asThe Policy Analysis Matrix for Agricultural Develop-ment, Cornell University Press, 1989

TABLE A3.1: Elements of the Policy Analysis Matrix

Financial Prices (f) Economic Prices (e) Transfers (t)

Revenues Rf Re Rt

Cost of Domestic Resourcesand Nontradable Inputs DFf Dfe DFt

Cost of Tradable Inputs Tlf Tle Tlt

Profits Pf Pe Pt

88

APPENDIX 3

columns, and represents economictransfers and distortions in resourceallocation and/or the market, orimprovements to efficiency that arisefrom the policy intervention. Separatingthe “costs” element into domesticresource costs and tradable input costspermits analysis of the policy’s impactarising from the relative use of domesticnon-tradable factors of production, inrelation to tradable inputs (the latter arepurchasable locally or imported).Domestic resources may include, forexample, labor and capital charges suchas interest on bank loans, licenses, andtaxes. Tradable inputs may include seed,fertilizer, and chemicals, fuel, equipmentand machinery. Revenues under marketcosts are the actual prices received byfarmers or traders. Where a subsidy,tariff, or quota is used, this will raise themarket price above economic prices. Inorder to assess the economic effects ofthese policies, it is necessary to convertmarket prices to economic prices. In thecase of the FSM, this is done using borderprices for imports (c.i.f. from the UnitedStates West Coast as the main source offood imports) or exports to Guam as aproxy for regional market prices.3

The information provided in the PAMpermits analysis of a variety of policyinterventions that affect prices: taxes andsubsidies on inputs and outputs; factor-market policies such as wage and landpolicies; and basic aspects ofmacroeconomic policy, especiallyexchange rate policy, the subsidy and taxaspects of fiscal and trade policies. Inaddition, the relative efficiency of

commodity systems in addressing thepromotion of domestic factor markets,such as employment, can be assessedusing PAMs, as well as the comparativeadvantage of different enterprises indomestic and export markets and theirability to meet import-substitution policyobjectives. To do this, the PAM uses aseries of indicators and ratios arisingfrom the matrix. Seven of the most usefulratios are described in the followingsection with reference to policies andprojects currently under discussion in theFSM. A concept that is frequently usedin PAM indicators is that of value added,which is the value of the output less thecost of inputs, and provides the returnsto factors used in production. Itsimportance is reflected in the PAM’sindicators as follows:

b. Ratio Indicators

Private Cost Ratio (PCR) =

DFf/(Rf – Tlf)

PCR is the ratio of domestic factor costsat market/private prices to the valueadded created by the enterprise.Enterprises that create a high return toprivate domestic resources used producea high value-added in relation to theseresources, and will be relatively attractiveto the enterprise or owner or manager,i.e., they have a “private” comparativeadvantage.

Domestic Resource Cost Ratio (DRC) =

DFe/(Re – Tle)

DRC is the ratio of domestic factorcosts at economic prices to the value

3 The United States, Guam, and the NorthernMarianas Islands are the FSM’s main tradingpartners for agricultural products.



AP

PEN

DIX 3

added created by the enterprise also atsocial prices. Enterprises that are “sociallyprofitable” have a high value-added inrelation to domestic factor costs, and havean “underlying comparative advantage.”As a rule of thumb, ratios below the valueof 1 have super-normal social profits anda high comparative advantage, while thosewith values exceeding 1 are at adisadvantage. Under a policy thatencourages comparative advantage,enterprises with the lowest values shouldbe emphasized while those with highvalues should be reviewed for possibleways to reduce the tradable input usethrough greater use of domesticresources. In the FSM examples (TableA3.2), many tropical fruits and vegetableshave a value well below 1 and, given thehigh level of imports and low domesticproduction, have good potential forcompetitive import substitution. Semi-intensive poultry units implementedthrough a donor and government-backedproject, however, had a DRC in excess of8 (and would be higher still if other costswere estimable). This means that thereturns to domestic resources and therelative comparative advantage were bothlow. In addition, the high cost ofimported inputs such as feed and day-old chicks significantly reduced theforeign exchange savings from domesticpoultry production. One way to improve,that is lower, the DRC in this case is tochange to production methods thatreduce the level of dependence onimported feed, provided that local feedsources are also economically competitive.Recommendations that enhance localproduction methods have a wideradoption rate. Only government-backedprojects promote intensive, import-dependent broiler production and have

proven commercially unprofitablewithout subsidies.

Nominal Protection Coefficient (NPC) =

NPC = Rf/Re

NPC on tradable commodities/outputs isthe ratio of the domestic market orproducer price of the commodity to thesocial or efficiency prices (taken as theborder prices as a proxy for world prices).The NPC output ratio is an indicator ofproducer price competitiveness withrespect to world prices (note there is alsoan NPC for inputs but that is notdiscussed here). Ratios over 1 indicate thatproducer prices are high compared toworld prices and either producers havesome form of protection or prices are setat uncompetitive levels. Ratios of less than1 are an indication of a tax on producewith farmers as the losers, and governmentor consumers as the beneficiaries. Thelimitations of NPCs as indicators include:the difficulties in identifying world anddomestic market prices; overvaluedexchange rates that will overstate domesticprices and understate the extent of effectivetaxation; and the implications of inputpricing, use, and policy cannot beconsidered. Given that differententerprises use different productionsystems and types of inputs, this ratio, ifused alone, is a poor comparative indicator.

The Effective Protection Coefficient(EPC) =

(Rf-Tlf) / (Re-Tle)

EPC is the value added of an enterpriseat market or private prices in relationto the value added at social or efficiency

CASE 1

90

APPENDIX 3

4 Based on an analysis of poultry production in the Federated States of Micronesia, using a template fromAgricultural Policy Analysis, Jonathan Kydd, Wye College External Programme, University of London, 1998.

Indicators:

Private Profit/Loss ($/batch) 706 Return to farmer/management.

Economic Profit/Loss ($/batch) -583 If negative, profitable production can only continue withsubsidy, import tax or quota

Private Cost Ratio 0.63 Ratio of domestic costs to value added at market prices.

Domestic Resource 2.49 Ratio of factor costs to value added at social prices. RatiosCost Ratio between 1 and 0 show efficiency.

Nominal Protection Coefficient(on tradable outputs) 1.40 Ratio of private prices to economic prices for the commodity.

(on tradable inputs) 1.16 Ratio of private prices to economic/market prices for tradableinputs for the commodity.

Effective Protection 4.93 Ratio of value added at private prices to value added atCoefficient economic prices.

Revenue/IncomeWholesale price incapital (batch of 2,700broilers)Import parity price

Domestic andNontradable Inputs

Hourly wage for labor

Building operationcostsBuilding depreciation

Tradable InputsImported day-oldchicks

Imported feedTransport outlays

Imported poultrydrugs ($0.02 per bird)

Units

bird/batchlbs/bird

lbs

hours/batch

$/batch

$/batch

no/batch

lbs feed/bird$/batch

no. of birds

Quantities

A B

2,7003

8,100 1

1,000 1

1 1

1 1

2,700 1

2,700 6.6160 1

2,700 1

MarketPrices

($)

1.05

0.75

1.00

200.00

25.00

0.75

0.250.25

0.02

Conver-sion

Factors

1.00

0.75

1.00

1.00

1.00

0.801.00

1.00

SocialPrices

($)

0.75

0.75

200.00

25.00

0.75

0.200.25

0.02

MarketValues

($)

8,505

1,000

200

25

2,025

4,45540

54

EconomicValues

($)

6,075

750

200

25

2,025

3,56440

54

Transfers($)

2,430

250

0

0

0

8910

0

TABLE A3.2: Example of a Price Policy Analysis Matrix4



AP

PEN

DIX 3

prices. It has the advantage over theNPC ratio in that it considers the type,source, and cost of inputs in relation tothe product’s price. As with the DRCratio, this is useful in identifyingmethods for improving the economicefficiency of production to overcomeproblems of lack of competitiveness andmarket failure.

c. Nonratio Indicators

Private Profits (Pf) =

Rf – DFf – Tlf

Pf is simply the profitability of theenterprise system to the producer oragency, and represents the basic elementsof an enterprise budget. Although itallows the profitability of the individualenterprise to be assessed, the value canonly be used to assess one enterprise,and should not be used to compare theprofitability among other enterprises.For comparisons, the earlier ratios aresuperior, particularly the private costratio. Where no labor is hired and onlyfamily labor is used, it is useful tomodify this indicator to differentiatebetween the return to management andthe return to labor. An example of thisis given in the FSM poultry PAM.Returns to labor are also useful whencompared with the opportunity cost oflabor (taken in the example as theminimum wage rate). However, ifreturns are calculated solely in terms ofreturns to family labor and management,the profit for a batch of broilers appearsmuch better. This was the basis uponwhich the original project appraisalestimated the viability of the proposedunit and concluded its worth, despite

the high DRCs and losses in terms ofsocial prices (which were not calculated).In cases where labor data are unavailableor considered inaccurate, an estimationof the returns to family labor andmanagement may, nevertheless, be auseful indicator. In this way, PAMs area useful tool for assessing the economicand policy implications of a project, aspart of the project planning andappraisal process.

Economic Profits (Pe) =

Re – DFe – Tle

Pe is the same calculation as for privateprofits but uses efficiency prices, allowingassessment of the economic efficiency ofthe enterprise. In the poultry example,the result is negative, indicating that theonly way the enterprise can survive is withprotectionist measures such as tariffs orquotas. This brings into question theproject advice to protect poultry farmers,especially as it will result in an increasein the price of poultry to consumers andis likely to draw a political reaction frominfluential importers and retailers. Incomparison, island staples (i.e., fruits andexotic vegetables) are profitable, exhibitcomparative advantages, and with a knownregional export market. Moreover, theproduction techniques of island produceare better known to farmers.

Transfers include output and inputtransfers, and net transfers and arecalculated by subtracting the market orprivate prices from the social orefficiency prices for the respectiveelements. For example:

Pp – Ps = Pt

CASE 1

92

APPENDIX 3

where Pt is the net of output, input, anddomestic factor transfers. The transferindicators are a means of assessingmarket failure, distortionary policies,and improvements in efficiency. Wherethe net transfer value (Pt) is negative,this implies that a subsidy either existsor is required for the enterprise tocompete with economic prices.Otherwise, efficiency improvements toproduction may be required. Where therevenue value (Rt) is positive, thismeans that the price of the commodityis in excess of the world market oreconomic price and that prices are eithersubsidized or there is market failure, asoccurred in the FSM. Where there arenegative transfers of tradable inputs, thismeans that a subsidy exists, and theinput price is less than the efficiencyprice. Although not applicable to theFSM, this can also occur as a result ofan overvalued exchange rate.

3. Technique Limitations

Despite the usefulness of the conclusionsthat can be drawn from these indicators,the PAM technique also has inherentlimitations. First, the results describe theeffects of price distortions but notnecessarily the cause, which couldextend beyond current governmentinterventions. This makes comparisonwith a counterfactual, which may be acomplex interaction of many endogenousfactors, difficult. Structural analysis ofthe industry is still needed. Furthermore,as the PAM is a static technique thatcannot assess dynamic economic behaviorand relationships, it should be usedcautiously in assessing desirable policychanges to effect structural adjustmentand transformation. In some cases,

certain short-term interventions may bedesirable to improve longer-termcomparative advantage.5

4. Conclusion

Given the complex interaction of macro-and microeconomic factors thatinfluence the large numbers ofagricultural enterprises and theirproduction-to-market systems, even ina small Pacific island economy, acomprehensive analysis of agricultureand food issues can be daunting,especially when information is in shortsupply. The risk of a limited analysiscan result in a costly error for a projectthat was designed to fulfill a policy, inthis case, of import substitution. ThePAM is a useful technique that buildson familiar enterprise budgettechniques, which can still be dataintensive, but is a reasonable alternativeto full supply (and underlyingproduction functions) and demandschedule analysis for which data may notbe readily available. In the context of theFSM, the PAM technique has assistedin a better understanding of enterprisesthat: (i) have comparative andcompetitive advantages, (ii) create valueadded in domestic and regional exportmarkets, and (iii) require minimaldependence on government subsidiesand other distortionary trade measures.This provides the basis for an initialassessment and for more in-depth andfocused policy analysis and projectidentification in the future.

5 A fuller discussion of the limitations of PAM, andits applications, can be found in Food andAgriculture Organization Training Materials forAgricultural Planning, No. 31, J. Harrigan, R. Loader,and C. Thirtle. 1992. FAO Training Materials forAgricultural Planning, No. 31. Rome.



AP

PEN

DIX 3

B. CASE 2. Philippines Power SectorRestructuring Program: TariffPricing for Cross Subsidies


In the case of a public power or watersupply scheme, in which metering is notprohibitively costly, alternative tariffstructures can be a crucial determinantof its benefit distribution profile andpoverty impact. Often, there is a trade-off between the aggregate economicefficiency of the supply and its equityoutcome depending on the tariffstructure. Therefore, the design orreform of the tariff structure is animportant part of poverty impactanalysis for sector-level reform programsas well as investment projects. ADB’sGuidelines for the Economic Analysis of Projects(1997b, Appendix 26) provides anillustration of a poverty impact analysisof a water supply project underalternative tariff structures. Here, thecontext is a sector reform program:Philippines Power Sector RestructuringProgram (PPSRP).

The PPSRP seeks to create competitiveelectricity markets through theprivatization of the National PowerCorporation (NPC) and the provisionof open and equal access to transmissionand distribution. Electric power in thePhilippines is traditionally provided bythe public sector. Utilities are regulatedas vertically integrated monopoliescontrolling the three components ofgeneration, transmission, anddistribution. In the case of thePhilippines, the NPC has a virtualmonopoly on generation andtransmission. The distribution function

is being handled by about 119 ruralelectric cooperatives, 15 private utilities,and a handful of public provincialutilities.

Tariffs often exhibit cross subsidies(between customer classes and regions)in view of political pressure to favor oneclass of consumers over another. In thePhilippines, industrial, commercial, andhigh-consumption residentialconsumers have been subsidizing poorresidential customers (those belongingto the lowest consumption bloc of 0–50 kilowatt-hour (kWh) monthlyconsumption category). The PPSRPseeks to phase out these cross subsidiesin the tariff structure in line with theobjective of configuring a morecompetitive power sector. Removal ofcross subsidies could lead to higherelectricity tariffs for poor residentialcustomers and lower tariffs forindustrial and commercial groups, aswell as residential consumers with higherkWh monthly consumption. This willimply a higher level of expenditures bypoor residential consumers for the sameamount of electricity. The extent of thetariff increase to the poor is estimatedas the difference between average actualmonthly charges to the lowest group ofpower consumers and the estimatedmarginal costs of supply to this group.It has been argued in the program loanreport that, in the long run, the fullrestructuring of the electric industry isexpected to introduce efficiency gainsthat will have substantial effects on themacroeconomy. In particular, thereduction in electricity tariffs is expectedto lower electricity prices to productionunits that would make industries morecompetitive.

CASE 2

94

APPENDIX 3

TABLE A3.3: Elements of Power Subsidy Determination

Generation/Transmission Distribution Supply LRMC Tariffs SubsidyCosts (P/kWh) Costs (P/kWh) (P/kWh) (P/kWh) (P/kWh)

(A) (B) (C=A+B) (D) (E=C-D)

Residential(0-50 kWh)

Residential

Commercial

Industrial

A more competitive environment isexpected to put downward pressure onthe tariff rates closer to the costs ofproducing the last kWh of electricity.This is in consonance with the economicefficiency principle of aligning the priceof a good or service equal to the cost ofproducing an additional unit of it (itsmarginal cost). The estimation of thetariff levels that are likely to be achievedby the full restructuring requiresestimating the long-run marginal costs(LRMCs) of supply (which includegeneration, transmission, anddistribution). These estimates are thencompared with the existing tariffs thatare being paid by various consumercategories. The difference between theLRMC supply for a particular customercategory and the average tariff they payis the subsidy. A positive subsidy for aconsumer grid/category means that it isbeing subsidized while a negative subsidymeans that the particular grid/categoryis subsidizing other grids/categories.

In principle, estimates of the LRMC foreach consumer grid/category are neededwhich would require computing the

LRMCs of generation, transmission, anddistribution. Under the technicalassistance (TA) attached to the PPSRP,LRMCs were determined using a regionalcapacity and energy market forecastingsystem that simulates the fundamentalsassociated with competitive powermarkets. A computer program wasutilized to advance market price fore-casting forward from a typical cost-basedsimulation approach to an approach thatis market-based or bid-based. A dual-commodity market was simulated usinglinear programming techniques (TAPHI-3127: Consumer ImpactAssessment, Final Report, 18 May 2001,Navigant Consulting, Inc.).

In the other LRMC studies for thePhilippines, proxies were used inestimating the LRMC in view ofmeasurement difficulties. A separatecomputation of the LRMC forgeneration and transmission was doneusing an incremental approach whereindemand for electricity was advancedincrementally in units of kWh. Given eachincrement, a system was optimized togenerate the additional 1 kilowatt of



AP

PEN

DIX 3

demand using the most efficientconfigurat ion of plants . Theresulting incremental cost was anapproximation of the marginal cost.Incrementa l capita l costs wereincluded by annualizing the presentvalue of investments in generationand transmission capacity (implied bythe optimal expansion), anddistr ibut ing these over theincrements in demand (Deloitte andTouche Consulting Group 1997).

Another study was conducted toestimate the LRMC for distribution.The average cost for distribution wasused as a proxy for marginal cost ofdistribution, where:

Average Cost =Operating Expenses/kWh sold

Data on operating expenditures (net ofdisallowed/unrelated expenses,depreciation, and purchased power costs)and kWh were taken from a sample ofprivate electric utilities and rural electriccooperatives. These costs were adjustedto reflect the customer-related costs ofservice to residential, commercial, andindustrial consumers based on the coststructure of the biggest private electricutility (i.e., Meralco). An economic costof capital is imputed and added to thisto obtain an estimate of total economiccost. Similarly, data on net operatingassets were generated from the sampleand the value was converted into anannuity with a period of 50 years toapproximate the economic cost of capitalfor distribution (Lee 1998).

Warford (1997) argues that any formof marginal cost pricing has to be

approximate and, ultimately, someaveraging of costs over a range of outputis always required. A suggested approachis to broaden the definition of marginalcost, and to set price equal to the averageunit of incremental output. The averageincremental cost (AIC) formula (see Box8) may be adapted to encompassgeneration, transmission, anddistribution components of powersupply. It can be estimated by dividingthe discounted value of future supply(investment costs) costs plus theincremental operation and maintenance(O&M) costs by the (similarlydiscounted) amount of additional kWh,representing consumption.

2. Technique Description andApplication

Munasinghe and Warford (1982)adopted a long-run AIC method inestimating capacity, energy, andcustomer costs for the Philippinesusing data from NPC, three sampleprivate utility firms, and six samplerural electrification cooperatives(RECs). These cost computations werebased on the expansion programs ofeach sample utility, historical data onthe utility’s operating performance, andsystem characteristics. The methodused, while it led to a morecomprehensive analysis, entailedenormous data requirements.

a. Long-Run Marginal Cost of BulkSupply from National PowerCorporation

Capacity Costs. Incremental capacitycost consisted of investments inadditional generation and high-voltage

CASE 2

96

APPENDIX 3

(HV) facilities, including relatedincreases in O&M costs as a result ofplant expansions. These yearly investmentstreams for generation and transmissionplant projects were derived from NPC’s10-year power expansion program. Theinvestment flows were discounted (usingthe social discount rate of 15%) togenerate the present worth of theprogrammed capital outlays. The resultantpresent worth of investments was thenannuitized over the estimated economiclife of the plants in each grid. The presentworth of the annual increase in generationand transmission O&M costs were,likewise, derived. Costs at the generationand HV transmission levels were thencompared with the present value of theincreases in capacity requirement duringthe planning period, to generate annualincremental capital and O&M costsestimates. These were then appropriatelyadjusted for power losses fromtransmission and were added to the annualincremental capacity cost at HV levels.

Energy Costs. These were derived fromthe fuel costs associated with the varioustypes of plants scheduled for operationover the 10-year planning period. O&Mexpenses of generation and HVtransmission plant, required for energyproduction and deliveries, were also added

as part of energy costs. The fuel costswere applied to the projected annualstream of plant generation to derive theannual fuel cost for each grid. Increasesin fuel expense over the period were thendiscounted and compared with theincreases in energy generationrequirements to derive the incrementalenergy fuel cost at generation level.Increases in O&M expenses related toenergy were, likewise, discounted. Thesum of incremental energy costs at thegeneration level was adjusted fortransmission losses.

Customer Costs. These are incrementalexpenses directly attributable toconsumers which include costs ofhookup, metering, and billing. Strictlyspeaking, some of NPC’s administrationand general (A&G) expenses were notdirectly related to customers. However,these expenses were considered since aconsiderable amount was associated withcustomer billing, accounting, collection,and other related services. Theincremental A&G costs were estimatedfrom the projected annual increases inA&G costs. Then, the present worth wasderived and their corresponding annuityvalues over the projections of the numberof customers were obtained. Since NPCis considered a bulk supply authority,

BOX 8: Average Incremental Cost (AIC) Formula

where It is the investment cost in year t, (Mt - Mo) is the operation and maintenancecost in year t due to incremental consumption of electricity in year t or (Qt - Qo), r is thediscount rate, and 0 is the base year.

Source: Warford (1997).

( ) ( )

+−∑ =

tott rQQ 1// 1( ) ( )

+−+∑= =

tottt rMMIAIC 1/1



AP

PEN

DIX 3

customer-related costs represent only asmall portion of NPC’s total costs.

b. Long-Run Marginal Cost ofRetail Supply from PrivateUtilities

Capacity Costs. Incremental capacity-related costs for retail power supplyinclude capital expenditures ontransmission and distribution (T&D)facilities, plus attendant O&Mexpenses. A disaggregation of T&Dcosts into customer-related andcapacity expenses was done since onlya segment of T&D expenses was relatedto capacity. The cost allocation tookinto account the number and spread ofcustomers. Hence, for HV andmedium-voltage (MV) suppliesinvolving relatively few customers,100% of capital and O&M costs wereclassified as capacity-related. For low-voltage (LV) deliveries, 70–90% ofinvestment and O&M costs wereallocated to customer-related costs.Capacity-related capital investmentswere then matched with additionalinvestments to be supplied as a resultof expected increases in demand. Thesewere subsequently discounted at 15%.Incremental capital costs were added tothe incremental O&M costs to obtainthe total incremental capacity at varyingdelivery voltages.

Energy Costs. Estimates of incrementalenergy-related costs of retail deliveriesof private utilities consist of costs ofenergy purchased from NPC andexpenses related to self-generatedenergy. A weighted average value ofenergy cost of these two was thencomputed for each utility.

Customer Costs. Annual incrementalcustomer-related costs for distributionfacilities (at the LV level) includeincremental capacity costs, incrementalO&M and A&G costs, and incrementalmeter costs.

c. Long-Run Marginal Cost ofRetail Supply from Rural ElectricCooperatives

Capacity Costs. These were based on theprojected investment requirementsneeded to develop primary and secondarydistribution lines and associatedfacilities. The annual stream of theseinvestments was compared with theforecast of additional load demand.Applying a 15% discount rate andassuming an average plant life of 25years, the resulting incremental capacitycost per kilowatt was converted into itsannual equivalent. Thirty percent oftotal incremental O&M, as well as A&Gexpenses were assumed to be capacityrelated.

Energy Costs. Energy-related costs werebased on the costs of energy purchasedfrom NPC’s bulk power supply andadjusted for T&D losses down throughthe LV delivery level.

Customer Costs. Components ofcustomer-related costs of power suppliedby cooperatives included total costs ofadditional secondary lines, transmissionfacilities, service drops, and meters. Alsoincluded were substantial percentages ofthe O&M expenses; generaladministrative and consumer accountsexpenses; and some provision for the costof architectural and engineering services,street lighting, and general plant and

CASE 2

98

APPENDIX 3

other contingencies. The stream ofprojected investments in distributionfacilities was discounted to derive theannual equivalent. Other customer-related incremental expenses were thenmatched with the forecast of additionalcustomers to be connected.

3. Existing Tariff Levels and LifelineRate

Following Lee (1998), average tariffs ofthe various categories can be taken fromthe tariff rate schedules of the RECs andthe sample electric utilities. A weightedaverage can be estimated (by using therespective kWh sold) to obtain theaverage tariff rate per consumer category.In the case of the tariff rates paid by thepoor, an average estimate can be derivedfrom the tariff schedules (i.e., ratesapplying to the lowest kWhconsumption) of the RECs and theprivate utilities. After deductingestimated taxes (in the case of utilities),an overall weighted average tariff rate foreach customer category can then becomputed by using the shares of privateutilities and RECs’ kWh sales to totalkWh sales, respectively.

In the Philippine case, statistics indicatethat the main uses of electricity by ruralhouseholds are on recreation, lighting,and space cooling. Only a very minimalproportion (about 3%) of ruralhouseholds use electricity for cooking andfood preparation. Prior to the reformmeasures, total expenditure ofhouseholds (with annual income belowthe poverty threshold) on electricity wasestimated to be about 1.6–3.0%. Thus,the impact of subsidy removal was judgedto be relatively small due to the small

percentage expenditure of the poor onelectricity (Table A3.4). This can beattributed to the fact that demand forelectricity is price inelastic (the quantityof electricity demanded increases anddecreases by a smaller proportion thanthe changes in electricity prices). Inestimating the impact of removal ofsubsidies, data on household expendituresby income and by region were obtainedfrom the Family Income andExpenditures Survey. The share ofhousehold expenditures on electricity(which is included in the fuel, light, andwater category) to total expenditures wasthen computed. Tariffs are assumed torise by the amount of subsidy removed.

The total increase in monthly expendituredue to tariff increases ranges from a lowof 2.5% in Luzon to a high of 3.6% inBohol. In absolute terms, these areincreases of between P55 and P86monthly (Table A3.5). The tariffincreases required to remove crosssubsidies are deemed to have only a mostmodest impact on total expenditure ofthe poor. Hence, no particular mitigatorymeasures are mentioned in the programloan report apart from a phasing in ofthe increases. The poor are expected tobenefit from lower electricity prices inthe long run as a result of the competitivemeasures introduced by the restructuringprogram, but no attempt is made toquantify the extent to which long-runpower tariffs might be lower due toefficiency gains. However, a similarprogram for Indonesia (Loan 1673-INO), while also estimating a smallincremental expenditure for the poor(valued equivalent to less than 1% ofannual rice consumption), mentions theintention to establish a Social Electricity



AP

PEN

DIX 3

TABLE A3.5: Impact of Subsidy Removal on Budget of HouseholdsConsuming 0–50 kWh

Increase in MonthlyElectricity Household Increase in

in Subsidy to Expenditure Expenditure NumberTotal Electricity of Total of Average of Poor

Grid Expenditure Tariff Removal Poor Family Households(%) (%) (%) (P)

Luzon 3.0 84.6 2.5 65.4 1,618,635

Bohol 2.1 172.0 3.6 85.7 —

Cebu-Negros-Panay 2.3 118.0 2.7 65.0 602,327

Mindanao 1.6 196.6 3.2 74.3 859,637

Leyte 2.4 113.7 2.7 55.5 233,512

— Data not available.Source: P. U. Lee. 1998. Report on Impact on Poverty of Philippine Power Sector Restructuring, University of Asia and

the Pacific, May, Table 11.

Development Fund to continue tosubsidize the most vulnerable.Presumably, this is in recognition of thebasic point that even small increases incost can have a substantial impact on thevulnerable households.

With the exception of the Meralcofranchise area, legislated subsidies orlifeline rates are not offered to low-income captive market consumers whocannot afford to pay at full cost. However,under the Power Reform Sector Law, alifeline rate is proposed to be set up to

supply the basic electricity needs ofmarginalized end-users. Aside fromequity grounds, a lifeline rate is deemednecessary in cases where the cost ofelectricity consumption is high vis-à-visa particular income level. In practice,proper targeting of beneficiaries isessential in view of possible “leakages”in the system (i.e., benefits may be partlycaptured by a nontarget group). On theother hand, narrowly defining thebeneficiaries, while it lowers the cost ofintervention, may unduly result inundercoverage of the targeted poor.

TABLE A3.4: Subsidies to Residential Consumers before Power Restructuring

Total Residential Residential Long-RunGrid Sales Sales to 0-50 kWh Marginal Cost Subsidy

(MWh) Consumers (MWh) (P/kWh) (P/kWh)

Luzon 8,474,408 152,539 4.3 2.0Bohol 37,566 676 8.3 5.2

Cebu-Negros-Panay 1,618,722 29,137 5.7 3.1

Mindanao 1,013,822 18,249 5.7 3.8

Leyte 91,811 1,653 6.4 3.4

Source: P. U. Lee. 1998. Report on Impact on Poverty of Philippine Power Sector Restructuring, University of Asia andthe Pacific, May, Table 9.

CASE 2

100

APPENDIX 3

C. CASE 3. Uzbekistan EducationSector Development Program:Integrating the Logical Framework,Policy Matrix, and Poverty ImpactAssessment


Enhancing the relevance and upgrading thequality of education are important factorsin Uzbekistan’s transition toward amarket-driven economy. The educationalsystem is beset by various concerns,including significantly lower participationrates in basic education than officialstatistics may suggest and lower literacyof much of the adult population broughtabout by the Government’s decision toswitch fromthe Cyrillic to the Latin script.The higher dropout rates afterindependence suggest rising direct andindirect costs of schooling, especially inrural areas, and poor levels of preparednessfor primary education as a result of a sharpdecline in preschool attendance. Thephasing out of the Cyrillic alphabet in thecoming years is also expected to affect themajority of teachers. These issues areexacerbated by gender disparities thatbecome more prominent as one moves upin education levels. Girls’ access totechnical education remains restricted,especially in the field of science, whichtends to confine employment of womento a relatively few sectors (e.g., education,health, food processing) and lessen theiropportunities for career advancement. Itis also expected that access to educationfor children from linguistic and ethnicminorities will become an issue in thefuture as a result of the recentgovernment decision to reduce thenumber of languages of instruction fromseven to three.

Structural problems concerning thequality, relevance, efficiency (bothexternal and internal), and effectivenessof the current educational system needto be addressed to generate substantiveoutcomes. There is a need to revamp thecontents, processes, and institutional set-up of the system to better serve the needsof changing socioeconomic and politicalconditions. Budgetary constraints hinderefforts to achieve qualitativeimprovements (e.g., teacher retraining,materials development). Suboptimalutilization of public resources (reflectedin small class and school sizes) and lowefficiency ratios and teaching loadscharacterize the present educationalsystem, highlighting the need forreforming the contents and methods ofeducation at all levels to make it morerelevant to the needs of its clientele andto provide a solid foundation for furtherlearning and entry into the job market.

A precondition for the success of theeducational system, as a catalyst in thetransition process, hinges on theimplementation of cost-effective andsustainable reforms. The EducationSector Development Program (ESDP)is envisaged to meet the twin objectivesof carrying out sector-wide policymeasures and implementing suitableinvestment activities to enhance thequality of basic education and improvemanagement practices in the sector.Broadly, ESDP is designed to address thefollowing areas: (i) modernization of thestructure, contents, and processes ofeducation; (ii) sustainability andefficiency improvements in the sector;(iii) sector governance reforms; and(iv) provision of protection for the poor.Crucial interventions will take the form



AP

PEN

DIX 3

of the following: (i) institutional changesrequired to improve system equity andefficiency; (ii) support to a cost-effectiveand pro-poor restructuring of educationbudgets; (iii) assistance for deploying theeducational personnel in a more efficientway and for enhancing staff performance;(iv) participatory research studies incritical policy reform areas;(v) establishment of a distance education(DE) capacity to retrain teachers in acost-effective way; (vi) development ofteacher education courses on teachingmethods and selected subject contents;(vii) provision of educational equipmentto teacher training institutions;(viii) skills development programs forschool principals and local-leveladministrators; (ix) improvement of thephysical conditions and resource base ofschools; (x) assistance to communitiesin establishing school boards; and(xi) provision of grant funds accessibleto schools on a competitive basis.

2. Adjustment Cost and MitigationMeasures

Reforms required to improve theefficiency of the system will entail short-term losses (e.g., staff retrenchment andredeployment, increased teacherworkload, reduced budget allocations forstipends and scholarships, closing ofschools, abolition of the system of freetextbook provision) for certainstakeholders and thus, may face resistancewithin the system. These will entailadjustment costs and compensatorymeasures. In particular, mitigationmeasures will have to be instituted toaddress the issues of planned downsizingof administrative personnel andrationalizing the existing school

networks. Personnel reduction will beachieved through natural attrition(retirement) and voluntary resignationof teachers unwilling to continue holdingtwo jobs (as was frequently the casepreviously). Outplacement and retrainingservices will be provided to staff to beretrenched and reassigned to newpositions.

Specific measures (e.g., free textbookprovision will be continued for certaincategories of pupils) will also beintroduced as social protection schemefor the poor who may be affected by theproposed introduction of user charges inthe educational system. Teacher salariesand benefits will be protected. Particularattention will be given to maintainingaccess of rural students to basic educationwho will be affected by mergers andclosures of schools. Schemes involvingdirect student assistance will be bettertargeted toward needy higher educationstudents by integrating social criteria inthe granting of scholarships and stipends.School-based project interventions (e.g.,poor school rehabilitation program andthe School Initiatives Fund) will betargeted toward schools located in poorand remote areas where the capacity ofthe community to support education isrelatively low.

3. Budgetary Impact of the Program

The share of education in total recurrentcentral government expendituresremained constantly above 22% since1995. Almost 90% of total educationexpenditure comes from the centralbudget, either directly (for financinghigher education institutions andacademic lyceums), or through

CASE 3

102

APPENDIX 3

transfers to regional governments (forprimary and secondary schools andprofessional colleges). The remaining10% is financed through tax revenuesgenerated by local governments.Funding for noncompulsory levels ofeducation (i.e., preschool and highereducation) relies in part on extra-budgetary sources (e.g. , namelypayments made directly to institutionsby students, families, companies, etc.).Extra-budgetary revenues account forabout one third of total expendituresby higher education institutions, andare primarily comprised of universityfees paid by employers for “contractstudents” (as distinguished from“budget students” who receivescholarships and in some instancesstipends).

A simple simulation was undertaken toassess the impact of ESDP on therecurrent education budget during theimplementation of the Program. Policymeasures are expected to generate cost

savings amounting to $208.7 millionwhich are much higher compared withthe estimated budgetary cost of about$54.5 million. In annual terms, a netsavings of about $38.5 million per yearwill be realized as compared with theyear 2000 recurrent education budgetof about $410 million. Table A3.6 setsout the details of each policy measure.

4. Linking the Program Framework toPolicy Measures, Poverty ImpactAssessment, and Adjustment Costs

A feature of the Uzbekistan EducationSector Development Program loanreport is the clear and explicit linkagebetween the Program’s logicalframework, the policy measures andconditions, the poverty impactassessment, including mitigation andenhancement measures, and the budgetimpact and adjustment costs identified.Table A3.7 exemplifies selectedlinkages from the ESDP’s frameworks.

TABLE A3.6: Financial Impact of Policy Measures($ million)

Policy Measure Costs Cost Savings

Implementation of an administrative staff redeployment plan 5.0 13.6Rationalization of the school network 5.4 61.7Increase in the teaching load 2.7 0.0Review of boarding facilities for SSE 0.0 92.6Introduction of cost recovery for boarding services (pilot basis) 0.0 1.5Improvement of teachers’ service conditions 17.2 0.0Transfer of supplementary resources to poor regions 21.5 0.0Introduction of SSE student allowances for the poor (pilot basis) 2.6 0.0Targeting of scholarships and stipends in higher education 0.0 39.2

Total 54.4 208.7SSE = senior secondary education.



AP

PEN

DIX 3

continued...

TABLE A3.7: Example of Linkages between Logical Framework, Policy Matrix,and Poverty Impact Assessment

1. Logical Framework

Design Summary Performance Monitoring Assumptions Targets Mechanism and Risks

Goal:

Objectives:

Outputs:

• To establish ageneraleducationsystemresponsive tothe needs of amodern market-orientedeconomy anddemocraticsociety

• Education structure stream-lined (Policy Matrix [PM] 1.1)

• Curriculum standardsreviewed and textbooksimproved (PM 1.2)

• Teachers trained in moderncurriculum (PM 1.3)

• Education planning,management, and super-vision improved (PM 3.1, 3.2)

• Staff redeployed and serviceconditions revised (PM 2.1,2.2)

• School facilities rationalizedand upgraded (PM 2.3)

• Remote rural schoolsrehabilitation and access tofinancial resources improved(PM 4.1)

• Nongovernment educationintroduced (PM 3.3)

• Governmentcommitment toefficiency improvement

• Consensus on newstaffing, workload,salary norms

• Counterpart fundavailability

• Distance Education(DE) systemestablished

• Access to reliable dataand statistics

• To improve thequality andefficiencyof the generaleducationsystem

• Strengtheningsector planningand managementcapacity

• Improving andextending teachereducation

• Strengtheningcommunity

• New school map adopted,organizational auditimplemented, principals anddistrict managers trained

• Policy studies in staff serviceconditions andredeployment, privateeducation, decentralizationand budget allocation

• Authorities collaborate,access to dataprovided,new functions agreedto

• Consensus on HRDissues, government/NGcooperate

• Qualified staff

• Universal access to basiceducation maintained

• Efficiency and effectivenessof basic education improved(main indicators brought upto international standards)

• Curriculum developmentcapacity strengthened

• Educational outcomesimproved

• Improved educationalpolicies and institutionalreforms implemented withintime frames provided in thepolicy matrix

• National legislationand policies oneducation

• Educational budgetsand statistics

• Reports of theeducation reformimplementationmonitoring unit(ERIMU) andmonitoring units of thevarious agencies

• Education SectorDevelopment Program(ESDP) policy matrixand efficiencyindicators

• Governmentcommitment andpolitical will to reformeducation

• Sustained budgetallocations to educationconsistent with thereform agenda

• Technical andmanagerial capacity todesign and manage thereform

• Coordination amongministries andinstitutions; individualincentives aligned withoverall priorities

• Educational budgetsand statistics

• New sector reformdecrees

• Reports of theERIMU andmonitoring units ofthe various agencies

• Midterm and finalproject reviews andevaluations

• Reports of the SIFregional facilitatorsand nationalcoordinator

• ESDP policy matrixand efficiencyindicators

• Government /Ministry of PublicEducation (MOPE)decrees orresolutions onservice conditions

• Studies• Midterm review /

evaluation

CASE 3

104

APPENDIX 3

2. Policy Matrix

Policy Areas and Actions Actions End ofMeasure Tranche 1 Tranche 2/3 Program Targets

PM 1: Modernize Structure, Contents and Processes of Education

1.1 Streamlining of • Submit projections for • Submit updated plans •Education structureeducation structure 2001–2005 of (i) student for SSE expansion streamlined

1.2 Curriculum review flows from primary to • Formulate strategic • Curriculum standardsand textbook higher education and directions for reviewed and textbooksdevelopment (ii) capital and recurrent restructuring higher development improved

1.3 Teacher education education expenses education • Teachers training1.4 Monitoring of broken down by • Maintain recurrent improved

education quality subsector and region budget allocation to • Education qualityeducation for fiscal monitoredyears 2003 and 2004

• Complete preparationof a master plan forhigher education

• Submit tracer studyof SSE graduates

PM 2: Improve Sector Sustainability and Efficiency

2.1 Staff redeployment • Prepare teacher • Organizational audit • Staff redeployed2.2 Service conditions number projections and redeployment plan • Teacher load increased

of education for 2001–2005 • Identify adjustment to 20 hrs/weekpersonnel • Policy studies on staff mitigation measures • New school map

2.3 Rationalization of service terms of (retraining, etc.) implemented and schoolfacilities reference conditions • Introduce new location network rationalized

and staff redeployment and performance• School and community reward systems

survey terms of reference • Prepare new schoolmap

Table A3.7 (cont’d.)

continued...

Design Summary Performance Monitoring Assumptions Targets Mechanism and Risks

• DE staff trained in programdevelopment, DE centersimproved, teacher trained

• School/community leaderstrained in self-management,school boards established,poorest schoolsrehabilitated, SIF introducedin poor areas

• Reports anddatabases on schoolmap and audit

• Administrative staffredeployment plan

• Reports on studyvisits and short-termcourses abroad

• Teachers training andDE modules

• Decree on theaccreditation of DEprograms

• Project monitoringsystem/database

identified and releasedfor training

• Budget available forDE units and civilworks

• DE accredited andaccepted

• Module writers haveincentives

• Local authoritiessupport legislation

• Effective targeting ofpoor communities andschools

• SIF facilitators andschool authoritiescollaborate

involvement withschools andimprovinglearningconditions



AP

PEN

DIX 3

PM 3: Reform Governance of Education

3.1 Policy formulation, • Establish Education • Program budgeting • Assess and considerplanning and Reform Unit test expansion of programfinancial • Policy studies for • Report on school budgetingmanagement budget allocation, boards and • Evaluation of SIF

3.2 School management decentralization mobilization of • Pilot project for schooland community • Legal basis for SIF extra-budget funds user charges initiatedparticipation • Determine modalities

3.3 Nongovernment for SSE user feeseducation

PM 4: Protect Poor and Vulnerable Population Groups

3.4 Targeting of poor • Education finance • Pilot test for poor • Rehabilitate 200 poorregions/districts for working group SSE student allowance schools and providesupplementary established • Define criteria for supplementary budgetbudget allocations • Rehabilitation plan stipends/scholarships for poorest districts

3.5 Assistance schemes prepared for 500 • Reduce proportion of • Proportion of budgetfor the poor schools in poor areas budget students to 25% students reduced to 20%

• Mechanisms identifiedfor additional resourcetransfer to areas belownational average

• Textbook and budgetneeds assessment

• Student loans decree

3. Poverty Impact Assessment

PM 2. Improve Sector Sustainability and Efficiency

Channel Direct Effect Indirect Effect Indirect Effect Other Enhancement andof Effect Short Run on the Short Run on the Medium Run on Stakeholders Mitigation

Poor Poor the Poor Affected Measures

Labor Direct short-run Indirect short-run School graduates Teachers assigned DistanceMarket effect during the effect during the better educated to remote and education

program period on program period on and skilled on economically for teacherslabor market and labor market and entry to the labor disadvantaged provided toreal wages for the real wages for the market with areas. remote areaspoor unlikely. poor unlikely. potential to Students in ($4.0 million).

command higher same areas. Compensationreal wages Negatively packages for(depending on affected: redundant staff.prevailing labor redundantmarket teachers andcircumstances). their dependents.

Prices PM 2.2 Cost of Better qualified Teachers assigned Supplementaryteachers increases teachers and to remote and incentive for(21,000 to 25,000 improved local economically teachers servingteachers in management of disadvantaged in disadvantageddisadvantaged schools reduces areas. areas ($3.9areas receive overall school Negatively million).


continued...

Policy Areas and Actions Actions End ofMeasure Tranche 1 Tranche 2/3 Program Targets

CASE 3

106

APPENDIX 3

special incentives). management cost. affected: studentsPM 2.3 Increased and families noservice cost to longer eligible forstudents no longer governmenteligible for subsidy (assumedgovernment nonpoor).boarding subsidies.

Access PM 2.1 Administrative Students in Teachers assigned Distanceto Public staff and teachers remote to remote and EducationServices have access to and poor areas economically Program for

professional have access to disadvantaged teachers.development improved school areas.

facilities and Students in samequality of areas.teaching.

Transfers PM 2.2 Transfers Intention to redirectof supplementary budget savings toresources to the poor.teachers ineconomicallydisadvantagedareas.

Analyti- PM 2.2 Teachercal motivation andBasis productivity isand/or enhanced byAssump- supplementarytions payments.

PMs 3 and 4: Reform Governance of Education and Protect Poor and Vulnerable Population Groups



Labor Direct short-run Indirect short-run School graduates CommunitiesMarket effect during the effect during the better qualified on and students in

program period program period entry to the labor economicallyon labor market on labor market market with disadvantagedand real wages and real wages potential to areas.for the poor for thepoor command higherunlikely. unlikely. real wages.

Prices PM 3.1, 3.2 Higher priced Establishment ofSchool education for SIF ($3.5 million).management nonpoor.costs reducedthroughimprovedplanning, self-


continued...





AP

PEN

DIX 3

management,privatization ofsupport services,incomegeneration andintroduction ofuser charges.

PM 3.2, 3.3Incrementalfinancial costto communities.

Access PM 4.1 Improved Communities Poor schoolsSupplementary participation and students upgradingbudget allocations rates and lower in economically program for 1,000provided to poor absenteeism disadvantaged schoolsareas, resulting rates in schools. areas. ($22 million).in improved Higherlearning achievementenvironment in of students inschools through beneficiaryrehabilitation schools.of school Improvedpremises. health status

of students.PM 3.3 Private/nongovernmenteducationintroduced,freeinggovernmentresources forthose unableto afford fees.

Transfers PM 4.1 Increased Communities Establishment ofbudget/ block and students in SIF ($3.5 million).grants provided economicallyto schools in disadvantagedeconomically areas.disadvantagedareas.

PM 4.2 Direct Negativelysupport to affected: studentspoorest pupils and communitiesthrough SIF no longer eligibleprogram, but for governmentreduced allowances.governmentsupport tononpoor.




continued...

CASE 3

108

APPENDIX 3

Policy Budget Impact, Costs and Benefits

Indicators Budget and Budget and Non-Policy Area and Measure Adjustment Economic monetized

Costs Benefits Benefits$‘000 $‘000

PM 2.1 Staff redeployment

• Compensation package for redundant staff• Retraining costs 1,113• Reallocation costs of transferred staff 3,761• Auditing and training needs assessment 157• Payroll reduction as a result of staff costs 13,619

PM 2.2 Revision of service conditions

• New promotion mechanisms 6,565• Incentives for teachers assigned to poor areas 3,979• Incentives for multigrade teachers 2,064• Salary increases for teachers with

DE certificates 4,615• Increase in teacher workload/ salary bill

augmentation 2,723

PM 2.3 Facility rationalization

• Rationalization of the school network• Compensation package for redundant teachers 3,955• Retraining costs for multigrade teachers• School mapping and community survey 1,423

• Payroll reduction 57,549• Reduction in school operating expenditure 4,166• Reduction in boarding costs in SSE institutions 92,663

PM 4.1, 4.2 Targeting of disadvantaged areas

• Transfer of supplementary resources todisadvantaged regions 21,457

• Introduction of SEE student allowances forthe poor (pilot) 2,619

• Targeting of scholarships and stipends inhigher education 39,191

TOTAL 54,435 208,686HRD = human resource development, NG = Non-government, PM = policy matrix, SIF = school initiatives fund, SSE =senior secondary education




AP

PEN

DIX 3

D. CASE 4. Philippine Grains SectorDevelopment Program: PartialEquilibrium Rice Model


The program loan report for the GrainsSector Development Program (GSDP)(Loan 1739-PHI) indicates that in thePhilippines, agriculture directlycontributes about 22% of GDP and agri-based industries another 13%.Agriculture also contributes about one-third of the value-added in the servicessector, and provides income to about43% of the Philippine labor force. Riceand corn alone contributed nearly 28%of agricultural value-added in 1995 withrice contributing two-thirds of thatfigure. Despite an increase in paddyproduction, and the introduction ofimproved technologies, net returns torice farming have steadily declined inrecent years, with an inability of farmersto improve individual productivity andfinancial viability. Farmers have alsobeen unable to move into higher-valuecrops. Overall, the constraints toimproving grains productivity can berelated to physical factors; farm-levelconstraints, including technicalinefficiencies; and allocativeinefficiencies, due to high transactioncosts. In turn, these are related to aninappropriate policy regime, includingprice distortions; structural bottlenecks,such as lack of infrastructure and otherpublic goods and services; and limitedinstitutional capacity. The NationalFood Authority (NFA) is the agencytasked with price policy implementation.

NFA is a government-owned andcontrolled corporation that is mandated

to stabilize rice prices in the Philippines.Since its inception, it has attempted tostrike a balance between the conflictingrequirements of three principal groupsin the grains subsector—farmers, traders,and consumers. NFA’s policies are aimedat responding to several, albeitconflicting, policy objectives, whichinclude the following: (i) stable year-round rice prices, (ii) affordable riceprices, particularly for the country’s poor,(iii) stable paddy prices at levels thatenable rice farmers to attain reasonablelevels of income, and (iv) sustaineddevelopment of an integrated grainsindustry. In fulfilling these objectives,NFA has become directly involved in thedomestic procurement of rice, bufferstocking, and importation.

The program loan report for thePhilippines GSDP argues that the policyinconsistencies, stemming from NFA’sexercise of both regulatory andmarketing/trading functions, have led tosignificant financial losses, a high degreeof retail price instability, and lower farm-gate prices. NFA has attempted toprotect both consumers and producersthrough its strategy of “buying high andselling low”, which has resulted infinancial losses over the years. The floorprice (i.e., the official price used by NFAto procure at the farm-gate level) is setrelatively high while the ceiling price (orNFA’s official price for distributing itsrice supply to rice consumption centers)is set relatively low. Consumers payprices that are 35–100% higher thanwould be possible under a free traderegime. This is attributed to the importban and the fact that NFA imports andreleases fail to make up for what wouldbe achieved by private traders. NFA

CASE 4

110

APPENDIX 3

determines the volume of rice importswhile, at the same time, serving as thesole importer of the commodity.Moreover, this “protection” against riceimports does not confer commensuratebenefits on the producers (Roumasset2000).6 Roumasset attributed thediscrepancy between official andequilibrium prices to the fact that NFAis a relatively small player in the totaldomestic rice market. NFA’s releasesaveraged around 11% of riceconsumption from 1995 to 1998.Related data on NFA procurementshowed an average of less than 1% ofdomestic rice production during thesame period (Table A3.8).

2. Options for NFA’s Reorganization

Two opposing schools of thought haveemerged from the debate among thestakeholders on the direction, extent, andmagnitude of NFA’s reorganizationwithin the context of the country’s foodsecurity program and current trends inglobalization. The “status quo” schoolcontends that NFA plays a critical rolein achieving food security and that the

proposed decoupling of NFA’s regulatoryfunction from its trading/marketingfunctions will result in welfare losses onboth consumers and producers. Whilesome operational reforms may beundertaken, NFA’s basic functions andstructure are proposed to be retained,especially its role in the domestic ricemarket. The opposing school of thoughtargues for decoupling to reduce thepresent level of inefficiencies in the ricemarket, based on the premise that theprivate sector can do a more efficient jobin the area of rice trading. Thus, theGovernment may, instead, opt to focusits efforts on providing the rightincentives to the private sector and onappropriately targeting food subsidies tothe poor.

There are two versions of the “statusquo” option in reorganizing NFA.Option A basically calls for a retentionof the present NFA structure, includingits expansion of coverage to othernongrain food commodities (e.g.,sugar, corn, and fertilizer). Its variant,Option B, seeks to retain both theregulatory and trading/marketingfunctions of NFA. However, coveragewill only focus on the rice commodity,since it is argued that markets for the

6 This section of the report draws heavily fromRoumasset (2000) and AGILE (2000).

TABLE A3.8: NFA Rice Distribution and Procurement, 1995–1998(thousand metric tons)

1995 1996 1997 1998 Average

DistributionQuantity (‘000 MT) 256.7 731.4 622.8 1,628.2 809.8

% of Consumption 3.6 9.3 7.9 22.2 10.7 Procurement

Quantity (MT) 8.2 124.3 100.5 97.4 82.6 % of Production 0.1 1.1 1.0 1.1 0.8

Source: Roumasset 2000.



AP

PEN

DIX 3

other nongrain food items have alreadybeen liberalized.

The GSDP (or Option C) proposed,among other measures, liberalized, morecost-effective grains pricing and importpolicies, and a decoupling of NFA’sregulatory function from its trading/marketing functions and institutionalizea transparent process for food securitypolicy implementation. Less governmentintervention is expected to spur greatercompetition, which would, in turn, lowerprices in the rice market to the benefit ofboth producers and consumers.Moreover, government savings can bechanneled to other basic support servicesand greater provision of infrastructurein rural areas.

Option D was an offshoot of the reviewundertaken and an extension of theschemes under Option C. It proposedthe following policy reforms:(i) implementation of a rice pricestabilization scheme without bufferstocks; (ii) creation of a National RiceBoard (NRB); (iii) transfer of NFA’sduplicative functions to appropriategovernment units; (iv) full privatizationof NFA’s proprietary functions,including the unrestricted sale of itsbundled assets to the private sector; and(v) adoption and implementation ofcomplementary grains and policyinstitutional reforms. These measuresare expected to ease the pressure on thegovernment’s budget and to contributetoward societal gains and welfare.

3. Agency Cost Approach

Roumasset (2000) attempted an agencycost approach to quantify NFA’s current

intervention mechanisms. In economics,agency costs pertain to the costs ofstructuring, monitoring, and complying/enforcing a set of contracts withconflicting interests, including theinefficiencies or losses that remain whenthe contracts are optimally, but not fully,enforced (Jensen and Meckling 1976).The central proposition of agency theoryis that rational, self-interested peoplealways have incentives to reduce orcontrol conflicts of interest so as tominimize the transaction costs thatthese conflicts engender. This impliesthat gains from minimizing conflictsand inefficiencies can be shared by thestakeholders/parties involved. It isposited that institutional structures,contracts, and informal arrangementscan be created to reduce conflicts, governrelations, and increase the extent ofcooperation and the benefits from it(Jansen 1994). For example, firms area form of internalization of a certaincategory of market transaction costs, forwhich vertical integration incurs fewertransaction costs.

Alternative government interventionsystems can be viewed as a set of contractsbetween the government (acting as theprincipal) and its agents (represented bymanagers, employees, contractors, andother stakeholders employed) to achievecertain objectives. In this case, agency costcovers structuring, monitoring, andcompliance/enforcement (or “bonding”)costs associated with establishing andoperating each government interventionsystem, plus the inefficiencies that stillremain despite the intervention schemeemployed. It then follows thatminimizing agency cost is equivalent tomaximizing net public gains and hence,

CASE 4

112

APPENDIX 3

can be considered as a decision rule indesigning an optimum food policyprogram. The greater the number ofcontracts that have to be enforcedbetween the government and its agents,the higher the structuring, monitoring,and compliance/enforcement costs(which can be collectively termed“organization costs”) that may beincurred to ensure that pertinentobjectives can be achieved by a givenintervention system. An increase inorganization costs will cause a subsequentdecrease in social costs in the form ofreduced welfare inefficiencies anddistortions in the rice market. Therefore,agency cost can be defined as the sum ofthese organization costs and residualsocial costs/losses (e.g., producer andconsumer welfare losses and opportunitycosts such as foregone fiscal revenues)for each type of intervention system(AGILE 2000).

In the context of the Philippine ricemarket, agency costs emanate from the

need for information, monitoring,compliance/enforcement, and otherattendant organizational activities thatare being undertaken to ensure adequacyin the supply of rice. Each alternativegovernment intervention system connotesvarying, and often redundant, organizationcosts. Thus, an increase in organizationcosts will bring about a concomitantincrease in social benefits in the form ofimproved welfare among rice consumersand rice farmers. However, the implicationof this analysis is that it may not be optimalto completely eliminate the inefficienciesin the rice market since it will entailinordinate organization costs. These willonly cause fiscal balances to deteriorateand result in diminishing marginal benefits(i.e., reduced residual costs/losses).Therefore, on efficiency grounds, the mosteffective system of government is theintervention that minimizes agency cost(point S* in Figure A3.1) (AGILE 2000).The following subsections summarize howAGILE analyzed and deducted thisconclusion.

Source: AGILE Consultnats, 2000. Strategic Reorganization of the NFA for the New Millenium.



AP

PEN

DIX 3

a. Agency Cost Estimation without theDecoupling of NFA’s Functions(Status Quo)

A prerequisite in the design of a foodsecurity program in the Philippines is theestimation of a cost-effective modalityon NFA’s reorganization. Broadly, thiswill involve the following steps asindicated in the AGILE study:(i) estimation of the organization costor the fiscal burden of alternative options;(ii) estimation of the inefficiencyindicators (i.e., residual inefficiencies)resulting from each NFA reorganizationscheme; (iii) derivation of the agencycosts by summing up (i) and (ii); and(iv) selection of the option that generatesthe leas-combined costs (i.e., the optionwith the lowest agency cost). As anillustration, in 1996–1998, the AGILEconsultants estimated the agency costsof maintaining the status quo in NFA.Estimates for 1997 are indicated below.The same methodology can be adoptedto estimate the agency costs for the otheryears.

b. Estimation of Organization Cost

A good indicator of organization cost isthe fiscal burden (including its impacton the credit sector) of operating a

delivery system that will meet foodsecurity objectives. Table A3.9 sets forthan estimate of the fiscal outlays incurredby the Government for the year 1997.

c. Estimation of Rice Price RegulationInefficiency Indicators

The residual welfare inefficiencies of eachNFA reorganization option can bemeasured by foregone tariff revenues,producer and consumer losses, and otherwelfare losses (resulting from reducedquantities purchased by consumers oroverproduction by farmers, due to priceor quantity restrictions imposed on therice market by the governmentintervention system). Roumassetestimated the following components ofrice market inefficiencies (both demandand supply curves are estimated asstraight lines for simplicity; see FigureA3.2 and Box 9):

In 1997, rice production was estimatedat 7.3 billion kilograms (kg), at aproducer price of P14.48/kg (includingthe conversion of palay (unhusked rice)to rice and marketing costs up to thewholesale in situ (warehouse point). Riceconsumption was calculated at 7.9 billionkg, at a consumer price of P15.31/kg,or about 55% above the estimated border

TABLE A3.9: Estimated Financial Burden of NFA Operations in 1997(billion pesos)

Item Financial Burden

Operational Subsidy (Budget) 1.5Equity Infusion 0.1Increase in Debt 3.9Less: Increase in Value of Rice Stocks 1.5Total Financial Cost 3.9

Source: AGILE Consultants. 2000. Strategic Reorganization of the NFA for the New Millennium.

CASE 4

114

APPENDIX 3

price of P9.90/kg. Rice imports ofaround 0.6 billion kg would have beenrequired to bridge the gap between locallevels of production and consumption.This infusion of additional supplythrough imports would have resulted ina domestic price of about P14.10/kg,which is equivalent to the equilibriumprice of rice without NFA intervention.On paper, rice producers (including thosein the trading/marketing segment) andconsumers should have been facing thisnew equilibrium price of rice (i.e.,P14.10/kg). However, actual conditionsin the rice market indicated thatconsumers were paying more(approximately P15.31/kg) andproducers were receiving less (aboutP13.48/kg). The difference between

consumer andfull-producer rice prices wasattributed to inefficiencies and non-competitive elements in rice marketing asa result of NFA policies and interventions.

Alternatively, a combination of a tariff,limiting rice imports to the supply gap(i.e., 0.6 billion kg), and taxes on bothproducers and consumers to lowerconsumer and producer prices, could haveachieved the same equilibrium price (i.e.,P14.10/kg) without the need for NFAintervention.

d. Estimation of Agency Cost

Roumasset estimated NFA’s agency costto be about P29.1 billion in 1997. TableA3.10 sets out the fiscal burden (i.e.,




AP

PEN

DIX 3

organization cost) and the variouscomponents of welfare losses (i.e.,residual inefficiencies). Agency costs ofNFA intervention were estimated to beP30.7 billion and P18.2 billion, in 1996and 1998, respectively.

Option A, which calls for an expansionin NFA’s current commodity coverage,is deemed not to be cost effective sinceit will create more inefficiencies anddiseconomies. While no formal

estimates were presented, NFA’sexpanded intervention will entailproportionate increases in fiscal outlaysand operational and welfare losses dueto trade restrictions and price controlsin the other commodities that areproposed to be covered. On the otherhand, the status quo (Option B), whichwill limit NFA’s intervention to the ricemarket only will, on the average, incuragency costs of about P26 billion (TableA3.10).

BOX 9: Residual Welfare Inefficiencies

Foregone Tariff Revenues: FTR = (the price of the permit to import a unit of rice or the differencebetween the wholesale price/kilogram and the border price of rice/kilogram) x (volume of importedrice or difference between local consumption and production levels), or the difference between theactual price of rice resulting from NFA intervention and the border price of rice. The Governmentwould have received this difference in the form of tariff revenues if rice import permits will be auc-tioned off to the private sector. Therefore:

FTR = P(15.31–9.90)/kg x (7.9–7.3) billion kg = P3.4 billion

Consumer Surcharges: CS = (difference between the actual price of rice resulting from NFA inter-vention and what could have been the equilibrium price of rice without NFA intervention) x (volumeof local rice production), or the penalties borne by rice consumers as a result of the inefficient ricemarketing system arising from the absence of a transparent and credible rice price policy. Hence:

CS = P(15.31–14.1)/kg x (7.3) billion kg = P8.9 billion

Producer Losses: PL = (difference between the equilibrium market price of rice and the actualfarm-gate prices received by producers as a result of NFA intervention) x (volume of local rice pro-duction), or the losses incurred by rice producers as a result of an inefficient rice marketing system.Thus,

PL = P(14.10–13.48)/kg x (7.3) billion kg = P4.5 billion

Excess Burden for Consumers: EBC = (one-half of the price difference between the border anddomestic market rice prices) x (difference between the quantity of rice that would have been con-sumed at the border price and actual consumption level), or the welfare losses arising fromconsumers buying lesser rice than they would have without the import restrictions on rice.

EBC =(1/2) x P(15.31–9.90)/kg x (9.58–7.9) billion kg = P4.5 billion

Excess Burden for Producers: EBP = (one half of the price difference between the border anddomestic market rice prices) x (difference between the quantity of rice produced if producers didnot receive any trade protection and actual production level), or the welfare losses imposed on thesociety from the rice producers’ expansion of their production to more than the amount they wouldhave produced if there were no import restrictions on rice.

EBP =(1/2) x P(13.48–9.90)/kg x (7.3–5.12) billion kg = P3.9 billion

CASE 4

116

APPENDIX 3

e. Agency Cost Estimations byDecoupling

Options C and D can be justified if itcan be shown that the decoupling ofNFA’s regulatory and marketingfunctions can lead to a lower agency cost(Figure A3.3). Operationalizing theseoptions will involve the full privatizationof NFA’s marketing functions and arefocusing of government efforts onregulatory activities. On the fiscal side,

these are expected to result in reducinggovernment subsidies, equity infusions,and credit guarantees. Private sectorresource mobilization is also expectedto be enhanced in the rice marketingactivities.

f. Estimation of Organization Cost

The setting up of National Rice Board(NRB) under Option D will require amuch lower financial outlay of about

TABLE A3.10: Estimated Agency Costs of NFA’s Intervention System, 1996–1998(billion pesos)

Item 1996 1997 1998 Average

A. Total Residual Inefficiencies 31.2 25.1 12.0 22.8 Foregone Tariff Revenues 4.3 3.3 3.6 3.7 Consumer Surcharges 6.4 8.9 4.7 6.7 Producer Losses 3.4 4.5 3.1 3.7 Excess Burden for Consumer 7.6 4.5 0.6 4.2 Excess Burden for Producer 9.4 3.9 0.1 4.5

B. Total Fiscal Burden (0.5) 3.9 6.2 3.2

Total Agency Cost (A+B) 30.7 29.0 18.2 26.0




AP

PEN

DIX 3

P116 million, as compared with NFA’sbudgetary outlays of P3.9 billion andP6.2 billion, in 1997 and 1998respectively (Table A3.10). However,there is no currently available estimateof the contingency fund that will haveto be added to the budget of theproposed NRB. Under Options C andD, budgetary resources will be requiredfor the envisaged targeted foodassistance scheme. Estimates range fromP1.5 billion to P2.0 billion to supportthis program and are derived by takingthe difference between the NFA-releaseprice and the wholesale price andmultiplying this with NFA’sdistribution volume from 1995 to1998. AGILE’s estimate of fiscal burdenwas about P1.7 billion for Option D.

g. Estimation of Rice Price InefficiencyIndicators

A more transparent rice policy regime isexpected to remove the wedges betweenthe prices that the consumers pay andthe prices that producers receive. Residualinefficiencies, in the form of consumerand producer losses, are still expected toexist in view of the imposition of animport tariff on rice that will continueto create a gap between the world price

and the domestic price. However, a morecompetitive rice market is expected toensue as tariff protection decreases overtime, thus minimizing residualinefficiencies. Estimated residualinefficiencies under Option D are shownin Table A3.11.

h. Estimation of Agency Cost

Using 1997 as an example, it can beshown that the estimated agency costwith the decoupling of NFA’s functionsis much lower (about P9.4 billion) thanthe agency cost of around P29 billionincurred under the present NFA systemof operation (Tables A3.11 and A3.12).

However, a political economy challengeis to find which components of theproposed decoupling modalities areresisted by whom and to what degreesome groups will be negatively affected.Compensatory and safety-net programsmust be designed for those who willsuffer short-term economic dislocations(AGILE 2000, p. 61).

i. Partial Equilibrium Analysis

The economic impact of the proposedpolicy reforms (para. 93 of the GSDP

TABLE A3.11: Estimated Residual Inefficiencies Under Option D, 1996–1998(billion pesos)

Item 1996 1997 1998 Average

Total Residual Inefficiencies 7.6 7.7 1.1 5.4 Foregone Tariff Revenues 0 0 0 0 Consumer Surcharges 0 0 0 0 Producer Losses 0 0 0 0 Excess Burden for Consumer 2.8 2.9 0.4 2.0 Excess Burden for Producer 4.8 4.8 0.7 3.4


CASE 4

118

APPENDIX 3

study) to NFA’s charter on procurementpricing, release pricing, and rice importswere analyzed using a partial equilibriummodel of the Philippines’ rice market,covering an 8-year period.

Table A3.13 sets out the structuralequations and identities used in themodel. Of the 15 equations, 10 areidentities and 5 are behavioral equations.Parameters of the model (not reportedhere) are estimated from historical data.In the model, prices received by farmersare linked with wholesale prices on theassumption that wholesalers deducttheir costs and then offer farmers theresidual. Area harvested is determinedby seasonal variations in growingconditions, which are incorporated bydummy variables for each quarter of theyear, and climatic conditions, which areallowed for by a rainfall variable. Totaldemand is made up of demand for foodpurposes, feed, seeds, changes in stocks,and exports. It is determined by riceoutput and the relative price of rice.Exports are given exogenously. Ricesupply is composed of rice output,stocks at the start of a quarter, andimports, which are also exogenous. Inthe model, area harvested is not relatedto the farm-gate price of paddy,although yield per hectare is, through

the relative price variable that comparespaddy and fertilizer prices. This is seenin the policy simulations below wherechanges in farm-gate prices have animpact via yields, not area harvested.

The model works through theequilibrium condition that, each quarter,wholesale prices adjust to clear themarket. If there is excess demand in aparticular quarter, prices are adjustedupward using an iterative process.Similarly when there is excess supply,prices are adjusted downward iteratively.However, price adjustment for eachquarter are kept within the historicalrange of proportionate changes from onequarter to another to avoid unrealisticprice swings. Any uncleared excessdemands after exhausting all priceadjustments within the historical range,are subtracted from stocks; similarly, anyexcess supplies are added.

The model is used in a quantitativeanalysis of two aspects of the policy loan(although this section of theconsultants’ report is not used in thefinal program loan report). The policychanges tested with the model involvedchanges to the procurement pricingpolicy of the NFA to offer farmers aseasonal premium to the price of rice

TABLE A3.12: Agency Cost Estimates of the Proposed Decoupling of NFA’s Functions, 1997(billion pesos)

Component Cost

Economic Welfare Inefficiencies 7.7Fiscal Burden 1.7

Total Agency Cost 9.4



AP

PEN

DIX 3

TABLE A3.13: Partial Equilibrium Rice Model

Equation 1: (Behavioral equation)

Pft = Constant + Log Pr

tFarm-gate paddy price Wholesale rice price


At = Constant - RFt + ASt + CIt -DV1 -DV2 -DV3Area harvested Rainfall Seasonal index * Dummy quarter 1 Dummy quarter 2 Dummy quarter 3


Yt = Constant -FPPR -DV1 + DV2 -DV3 +TtYield Fertilizer price/ Dummy quarter 1 Dummy quarter 2 Dummy quarter 3 Technology index

Paddy price

Equation 4: (Identity)

Qpt = At.Yt

Paddy output Area harvested times yield


Qrt = 0.654Qp

tRice output 0.654 times Paddy output


BSrt = ESr

t-1Beginning of quarter rice stock End of previous quarter rice stock


ESrt-1 = Qr

t (constant -RPrt +DV1 +DV2 +DV3 - Pr

t) -uEnd of previous Rice CPI Dummy Dummy Dummy Whole Un-

quarter rice stock output deflated quarter 1 quarter 2 quarter 3 sale cleared real rice excess

wholesale price demand rice price for price


Mt = MBARtRice Imports Policy determined rice imports


Rst = 0.075At

Rice for seed 0.075 times area harvested


Rft = 0.04251Qr

tRice for feed 0.04251 times rice output


Xt = XBARt Rice exports Exogenously determined rice exports

continued...

CASE 4

120

APPENDIX 3

imports as a means of stabilizing pricesand reducing shortages.

The case for seasonality in riceprocurement pricing arises from the factthat there are two distinct seasons—onedry (March to August harvest) and theother wet (September to Februaryharvest)—and the cost of productionand yield, as well as of storage, variessignificantly between these two seasons.The dry season period is one of relativelyscarcity when market prices tend to goup and the seasonal procurement

pricing, with a dry season premium,would encourage farmers to increase thedry season harvest.

Rice imports in the Philippines havebeen made on an ad hoc basis in light ofdomestic supply conditions, with theconsequence that they have often arrivedat the wrong time in the agriculturalcycle. The consultants recommendedformalizing a decision on imports by 30April each year, as well as allowing theprivate sector to import a proportionof the import quota.


PCRct = (constant + Qr

t - RPt ) aPer capita rice Rice output CPI deflated real Adjustment toconsumption wholesale rice calibrate estimated

price consumption tohistorical levels


Rct = POPt PCRc

tRice consumption Population times per capita rice consumption


DEMt = Rct + Rs

t + Rft + ESr

t + XBARtTotal rice use Rice consumption Rice for seed Rice for feed Rice stocks Rice exports


SUPt = Qrt + BSr

t + MBARtTotal rice supply Rice output Beginning of quarter rice stocks Rice imports

Equilibrium condition (1):

DEMt = SUPtTotal rice use Total rice supply

Equilibrium condition (2):

Historical oL < (PRt /P

Rt+1) -1< Historical oh

Notes: *CI is not defined in the source.Equilibrium condition (2): Price adjustments were confined to historical range of proportionatechanges from quarter to quarter.t refers to quarters.

Source: DAI Consultants. 1998. Philippines Grains Sector Development Program Project. Final Report, Vol. 1,Chapter 2 (ADB TA 2717-PHI).

Table 13 (cont’d.)



AP

PEN

DIX 3

Of these policy changes, the one thatallows a regular flow of imports ishandled very simply in the model bysetting the imports at 300,000 metrictons (or approximately 5% of domesticoutput) each year for the projectionperiod 1997–2004. Hence, MBAR inequation 15 is set at 300,000 annually.The timing of import arrivals during theyear is not made clear in the text. Themethod of allowing for the changes inNFA procurement prices is even moreunclear however (particularly because itis not given in equational terms). Itappears from the text that equation (1)is re-specified to make farm-gate pricesfor paddy a function of the NFAprocurement price—as well, it seems, asthe wholesale price, but the variablesincluded in the new version of equation(1) are not discussed. The NFA price isincreased above current levels to includea seasonal premium presumably takenfrom historical data on the averagerelation between market prices for paddyin the dry and wet seasons. The adjustedNFA price is then projected into the future.

These two adjustments work in oppositedirections. The higher NFAprocurement price increases the pricereceived by farmers for paddy relativeto what it would be without the seasonalpremium. Higher paddy prices increaseyields due to their assumed impact viahigher fertilizer use (see equation 3) andthis increases output since area harvestedis assumed to be unchanged. However,this upward pressure on farm-gate pricesfrom the change in procurement policyis swamped in the model by thedepressing impact on prices of higherrice imports (the level of rice importswithout the policy reform is not stated

in the text). Higher imports depresswholesale prices through their impacton supply (via equation 16) and lowerwholesale prices mean that farm-gateprices for paddy are also depressed.Lower farm-gate prices mean lower riceoutput, as yields respond positively topaddy prices (see equation 3). Hence,in the policy reform scenario, both farm-gate prices and domestic rice output arelower than they would be withoutreform. However, wholesale prices andhence, prices also paid by consumers arelower than they would be without reformand consumption of rice would behigher. Tables A2.14 and A2.15summarize the model results for thewith- and without-reform scenarios,showing annual income changes for riceproducers, who loses from the reforms,and for rice consumers, and who gains.

The logic of the analysis can be illustratedin Figure A3.4 using a simple competitiveframework and linear demand and supplycurves. The analysis is slightlycomplicated, since in the rice model, thesupply line refers to paddy output whilethe demand line refers to the final goodrice. Hence, we show supply and demandlines in different diagrams. Producers facea lower paddy price (from Pf1 to Pf 2)and a falling paddy output (from Qp1 toQp2). Producer losses are, therefore, acombination of a lower price for theirnew output

Qp2* (Pf1-Pf 2)

and the net loss as a result of their fallin output

0.5((Pf1-Pf 2)* (Qp1 - Qp2)).

CASE 4

122

APPENDIX 3



AP

PEN

DIX 3

This latter term allows for the fact thatalthough output is lost, there is also acost saving so that the net loss is onlyequivalent to the triangle ABC inFigure A3.4.

A similar analysis applies to consumers.Consumers face a lower rice price (fromPr1 to Pr 2) and a rise in consumption(from Rc1 to Rc2). Consumers gainfrom the lower rice price for theiroriginal consumption

(Rc1*Pr1-Pr2)

plus their consumer surplus from theiradditional consumption

(0.5*(Pr1-Pr2)*(Rc2-Rc1)).

This latter term is the differencebetween what they actually pay for theadditional rice and what they are willingto pay. It is shown as the triangle XYZin Figure A2.4. The analysis assumesthat none of the change in wholesaleprice of rice is captured by traders inhigher profit margins at the retail stage.

Tables A3.14 and A3.15 show theannual income changes for riceproducers and consumers respectivelyfor the period 1997–2004. These arediscounted at a 12% rate to give presentvalues for income changes resulting fromthe policy intervention. In present terms,producers lose P46.8 billion, whileconsumers gain P49.5 billion. There is,therefore, a net gain to society of nearlyP3 billion. However, from a povertyperspective, the important issue is whatproportion of the income changes forconsumers and producers accrue tothose below the poverty line.

A poverty profile of consumers andproducers is not given in any detail inthe consultants’ report. However, somerelevant data can be extracted. First,there is an estimate taken from anothersource that for 1995, the poor take47% of national rice consumption(DAI Consultants 1998, Table 4.13).In relation to production, no direct dataare available. However, some inferencescan be drawn. There are statistics onearnings per hectare on paddy farmstaken from the Bureau of AgriculturalStatistics (Table 2.6, consultants’report). These show an average incomefor paddy farmers (including off-farmincome) of P57,899 in 1995. At thethen exchange rate (P39/$), this is$1,484 per family per annum.Assuming six family members, this isabout $247 per annum per person orwell below the minimum $1 a dayinternational poverty line. Hence, onthe average, paddy farmers will be belowthe poverty line, so the expectation isthat a majority will be poor.

A further insight can be gained byconsidering net income per hectare datafor paddy, again taken from the Bureauof Agricultural Statistics (Table 2.9,consultants’ report). Tables A3.15 andA3.16 give the net income per hectarefigures and the number of hectaresrequired to reach the $1 a day povertyline (assuming six family members).For irrigated farms, approximately 17hectares are needed and fornonirrigated, 21 hectares. Data on landdistribution are not presented in theconsultants’ report, but if we knowwhat proportion of land is farmed inunits of less than these sizes, we canmake approximate inferences on the

CASE 4

124

APPENDIX 3

share of rice production grown by poorfarmers.

While not without some apparenttechnical problems, Table A3.13 is an

illustration of what can be done using apartial equilibrium approach. Additionalinformation is needed to convert thisinto a full general equilibrium study butit provides a helpful case illustration.

TABLE A3.14: Rice Production With and Without Policy Reforms

Paddy PaddyFarm- Farm- Fall in Fall in Fall in

Paddy gate Paddy gate Paddy Paddy IncomeOutput Area Price Output Area price Output Price for Paddy‘000 MT Harvested Yield P/kg ‘000 MT Harvested Yield p/kg ‘000 MT P/kg Producers

Year (Q2) ‘000 ha MT/ha (P2) (Q1) ‘000 ha MT/ha (P1) (Q1-Q2) (P2-P1) P million

With Reform Without Reform

1997 11,402 3,906 2.918 9.1 11,440 3,906 2.9 9.7 38 0.6 6,8751998 11,467 3,906 2.935 10.1 11,505 3,906 2.9 10.7 38 0.7 8,0671999 12,214 4,138 2.951 11.2 12,254 4,138 3.0 11.9 40 0.7 9,2062000 11,629 3,906 2.976 12.0 11,652 3,906 3.0 12.6 23 0.6 7,2312001 11,647 3,906 2.982 13.5 11,681 3,906 3.0 14.5 34 1.0 11,1132002 11,702 3,906 2.996 15.0 11,735 3,906 3.0 16.0 33 1.0 11,8692003 11,753 3,906 3.009 16.5 11,785 3,906 3.0 17.5 32 1.0 12,2732004 11,593 3,837 3.021 18.0 11,623 3,837 3.0 19.1 30 1.1 12,802

Present value of producer loss at 12% = P46,759 million

Note: Loss to producers = (P1-P2)*Q2 + 0.5(P1-P2)*(Q1-Q2), where P1 is original paddy price, P2 is new paddy price, Q1is original domestic supply, Q2 is new domestic supply.

Source: DAI Consultants. 1998. Philippines Grains Sector Development Program Project. Final Report, Vol. 1, Chapter 2(ADB TA 2717-PHI).

TABLE A3.15: Rice Consumption With and Without Policy Reforms

Year Rice Rice Rice Rice Rise in Fall in Gain inConsumption Wholesale Consumption Wholesale Paddy Wholesale Income for‘000 MT (C2) Price P/kg ‘000 MT Price P/kg Consumption Price P/kg Rice

(P2) (C1) (P1) ‘000 MT (P1-P2) Consumers(C2-C1) P million

With Reform Without Reform 1997 7,191 18.4 6,912 19.5 279 1.0 7,228 1998 7,292 20.3 6,996 20.7 296 0.4 2,808 1999 7,587 22.6 7,295 24.0 292 1.4 10,127 2000 7,558 24.3 7,320 25.5 238 1.3 9,336 2001 7,466 27.5 7,129 29.4 337 1.9 13,741 2002 7,472 30.5 7,144 32.5 328 1.9 14,104 2003 7,499 33.6 7,177 35.6 322 2.0 14,669 2004 7,441 36.9 7,117 39.0 324 2.1 15,199

Present value of consumer gain at 12% = P49,500 millionNotes: Gain to consumers = (P1-P2)*C1 + 0.5(P1-P2)*(C2-C1) where P1 is original wholesale rice price, P2 is new

wholesale rice price, C1 is original consumption, C2 is new consumption.Some data from the original report have been modified to avoid an apparent inconsistency in the 1997 values.

Source: DAI Consultants. 1998. Philippines Grains Sector Development Program Project. Final Report, Vol. 1, Chapter 2(ADB TA 2717-PHI), Tables 4.8 and 4.10.



AP

PEN

DIX 3

Overall, as described in the program loanreport, the model indicates that thecumulative impact of the reforms arehigher rice yields, farm-gate prices, andtotal rice production, and lower pricesfor consumers, with a marginal increase

in rice imports. Aggregate benefitsinclude a combined income gain ofproducers and consumers of $0.2 billionover the simulation period, and an 11%income gain for the poor throughtargeted subsidies compared to thecurrent 1.4%.

All 4,129 105.9 17.7 20.7Irrigated 5,018 128.7 21.4 17.0Nonirrigated 4,023 103.1 17.2 21.2

Note: Calculations are based on an exchange rate of P39/$, 365 days per year, and an assumption of six members perfamily.

TABLE A3.16: Net Income from Paddy Farming, 1995

Annual Net FamilyIncome (P/ha)

Annual Net FamilyIncome ($/ha)

Annual Net Incomeper Person ($)

Hectares Needed for$1 a Day per Person

E. CASE 5. Nepal: Second IndustrialSector Study: Computable GeneralEquilibrium Modeling for Trade andIndustrial Policy Measures

1. Relevance of CGE Analysis

As introduced in Chapter 4, computablegeneral equilibrium (CGE) analysis isapplied to a broad range of policy issues,which include the following:(i) exogenous shocks (e.g., increase inthe price of imported oil or a decline inthe price of a country’s main exports);(ii) economic policy changes (e.g.,changes in the size and composition ofa government’s current expenditures andinvestment); and (iii) changes in thedomestic economic and social structuresuch as technological change inagriculture, asset redistribution, andhuman capital formation (Sadoulet andde Janvry 1995). A CGE model is an

aggregate representation of the economyand provides quantitative evaluations ofthe outcomes of these policy changes.It simulates the behavior of a marketeconomy wherein quantities and pricesadjust to clear all markets. CGE modelstake into account the interactions of theeconomic agents (i.e., households, firms,government, etc.) and reflect the circularflow of income or resources among theseagents. Because they emphasize theimpact of reallocating resources acrosssectors of an economy, these models areuseful tools for identifying gainers andlosers under a policy change (Kehoe andKehoe 1994).

In essence, a CGE model attempts tocapture the sets of relative prices, initialconditions, and quantities thatcharacterize a general equilibriumcondition in an economy. The decisionsof economic agents are based on their

CASE 5

126

APPENDIX 3

7 The World Bank’s Integrated MacroeconomicModel for Poverty Analysis (IMMPA) Project (WorldBank 2001a) is an attempt to merge this withmonetary economic models.

responses to prices. These result in levelsof production for each economic agentso that demand equals supply in allproduct and factor markets (i.e., thereare no excess demands for commoditiesand services). Starting with thisequilibrium condition (the “base case”),an external change in economiccondition is introduced which causes ashift in resources, leading to theestablishment of a new equilibrium. Anexogenous shock (e.g., an oil priceincrease) tends to alter relative pricesand the allocation of goods amongconsumers and of resources amongproductive activities (in view ofinterconnected markets), putting theeconomy into disequilibrium at existingdemand and supply levels. Given such ashock, a CGE model will simulate theimpact and the succeeding interactionsof the various sectors of the economy,and it will estimate the new set of relativeprices necessary to reestablish theequilibrium solution (i.e., supply equalsdemand in each market). Comparisonsare made on the new values vis-à-vis theoriginal base case figures for the wholeeconomy and for each sector. In CGEmodeling, the objective is usually notto forecast the exact outcome of policymeasures but to give only an indicationfor the direction and size of the effects(Thissen 1998).

CGEs are almost all set up in “real”terms. A numeraire or a price index (e.g.,consumer price index, exchange rate,aggregate producer price) is chosen as aconstant to define a unit of account forall nominal values. There are no assetmarkets, money is neutral, and alleconomic agents make decisions on thebasis of relative prices rather than to the

absolute level of any price. Hence, themodel will then only solve for relativeprices (Sadoulet and de Janvry 1995).This implies that the economy does notsuffer from money illusion (i.e.,doubling prices will increase themonetary value but will not affect thereal value of transactions in theeconomy).7

2. Steps in Computable GeneralEquilibrium Modeling

Reed (1996) indicated that the basicidea behind a CGE is to build ananalytically consistent mathematicalmodel of an economy; collect data forsome time period on those variables forwhich data are available; and then usethe characteristics of the economy inthat time period to solve the modelnumerically. This last stage is performedcomputationally using computersoftware. The initial step in building aCGE model involves the formulationand collection of economic data into asocial accounting matrix (SAM) to forma consistent data set for a given baseyear. A SAM is a square matrix thatindicates the national accounts data inconsistent format and describes a flowof resources between various economicagents and factor markets andinstitutions. It is an extension of aninput-output (I-O) table and uses adouble entry bookkeeping system totrace monetary flows through debits andcredits. A SAM is organized in such away that receipts (rows) andexpenditures (columns) are in balance.



AP

PEN

DIX 3

It records the values of transactions inthe base year. In effect, it is a portrayalof the statistical state of the economyat that point in time (i.e., the base year).The basic structure of a SAM is providedin Figure A3.5.

After the establishment of a consistentdatabase, the next process concerns thegeneral design of the economicrelationships to be modeled. Thisinvolves choosing the number and typesof economic agents in the model. Then,a choice or optimization problem is

specified (through algebraic equations)for each agent. As an example, producersin each sector maximize profits, subjectto their available technology andproduction costs. They can also decidewhether to sell on the domestic marketor to export on the basis of relativeprices. It can also be assumed thathouseholds maximize utility, subject totheir budget constraints. Factors ofproduction (i.e., land, labor, and capital)are paid in accordance with theirrespective marginal productivity. CGEmodels also incorporate macroeconomic

Figure A3.5: A Typical Social Accounting Matrix

Industry Commodity Factors Institutions Government Trade Total

Industry Produce Totalindustryoutput

Commodity Utilize Consumption Exports TotalOutput commodity

Factors Returns to Exported Total –land primary primary factor –labor factors factors income –capital (value (e.g.,

added) labor flow)

Institutions Sales Sales Distribution Transfer Exports Total –households of factor payments institu- –others income tional

income

Government Indirect Sales tax Factor taxes Inter- TotalBusiness govern governmenttaxes mental income

transfers

Trade Imported Imports Imports Trans- Totalpurchased ship importsinputs ments

Total Total Total Total Total Total Totalindustry commodity factor institutional government exportsoutlay outlay outlay outlay outlay

Source: Vargas et al (undated).

CASE 5

128

APPENDIX 3

relationships, such as the balance ofpayments, savings-investment gap, andgovernment budget.

Once the economic agents are identifiedand their optimizing behaviors specified,the parameters in these equations mustbe solved or “calibrated” to reproduce(or exactly replicate) the benchmarkdata on the observed economictransactions in the database, as indicatedin the SAM. This involves using theSAM data along with a limited amountof additional information to determinethe values of all parameters in the model.For some of the equations, calibrationcan be done directly from the SAM(Buehrer and di Mauro 1993). A CGEallows for an arbitrary choice of anumeraire. A typical practice is tonormalize (i.e., data and the model aretested by running the base-year modelwith all prices set to one) prices andfactor rents (e.g., wages) in the initialequilibrium so that a certain price indexremains constant. For example, pricescan be normalized according to a priceindex based on consumption weights(Kehoe and Kehoe 1994). Sadoulet andde Janvry (1995) indicated thatmeasurement units for labor categoriescan be chosen so that all wages are initiallyequal to one. Similarly, measurementunits for quantities of domesticcommodities, imports, and exports arechosen so that the consumer prices ofdomestic goods and imports, the worldprice of exports, and the exchange rateare all equal to one in the base year. Withthe normalization rule, all initialquantities and prices can be computedand parameters that are directly computedfrom shares can be easily derived(Sadoulet and de Janvry 1995).

In CGE, “closure” of the model is doneto introduce macroeconomic constraintsthat limit the microeconomic behaviorof economic agents. This also involvesthe selection of which variables in themodel are to be exogenous (e.g., policyparameters such as tax rates, tariffs, andsubsidies). Additional endogenousvariables may be introduced to balancea particular constraint (Thissen 1998).Given the limited number of parametersthat can be determined by the calibrationtechnique, calibrating other parametersmay require data not indicated in theSAM. It is necessary to supplement thisprocess by assigning parameter values,such as price and income elasticities,either derived from secondary sources(i.e., research/econometric studies) orbased on the model builder’s judgment.

The equations in the CGE model maybe substituted and rearranged in orderto reduce the solution problem as muchas possible before solving it by anappropriate software package. The ideais to reduce the model to a set of excessdemand equations for product andfactor markets so as to generate the setof equilibrium prices and factor rents.For example, given some initial pricesand wages, excess demands in bothproduct and factor markets can becalculated by the computer. Then, thewages and prices are revised iterativelybased on the calculated excess demands.This iteration process will continue untilan equilibrium is reached (i.e., a set ofprices and wages is found so small thatall excess demand is sufficiently closeto zero). If both the model and the dataare consistent, the model will be able tosolve and generate the benchmark data.Inconsistency exists in either the



AP

PEN

DIX 3

specification of the model or the datawhen there is a discrepancy between theactual and generated data set (i.e., pricesdeviate from unity). In such a case, arevision will be necessary. Thisprocedure also implicitly provides anindication on whether the completecircular flows of income andexpenditures in the SAM are in balance.

Counterfactual experiments can beconducted by introducing changes in theexogenous variables or parameters, inpolicy reforms, and in marketconditions, and by rerunning the modelusing a software program. A comparativenumerical static analysis of the changein external condition, to examine thebase-case equilibrium situation vis-à-visthe new equilibrium after the exogenousshock has occurred or policy measureshave taken place, can then be undertaken.The results of static simulations areoften interpreted as representing howthe economic system in question wouldhave looked, had the new policy been inplace in the base year, after all relevantadjustments had been made (Gilbert andWahl 2000).

3. General Description of the NepalComputable General EquilibriumModel

The Nepal CGE model is adisaggregated multisector model with12 productive sectors and threehousehold groups. The database for theNepal CGE consists of an aggregateSAM and estimated elasticities. TheSAM consisted of 11 separate accountswith all productive activities andhouseholds aggregated. The first twoaccounts in the SAM are classified under

the heading of Production Accounts.Account 1 (activities) represents thetransactions of producers with entriesin the activities column representing theaggregation of the sector accounts. Onthe other hand, Account 2 (commodities)indicates the transactions of theconsumers. Combining these twoaccounts yields the standard nationalincome accounts identity for gross totaldomestic production equal to total useof resources. Accounts 3 and 4 (factors)describe the payment to factors ofproduction and distribution of factorincome to institutions, while accounts5, 6, and 7 (institutions) represent,together with enterprises, the majoreconomic actors in the economy.Government, households, and touristsare treated separately, reflecting theirbehavioral specifications. Account 8(capital) represents the balance betweentotal investment and total saving.Foreign saving is defined in accounts 9and 10 (India and Rest of the World)as the difference between foreignexchange inflows and outflows. Account11 (totals) sums up the totalexpenditure and receipts for eachaccount and indicates consistency in thedatabase. Table A3.17 sets out the SAMfor Nepal for fiscal year 1986/87.

The explicit parameters, such as themarginal propensity to save,consumption expenditure shares, andinput-output coefficients, are sharesderived from the SAM data. On theother hand, the values of exogenousparameters (e.g., export demandelasticities) are either assumed or basedon estimates from other econometricstudies. The endogenous parameters(i.e., price and income elasticities) are

CASE 5

130

APPENDIX 3

TABLE A3.17: Aggregate Social Accounting Matrix for Nepal, 1986/87(NRs million)

Expenditures

Rest ofAccount Acti- Commo- House- Tou- Govern- Capital the Total

vities dities Labor Capital holds rists ment Acct. India World Receipts

Activities 81,773 1,630 12 3,044 2,771 89,229Commodities 31,876 48,578 5,692 15,702 101,848Labor 21,182 21,182Capital 34,104Households 21,151 30,993 1,203 53,347Tourists 532 1,208 1,741Government 2,067 6,944 31 407 111 1,469 11,028Capital Acct. 2,705 4,769 5,324 12,798India 7,056 -3,480 3,576Rest of the World 6,076 575 6,651

TotalExpenditure 89,229 101,848 21,182 34,104 53,347 1,741 11,028 12,798 3,576 6,651 335,502

Source: Maxwell Stamp Consultants. 1991. Nepal Second Industrial Sector Study, Final Report, Annex (ADB TA 1047-NEP).

implicitly derived from the underlyingrelationships and the data set out in theSAM, and can be estimated throughsimulation.

The Nepal CGE is closed with anassumption about the level of saving andinvestment. Saving is fixed by specifyingthe level of capital inflows. Totalinvestment is also fixed but isdistributed between sectors by marketforces. With capital inflows specified,the trade deficit in real terms remainsfixed and varying only in relative priceterms.

4. Policy Simulation and Implications

The primary application of the NepalCGE modeling exercise is to identify theeconomy’s potentials in undergoingpolicy-induced changes in the industrialand trade regimes through variations intariffs, export subsidies, and the exchange

rate. A series of case studies wereundertaken to identify the mix of policyoptions. The “no policy change” scenariorepresented the base case against whichthe proposed policy packages were judged.These results are based on the technicalassistance (TA) consultants’ report onthe Nepal Second Industrial Sector Study(ADB TA 1047-NEP). Table A3.18 setsout these various policy experiments.

A decrease in the nominal protection ratewas conducted under C2 with theimposition of a uniform 30% tariff anda corresponding removal of excise taxesand import licensing premium. Tocompensate for the shortfall in revenueas a result of the tariff rate cuts, directtaxation was introduced, which was theleast distortionary. A 35% devaluation ofthe currency was carried out to improveexport competitiveness. In the C3 policyexperiment, a 20% subsidy was addedto the tariff, foreign exchange, and tax



AP

PEN

DIX 3

TABLE A3.18: Trade and Industrial Policy Case Studies, Nepal

Scenarios Characteristics

C1 Base case that assumes no policy change.C2 The immediate introduction in 1991 of a uniform 30% tariff rate on all imports; removal of import

licenses, excise, and export taxes; a revenue-compensating increase in direct taxation by 10%; anda 35% devaluation against the Indian rupee and other foreign currencies.

C3 Same as C2 plus a 20% subsidy on third country exports.C4 The phased 3-year introduction of policy measures specified in C3 with the same tariff and tax

rates achieved by 1993.C5 Phased introduction of 15% uniform tariff rates, introduction of a 10% export subsidy, harmoniza-

tion of 10% excise and sales taxes, removal of export taxes, and an increase in the rate of directtaxation.


changes, which were introduced in C2,to offset the 30% uniform tariff rate.The subsidy will be covered through anincrease in the direct taxation rate ofabout 15%. The C4 simulation test wascarried out to accommodate timing/phasing of structural reform proposals.The same policy reforms introduced inC3 were carried out but these werephased over a 3-year period. The C5policy experiment was a variant of C4.To illustrate the impact of policy reformon economic growth, results from C2,C3, and C4 simulations were comparedwith those of C1 to illustrate the impactof these policy changes; while resultsfrom C5 were compared with those ofC3 and C4. Table A3.19 reports theresults of the simulation runs (thoughsome macroeconomic indicators such asGDP, and export and import growthcould not be readily derived/inferredsince these were presented in graphicalformats).

Some policy implications can be derivedfrom these simulations on the reformproposals (pp. 241–242 of theConsultants’ Final Report). The

introduction of a uniform 30% tariff,together with the 20% export subsidy,is deemed to have the following effects:(i) A stable balance of payments

position can be maintained if thesereform measures are introducedwith a corresponding 35%devaluation. Stability in thegovernment budgetary position canalso be sustained if there is anattendant increase in revenue, eitherthrough direct taxation or a neutralsales tax, and without a concomitantrise (in real terms) in publicrecurrent expenditures.

(ii) Trade liberalization leads to acontraction in import-substitutionactivities. There is only a partialexpansion in exports owing to thelimited ability of producers toswitch markets in the short run.Third country export growth isprojected to increase partly at theexpense of exports to India.

(iii)Policy reforms result in strongermanufacturing sector activity.Export-led growth in themanufacturing sector results in ahigher rate of capital accumulation

CASE 5

132

APPENDIX 3

TABLE A3.19: Simulation Results, Nepal (%)

Scenarios

C1 C2 C3 C4 C5

Average Annual Growth rates(1990–2000)

Capital Stock 8.2 10.1 11.2 10.7 11.1Real Rural Wages 3.2 3.9 4.4 4.3 4.2Real Urban Wages 2.6 3.7 4.7 4.5 4.4

Terminal Year Shares (2000)Composition of GDP

Agriculture 45.1 42.3 42.3 42.7 41.8Manufacturing 9.3 10.9 12.6 12.6 12.0Services 45.7 46.7 45.1 44.6 46.2Investment Rate 12.6 14.5 15.4 15.5 15.4Economy-Wide Profit Rate 21.4 18.6 16.4 17.6 15.5Exports/GDP 15.4 18.2 20.3 19.8 20.2Imports/GDP 19.1 24.7 28.3 28.5 27.1


and a more robust growth in realwages across the economy. Giventhat the vast majority of householdsown labor rather than capital, it issurmised that the distribution ofincome will improve with tradeliberalization.

(iv) A delay in the implementation ofreform proposals causes efficiencylosses to the economy and translatesinto a drop in real GDP, ascompared with the scenario ofimmediate implementation. Withdelayed implementation, capitalstock accumulation and real wage

growth are slightly lower, ascompared with the figuresgenerated under the “immediateimplementation” scenario.

(v) Growth rates for the carpets,garments and textiles, othermanufacturing, construction, andcash crops subsectors aresignificantly higher withliberalization. However, the growthrates experienced by the food crop,agro-industry, tourism, and publicsectors are similar under theliberalization and “no change” casestudies.

APPENDIX

ADJUSTMENT COSTSIN ADB POLICYOPERATIONS—AN INVENTORYOF APPROACHESUSED

134

APPENDIX 4

Loan Main Policy Measures Adjustment Costs Analysis UsedIdentified for Cost Estimation

KAZ: Pension • Creation of an enabling, • Costs of transition to • Estimation of the fiscal costsReform regulatory, and institutional the new pension system of transition (funding of theProgram framework for a modern including payment of deficit) over the medium term($100 million) pension system arrears from existing (1997– 2002), discounted by

• Increase public system (to establish 12% to estimate the total andawareness of the a sound financial annual adjustment costspension reform basis for the new • Estimation of the fiscal costs

• Enhance administration system)—specifically of establishing regulatoryand management capacity the fiscal costs/funding agencies, establishment of aof the pension system of deficits in the residual system for pension

• Ensure financial of the old system for contributions information,sustainability existing and future payment and clearing

• Ensure social retirees functionssustainability

Supporting studies:• ESW and policy options

paper (Health SectorDevelopment ADTA)

• Health Sector ResourcesDevelopment TA

• Strengthening HealthInsurance ADTA

• World Bank Study Mongolia Poverty Assessment in aTransition Economy

MON: Health • Promote family group • Costs of transition • The cost of FGP is estimatedSector practice (FGP)-based to the FGP-based at about $5 per capita perDevelopment primary health care system including yearProgram (PHC) as opposed to the capitation(Policy $4 hospital-based payment for the Supporting studies:million, curative care. family doctors • ESW and policy optionsInvestment Loan • Encourage private and salaries for paper (Health Sector$11.9 million) sector participation nurses, rent, Development ADTA)

• Rationalize health heating, electricity, • Health Sector Resourcesfacilities and personnel water, essential drugs, Development TA

• Improve finance and basic laboratory • Strengthening Healthmanagement equipment, and supplies Insurance ADTA

• Protect poor and • Rationalization of • World Bank Study Mongoliavulnerable groups hospital facilities in Poverty Assessment in a

Ulaanbaatar Transition Economy• Relocation/training

costs for nonmedicalpersonnel

• Primary health careservices upgrading

• Reform of hospitalservices

• Development of ruralhealth services andreferral system

• Managementstrengthening

• Institutionalstrengthening

continued...


ADJUSTMENT COSTS IN ADB POLICY OPERATIONS—AN INVENTORY OF APPROACHES USED

AP

PEN

DIX 4

UZB: Education • Modernize structure, See Case 3 in • Aggregations are based on aSector contents, and Appendix 3. few basic data such asDevelopment processes of average salary per staffProgram education services category, estimated training($38.5 million) • Improve sector and relocation costs, level of

sustainability and stipends, operating costs perefficiency student. Details are

• Reform sector presented in Supplementarygovernance Appendix: Budget Impact of

• Protect poor and the Principal Policyvulnerable groups Measures.

SRI: Private • Privatization, divestment • Public Enterprise • PPTA/Mission financialSector or restructuring of Reform Commission analysis of public enterprisesDevelopment identified public restructuring-related • PPTA/Mission financial(Cluster) enterprises engaged operating costs analysis of proposedProgram in commercial activity ($5 million) institutions(Subprogram I • Commercialization of • Cost of enterprise • PPTA/Mission demand$100 million public enterprises under restructuring assessment for skills trainingand Subprogram II government control ($271 million) and financial cost estimation$50 million) through a corporate • Costs of retiring • Institutional and financial

governance framework workers for public analysis of bank restructuring• Private sector enterprises for • Financial analysis of deposit

participation divestment insurance schemein infrastructure ($395 million) introduction

• Improvement of • CPA establishment • Financial analysis of shiftcompetition and operating costs from direct borrowing topolicy and consumer ($1.4 million) market-based instrumentsprotection through • Insurance BoardConsumer Protection and Private Supporting studies:Authority (CPA) legislation Superannuation • ESW updates carried out byand establishment of Benefits Funds the mission and supported byan operational authority Regulatory findings of a USAID study on

• Remove labor-market Commission private sectorrigidities through establishment and competitivenessstrengthening of operating costsprocesses of dispute ($3 million)resolution and • Support for Socialarbitration Safety Net

• Improved private sector Fund (Skillsaccess to finance Developmentthrough further fiscal Fund) ($17 million)consolidation • Costs of formulatingand reduction of the restructuring planscrowding out effects of for state banksgovernment borrowing ($6 million)in the market, improved • Costs of establishingefficiency of financial deposit insurancemediation, and capital scheme (zero)market development • Cost of government

shift from directborrowing to market-based instruments($202 million)


Appendix 4 (cont’d.)

continued...

136

APPENDIX 4

IND: Gujarat • Strengthening of state • Revenue implications • A Medium-Term FiscalPublic Sector finances and their of tax restructuring Policy FrameworkResource management, including revenue was developed by theManagement including revenue losses associated mission in collaborationProgram augmentation with tax reforms with Gujarat Finance($250 million) • Expenditure, and loss offsetting Department to analyze

prioritization, through improved and project the fiscalmonitoring and tax compliance situation (e.g., revenuecontrol in the resulting from receipts, tax transfers,Core Investment reduction and current and capital outlays,Program, raising cost rationalization of loans and advances, etc.)recovery, subsidy tax rates, widening • Public finance analysisreduction, privatizing of tax base, and (including detailed taxselected services strengthening of analysis) through TA support

• Development of fiscal tax administration • Detailed case analyses ofpolicy and management • Cost offsetting SOEs through supporting TAcapabilities and factors and overall • Institutional analysis ofstrengthening of tax budget implications Gujarat Infrastructureadministration of restructuring, Development Board

• Privatization or including:divestment and • SOE debt settlement Supporting studies:restructuring of ($283 million) • Support for Gujarat’s Public 23 of 54 SOEs, and • cost of SOE labor Finance ADTAprovision of a social rationalization • Restructuring of State-Ownedsafety net for ($129 million) Enterprises of Gujarat ADTAretrenched workers • divestment proceeds

• Development of the • improved costenabling environment recovery (e.g.,for private sector arising from powerinvolvement in tariff increase)infrastructure ($198 million)projects (power, • SOE subsidy savingsports, roads sectors) ($71 million)

• Institutional • Revenue lossesstrengthening of associated with taxthe State Finance reforms ($302 million)Department • Revenue gains from

enhanced taxcompliance and tax-base broadening

• Project developmentand feasibility studiesfor power, ports, androads ($25 million)

VIE: State- • Promotion of an enabling • SOE labor retrenchment • Mission financial estimates ofOwned environment for foreign costs ($112 million) retrenchment pay andEnterprise direct investment • Costs of financial pension contributionReform and • Supporting development restructuring for medium • Based on IMF estimates andCorporate of private enterprises and large industrial mission financial estimatesGovernance by allowing non-state SOEs, including debt • Financial estimates for costsProgram enterprises to use their write-offs ($126 million) of institutional development($100 million) land-use rights as equity • Establishment and

contribution in joint operating costs (No underpinning ESW)ventures, unifying tax associated with



continued...



AP

PEN

DIX 4

treatment, strengthening institutionalbankruptcy procedures, development,introduction of summary including one-stopjudgment procedures shops for foreignfor noncompliance investment approvals/with contractual obligations licensing, and securities

• Industrial SOE restructuring trading centers,by improving the accounting and auditinginstitutional and standards office, costs ofregulatory framework for adoption of internationalSOE reform, accelerating accounting standardsindustrial SOE reform, by SOEs ($50 million)improving corporategovernance and financialdiscipline

• Enhanced labor mobilitythrough greater flexibilityin the social securitysystem

PAK: Trade, • Augmentation of trade • Estimation of the • Analysis of publicExport liberalization process, revenue losses from manufacturing enterprisesPromotion including reduction, tariff rationalization and trade bodies includingand Industry rationalization and and export costs of debt settlement andProgram simplification of the duties ($310 million) retrenchment (undertaken($300 million) tariff structure, • Loss of revenue from by mission)

rationalization and removal of export duties • Institutional restructuringreduction of the • Contribution to the analysis (by mission)Statutory Rules and Preshipment ExportOrders, phasing out of Finance Guarantee Fund, Supporting studies:nontariff barriers for National Accreditation ESW including:imports and exports Council, export • In-depth update of sector

• Modernization of the free marketing and profile (Industry and Tradetrade status policies and product upgrading Sector Study)administration for fund, Foreign Currency • Development of a medium-exporters Import Finance term export strategy (Export

• Broadening the coverage ($200 million) Development Strategy)and access to financing • Settlement of debt • Analysis of the effects offor exporters liabilities and severance changing tariff regimes on

• Restructuring of the Export payments for privatized industrial protectionPromotion Bureau, Public Manufacturing (Analysis of Tariff Reform indevelopment of private Enterprises (in part to be Pakistan) and Missionsector trading companies, offset from sale updatesabolition of state trading proceeds) ($200 million)bodies, capacity • Costs of redundancydevelopment to deal with payments for workers frominternational trade restructured or liquidatedenvironment trade bodies ($200 million)

• Development of investment • Cost of restructuring thepolicy and its effective Export Promotion Bureauimplementation ($10 million)

• Acceleration ofrestructuring andprivatization of publicmanufacturing enterprises



continued...

138

APPENDIX 4

INO: Financial • Adoption of best practices • Basis of cost estimates • The program aimed toGovernance in banking, capital markets, not provided support efforts to improveReforms Sector and public financial resource allocation efficiencyDevelopment management within the sector byProgram • Increasing disclosure and improving governance of($1.4 billion) transparency in banking, financial and public sector

capital markets, and resource allocation. Whilepublic financial measures were described formanagement restructuring and

• Strengthening legal liquidations, improvedframework in banking market orientation, bankand financial management supervision, increased

disclosure, anti-corruptionetc., the RRP contains noquantitative details.

THA: Social • Labor market and social • Support for laid-off • The loan amount was basedSector Program welfare program to address workers and various on estimated financial costs,($500 million) (crisis-related) initiatives to address in terms of budget

unemployment, poverty growing unemployment implications, forprogram effectiveness, ($200 million) implementing both crisisprivate sector training, • Support for students (e.g., rising unemployment)and labor force to prevent dropout and institutional reforms.competitiveness ($550 million)

• Education sector reforms • Expansion of health Supporting study:to address dropout rates coverage for low- • Social Impact Analysis of the(crisis-related), education income families Economic Crisis ADTAquality, staff rationalization, ($150 million)decentralization, andprivate sector education

• Health sector reforms toensure access by the poorand rural health caredelivery improvement

KAZ: • Creation of an agricultural • Issuance of land share • The total financial adjust-Agriculture land market certificates to individual ment cost was estimatedSector • Transfer of ownership of farm workers ($90 million) to be $400 million in presentProgram state farms to private • Revenue losses as a value terms in 1996($100 million) entities result of removal of (a medium-term horizon of

• Liberalization of external export tax ($40 million/ 5 years and a discount ratetrade year over the medium of 20% were used in

• Dismantling of state term) estimating the discountedmonopoly companies and • Shift to a competitive present value).restructuring of Agroprom public grain procure-bank ment system will result Supporting study:

• Promotion of a market- in a net additional • Strengthening thebased rural credit and expenditure due to Implementation of Agriculturesavings system higher grain prices Sector Reforms ADTA

• Strengthening social ($40 million/year overprotection for adversely the medium term)affected vulnerable groups • Fiscal measures to

• Strengthening strengthen socialenvironmental protection ($8 million/management year over the medium

term)



continued...



AP

PEN

DIX 4

COO: • Public sector reform • Retirement of a part of • No detailed aggregationEconomic • Promote private sector government’s short-term documented. The loan sizeRestructuring growth liabilities to domestic was on the basis of “theProgram • Social equity and private sector creditors significance of the policy($5 million) sustainability • Establishment of changes and economic

Business Ventures restructuring being pursuedDevelopment and the historical levels ofCorporation under Bank lending to the CookCook Islands Islands.” (RRP, para. 75)Development Bank(CIDB) Supporting study:

• Equity injection in CIDB • 1995 Economic Report served• Given the outstanding as ESW

government arrears andother external debttotaling NZ$200 million,the loan size ofNZ$7.2 million wouldnot have bridged thegap (Knapman andSaldanha, 1999, p. 34).

RMI: Public • Stabilize • Retirement-in-force • Based on salaries paid in theSector Reform government (RIF) program civil service and averageProgram finances in ($5.5 million) length of service, and($12 million) the short run. • Pay off Air Marshall expected number of

• Ensure long- Islands’ (AMI) commer- retrenchments (RRP,term structural cial debt ($4 million) para. 69)stability of • Seed injection to • Two thirds of AMI’sgovernment Financial Reserves outstanding commercialfinances. Trust Fund debt.

• Create an ($2.5 million)improved (Knapman and Supporting study:environment Saldanha 1999, • 1996 Economic Report servedfor the private pp. 89–90) as ESWsector

FSM: Public • Reduce size and • Direct costs of early • No detailed documentation inSector Reform operating costs of retirement schemes the RRP but the EMPAT teamProgram civil service for the national and (ADTA consultants) provided($18 million) • Increase domestic 4 state governments detailed estimates and also

revenue generation ($3 million for the developed monitoring• Restructure government national, $5.3 million schemes for the retirement

operations and public for Chuuk, $4.2 million programenterprises for Pohnpei,

• Mitigate social and $3.5 million for Supporting study:economic impact Yap, $2 million for • 1996 Economic Report served

• Foster development of Kosrae) as ESW.private sector

SAM: Financial • Deepen financial markets • Incremental budgetary • A loan of $7.5 million isSector Program • Strengthen Central Bank of cost of issuing CBS aimed at covering a major($7.5 million) Samoa (CBS) and enhance bills ($1.9 million share of these costs to the

its operational autonomy over a 3-year period) government, with the• Strengthen prudential and • Temporary remaining part coming from

regulatory framework compensation for the other sources, including



continued...

140

APPENDIX 4

• Strengthen National loss of revenue from revenue from privatization ofProvident Fund (NPF) elimination of the state-owned enterprisesand Development Bank foreign exchange (SOEs). (RRP, paras. 101–104)of Samoa (DBS) levies ($1.3 million • Short- to medium-term costs

• Privatize and corporatize annually) of the reform program werepublic enterprises and • Temporary estimated in fiscal andutilities compensation for balance-of-payment terms

the loss of revenue • Incremental cost of issuingfrom the planned CBS bills was calculated ascorporatization of the total incremental interestthe Posts and rate costs associated withTelecommunications interest rate differentialDepartment($5.72 million annually)

VAN: • Redefine the role of • Restructure and right- • Fiscal outlook covering 1997–Comprehensive government, and size the public sector 2000 and GovernmentReform Program enhance the quality • Restructure and funding plan 1998–2000($20 million) and delivery of its rehabilitate formed the basis of external

policy, regulatory, government-owned financing need for the reformand development financial institutions program (RRP, paras. 73–83)services • Fiscal stabilization

• Increase the (reduce the need for Supporting study:productivity and an inflationary • 1997 Economic Report servedgrowth of the domestic financing as ESW.private sector in of budget deficit)both urban and • The loan amountrural areas ($20 million) was to

• Support social cover 35%, 40%, anddevelopment of 25% of the abovethe disadvantaged three adjustmentand rural population costs. (Knapman and

Saldanha 1999, p. 152)

SOL: Public • Macroeconomic • No specification • The loan amount is based onSector Reform stabilization and fiscal documented in the the “scope and the costs ofProgram reforms RRP. Ex-post, the loan the policy reforms, the($25 million) • Public sector amount ($25 million) importance and urgency of

management reforms permitted clearance of the reforms and the state of• Public enterprise reforms approximately 63% of the public finances of the• Strengthen governance government arrears Borrower.” (RRP, para. 69)

institutions ($23.8 million);the financing of public Supporting study:service retrenchments • 1997 Economic Report served($1 million); and as ESW.incremental costs ofhiring staff toimplement reforms($0.2 million).(Knapman andSaldanha 1999. p.131)



continued...



AP

PEN

DIX 4

NAU: Fiscal • Fiscal management • Separation payments Supporting study:and Financial reform (450 public service • National Reform ProgramReform Program • Improve asset employees and 250 ADTA($5 million) management and Nauru Phosphate

enhance returns to corporation staff,the economy totaling $2.7 million)

• Establish a viable • Direct budget supportbanking system to pay arrears, deferred

• Improve efficiency obligations, andand effectiveness restructured debtof the public sector payments ($1 million)

• Finance specialistinputs to improve publicsector and SOEperformance($0.8 million)

• Payment of nonwage-related health andeducation expendituresto improve servicedelivery ($0.5 million)

TUV: Island • Decentralize and enhance • Budgetary provisions • The size of the ADB loanDevelopment regional autonomy to cover annual capital ($4 million) is based on theProgram • Foster an enabling assistance grants to the assumption that the($4 million) environment for regional islands at about the government and the island

development same level as expected communities will match theirfrom the trust fund when contribution by $4 million andit becomes operational, $0.5 million, respectively, soplus the cost of that the initial capital canadministrative generate reasonable returns.reorganization and (RRP, para. 103-104)upgrading on the islands

• Reorganization of thefinancial institutions(Development Bank ofTuvalu merging with theNational Bank of Tuvalu).

IND: Madhya • Enhance resource • Revenue losses from • Estimated financialPradesh Public allocation to social tax measures implications of the reformResource sectors through ($185 million) Program: $601.2 millionManagement focused interventions • Cost of public • Estimations done by MissionProgram • Implement public enterprise reform based on information($250 million) sector reforms including non- provided by the Government of

including capacity personnel outlays Madhya Pradesh andbuilding and public sector individual PSUsinstitutional undertaking (PSU) • Excluded outlays on accountstrengthening reform ($296 million) of voluntary retirement

• Promote an enabling • Cost of voluntary schemeenvironment for private retirement schemesector participation and social safety net

($120.4 million)



continued...

142

APPENDIX 4

PAK: Energy • Enhancement of • Restructuring cost of • Broad estimate of adjustmentSector governance in the the energy sector, cost is about $1.71 billionRestructuring energy sector including the cost to • Restructuring cost estimatedProgram • Enhancement of the help address the circular by the Government of($350 million) legal and regulatory debt problem in the Pakistan

framework and power sector and thestrengthening of the adjustment cost ofcapacity of the sector the natural gas andregulator petroleum subsectors

• Financial restructuring ($1.6 billion)and privatization of • Operating costs andutilities and corporatized other related costs ofentities reform (e.g., labor

• Creation of an enabling restructuring cost,environment for a cost of restructuringcompetitive electricity plans) ($105 million)market

• Enhancement of reformin the natural gas andpetroleum subsectors

INO: Industrial • Strengthening competition • Lost revenues from • Total adjustment cost relatedCompetitiveness by establishing an tariff reform to the Program wasand Small appropriate legal ($150 million) estimated at aboutand Medium environment for • Cost of administrative $650 million during theEnterprise (SME) commercial activity, reforms linked to the Program periodDevelopment simplifying the regulatory computerization and • The elimination of protectionProgram framework, removing reorganization of in a number of sectors is($200 million) barriers to domestic customs ($25 million) expected to impose costs to

competition and trade, • Establishment of enterprises in terms ofand liberalizing Commission for restructuring and retoolinginternational trade Competition needs. These costs are

• Facilitating investment Supervision and expected to be considerableand trade by other administrative but are difficult to quantify.strengthening the policy reforms ($5 million)framework governing • Lost revenues Supporting study:domestic investment and resulting from the • Trade and Industry Planningforeign direct investment, simplification of the and Strategy Formulation forand improving duty licensing and approval Repelita VII ADTAdrawback and customs regime ($470 million)procedures • Elimination of protection

• Rationalizing assistance in a number of subsectors,to SMEs by improving including fertilizerpolicy coordination and distribution, petro-implementation, improving chemicals, and otherfinancial intermediation, domestic trade subsectorsand strengthening is expected to imposetechnical and business upfront costs toservice support enterprises in terms of

restructuring andretooling needs(These costs areexpected to beconsiderable but aredeemed difficult toquantify)



continued...



AP

PEN

DIX 4

VIE: Agriculture • Stimulate competition in • Cost of establishing • Partial estimate ofSector Program the domestic and export a modern cadastral adjustment costs is about($80 million) marketing of rice and in service and issuance $190 million

the importation of fertilizer of certificates of landto reduce marketing transfer for parcels Supporting studies:margins, improve of land ($90 million • The Program evolved fromproducers’ prices, and during the program discussions with thereduce the cost of fertilizer period or a total of government over a period

• Support equal participation $360 million over a of 2 years.by the private sector and 12-year period)trade liberalization • Cost of upgrading/

• Establish the legal replacing facilities forframework for voluntary the storage andcooperatives, and farmers’ processing ofand traders’ associations agricultural outputs

• Improve the efficiency and • Cost of restructuringprofitability of formal rural and modernizing thefinancial intermediation agricultural extension

• Accelerate land tenure research ($100 million)reforms, including theallocation of land-userights to rural households

VIE: Financial • Establish the essential • Adoption of new • Adjustment costs estimatedSector Program market infrastructure to accounting and audit to be in the range of $265($90 million) facilitate private sector standards that are in million

participation in relation close conformity withto the legal and regulatory international standards Supporting studies:framework, the accounting • Setting up a deposit • Development of Small-Scaleand audit system for banks, insurance fund to Rural Credit TAinformation disclosure, maintain and build the • Commercial Banks Reviewand investor protection corpus of the fund to and Training TA

• Commercialize and eventually match • Financial Marketsmodernize domestic actuarial estimates Development TAbanking operations by of possible claims or • Financial Sector Review TAproviding greater losses from insurableautonomy to the state- risks ($10 million)owned commercial banks, • Organize a securitiesimprove risk management agency, a stock exchangepractices, and raise and an automatedcapacity for deposit central depositorymobilization system ($20 million)

• Promote competition by • Other programleveling the playing field components that willand diversifying the range cover collateralof financial institutions registration system,

• Initiate the development credit informationof a capital market by system, centralizedpromulgating basic system, and translationsecurities legislation facilities for variousand establishing a legislation, upgradesecurities agency to of accounting andregulate and develop the audit systems,market for securities including

computerizationrequirements,



continued...

144

APPENDIX 4

training costs, andengagement ofconsultants($10 million)

INO: Power • Restructuring of the power • Obligations for power • Adjustment costs were basedSector sector and creating an purchase on reasonable macro-Restructuring enabling environment for • Budgetary support to economic assumptions andProgram a competitive electricity compensate for the were expected to be between($380 million) industry market phased tariff increases $1 billion–$1.5 billion

• Establishing competition • Cost of debt annually in 1999 and 2000,in the supply of bulk restructuring and and about $4 billion in 2001electricity in Java-Bali expenditure related • These costs were expected to

• Adjusting tariffs to ensure to staff reduction be met by direct budgetaryfinancial viability of the associated with outlays, loans fromstate-owned utility and organizational multilateral and bilateralthe newly created restructuring sources, and part of thesubsidiaries during the proceeds from the sale oftransition period the assets of the state power

• Increasing private sector utility in Java-Baliparticipation

• Strengthening theregulatory environment,including protecting theinterest of the end-consumers

IND: Gujarat • Establish independent • Rationalization of • The total cost of adjustmentPower Sector tariff setting and regulation electricity duty is to be incurred by the stateDevelopment • Rationalize the imposition expected to cause a government over the reformProgram of tariffs and duties in the revenue loss of about period is estimated at $235($150 million) sector to maintain equity $33 million each year million

among consumer in the next 5 years for • It is felt that the program loancategories a total of $165 million of $150 million is adequate

• Introduce competition and • Long-term loans to and justified by the statecommercialization Gujarat Electricity government on account of

• Improve conservation of Board (GEB) to retire the reforms and the politicalwater and electricity expensive commercial costs it will incur during

debt and adjustment of implementationindependent powerproducer prices will be Supporting studies:about $43 million • Preparation of a power

• Payment of arrears of system master plandues of municipalities Preparation of a frameworkand other local bodies for electricity tariffsuntil FY2000 was • Review of electricityestimated to cost legislation and regulationsabout $15 million. It was • Financial support to GEB forexpected that in FY2001, the formation of twoan additional $5 million independent distributionmay be required profit centers

• About 700,000 consumer • Solicitation for private sectormeters will be procured implementation of theand installed over the Chhara Projectnext 3 years which willcost around $7 million



continued...



AP

PEN

DIX 4

MON: Agriculture • Reduce price and other • Preparation and • Adjustment costs areSector distortions to improve implementation of estimated at $8.0 million overDevelopment resource allocation and the plan to achieve the next 4 yearsProgram efficiency operational and • Increased costs of the($7 million) • Promote competitive financial autonomy structural reforms will be

markets for agricultural for Biokombinat partly offset by reducedinputs, outputs, and ($0.4 million) costs, or increased revenues,processed goods • Implementation of induced by the reforms from:

• Ensure delivery of the Cooperative Law (i) the more effective use offinancial services and and Cooperative financial resources resultingprovide improved access Development Program from the sale of commodityto credit for the rural ($0.3 million annually) aid; (ii) use of competitivepopulation • Implementation of an bidding procedures for

• Rationalize tax incentives action plan to promote procurement and sales byto promote investment investment in government agencies; andin rural areas disadvantaged rural (iii) proceeds from the

• Improve productivity and areas ($0.4 million privatization of the remainingsustainability in extensive annually) crop farmslivestock production • Control of animal

• Strengthen agricultural diseases that are public Supporting study:research and extension to health hazards • Agriculture Sectorsupport private sector ($0.3 million annually) Development Program ADTAagriculture • Establishment and

• Mitigate risks in operation of a licensingagriculture and ensure system for veterinariansfood security, income, supervising meatand employment for hygiene and productionvulnerable groups ($0.1 million annually)

• Operations of thesecretariat to maintainthe register forveterinarians andparaveterinarians($5,000 annually)

• Implementation of pilotprograms for improvedpasture management($0.4 million annually)

• Operation of the Scienceand Technology Council($0.4 million annually)

• Implementation of amedium-term plan forextension ($0.4 millionannually)

• Implementation of thepolicy on riskmanagement is alsodeemed to involvesignificant costsalthough these werenot indicated in theRRP



continued...

146

APPENDIX 4

MON: Financial • Strengthen the legal and • There will be costs to Supporting studies:Sector Program regulatory framework enterprises, which will • Institutional Strengthening of($35 million) for banking operations by lose up to 10% of their the Financial Sector TA

improving supervision deposits. • Institutional Strengthening ofand the regulation of Agricultural Banking Servicesnonbank financial • Strengthening of theinstitutions; improving Commercial Banking Systemthe financial information • Strengthening of Financialon banks, improving the Intermediarieslegal basis for debt • Development of Proceduresrecovery; and facilitating for the Reconstruction andthe enforcement of financial Liquidation of Insolvent Bankssector-related laws • Development of Bank

• Strengthen financial Restructuring Strategiesintermediaries byfacilitating the operationaland financial restructuringof banks

• Improve the efficiency ofthe financialintermediation processby reducing governmentinvolvement in bankingoperations, and byestablishing market-determined interest rates

PHI: Grains • Liberalize and promote • Government revenue Supporting study:Sector more cost-effective losses from the • Study on Food Crop PoliciesDevelopment grains pricing and import elimination of tariffs ADTAProgram policies on agricultural and($100 million) • Improve administration agriculture-related

of grain buffer stocks inputs, have been• Restructure National officially estimated

Food Authority from a at about $25 milliongrains marketing annually.monopoly into a public • Revenue lossesregulatory agency and from the loweringseparate private sector of tariffs on grains,marketing corporations primarily corn and

• Promote a more targeted corn substitutes,and effective food subsidy are estimated atprogram for the poor $10 million–

• Improve the coordination, $20 million annually,organizational structure, depending on theresearch and development, level of importsgrains and seed production, • Targeted food subsidyand marketing and/or food-for-work

program will incur aminimum annual costof $50 million startingin the third year, but maydouble or triple in cost,depending on thepercentage of poorfamilies targeted.



continued...



AP

PEN

DIX 4

• Compensationpackages for affectedstaff range from$20 million–$60 million.

• Increased annualbudgetary allocationsfor agriculturalinfrastructure, research,extension, and otherservices supported bythe Grains SectorDevelopment Programwill require$125 million–$250 million annuallyfor the next 4 years.

CAM: Financial • Enhancing banking • Adoption and • The costs of structuralSector Program intermediation and public enforcement of reforms are estimated atLoan ($30 million - confidence by establishing international $80 million–$90 million.program loan and strengthening banking accounting and auditcluster of about supervision, developing standards Supporting study:$10 million key information ($37 million) • Preparing for the Financialper program infrastructure and safety • Establishment of Sector Development Programloan) net, and building capacity banking sector ADTA

in both public and private infrastructure andsectors by reinforcing safety net includingcapacity building public registry ofinstitutions secured transactions,

• Establishing legal and credit informationregulatory framework for exchange arrangement,insurance development by and deposit insuranceinstitutionalizing a system ($18 million)supervisory system and • Strengthening bankprudential regulations; prudential regulationsfostering private sector and bank restructuringdevelopment in the ($16 million)insurance business; and • Development of inter-developing compulsory bank and moneyinsurance markets and laying the

• Laying the foundation for foundation for thethe development of inter- capital marketbank and money markets developmentthrough public-private ($7 million)sector partnership and • Development ofestablishing a basic legal legal and regulatoryframework framework for the

• Developing financial financial sectormarket infrastructure by ($6 million)establishing commonaccounting standardsand enforcementmechanisms andpromulgating relevant laws



BIBLIOGRAPHYAND REFERENCES

150

BIBLIOGRAPHY AND REFERENCES

Abonyi, G. 2002. Toward a Political EconomyApproach to Policy-Based Lending. Economicsand Research Department (ERD)Working Paper Series No. 14. ADB, Manila.

Agenor, P-R. and P. Montiel. 1999.Development Macroeconomics. second edition.Princeton University Press.

AGILE Consultants. 2000. StrategicReorganization of the National FoodAuthority (NFA) for the NewMillennium. Draft final report submittedto USAID. January.

Ahuja, V., B. Bidani, F. Ferreira, and M.Walton. 1997. Everyone’s Miracle: RevisitingPoverty and Inequality in East Asia .Washington, DC: World Bank.

Ali, I. 1990. Interrelationship between ShadowPrices, Project Investment and Reforms: AnAnalytical Framework. Economics andDevelopment Resource Center (EDRC)Report Series No. 52. ADB, Manila.

Aoki, M. 2000. What are Institutions andHow Should We Approach Them? Paperpresented at the Deutsche Stiftung furinternationale Entwicklung (DSE)workshop series on The InstitutionalFoundations of a Market Economy.

Asian Development Bank. 1986. Handbookon Management of ProjectImplementation. ADB, Manila.

———. 1987. A Review of ProgramLending Policies. Doc. R117-87.November. Manila.

———. 1995. Staff Instruction on PovertyImpact Assessment. Strategy and PolicyOffice. Manila. March.

———. 1996. Review of the Bank’sProgram Lending Policies. Doc. R143-96. Strategy and Policy Office. Manila.

———. 1997a. Guidelines onOperational Procedures for ProgramLending. Manila.

———. 1997b. Guidelines for the EconomicAnalysis of Projects. Economics andDevelopment Resource Center. Manila.

———. 1999. Review of ADB’s ProgramLending Policies. Doc. R210-00. Manila.November.

———. 2000. A Pacific Strategy for theNew Millennium. Manila. September.

———. 2001. Special Evaluation Study onProgram Lending. Operations EvaluationDepartment. Manila. November.

———. Forthcoming. Economic Analysisof Rural Development and AgricultureProjects: Concepts to Rationale. ERDWorking Paper Series.

Aron, J. 2000. Growth and Institutions: AReview of the Evidence. World BankResearch Observer, Vol. 15, No 1.

Artis, M. 1984. Macroeconomics. Oxford:Clarendon Press.

Bardhan, P. (undated.) The New InstitutionalEconomics and Development Theory: A BriefCritical Assessment. Available:www.lib.pku.edu.cn/webcourse/nie/readings/bardhan.doc.

Behrman, J. 1993. Macroeconomic Policiesand Rural Poverty. In Rural Poverty in Asia:Priority Issues and Policy Options. Edited by



RE

FE

RE

NC

ES

M.G. Quibria. Oxford University Press,Hong Kong for ADB.

Bolt, R. and M. Fujimura. 2002. Policy-BasedLending and Poverty Reduction: An Overview ofProcesses, Assessment, and Options. ERDWorking Paper Series No. 2. ADB, Manila.

Branson, W. and Hanna, N. 2000. Ownershipand Conditionality. Working Paper SeriesNo. 8. Operations EvaluationDepartment. World Bank, Washington, DC.

Brinkerhoff, D.W., and B.L. Crosby. 2002.Managing Policy Reform. Concepts and Toolsfor Decision-Makers in Developing andTransitioning Countries. Kumarian Press:Bloomfield.

Brinkerstaff, D.W. 1996. ProcessPerspectives in Policy Change:Highlighting Implementation. WorldDevelopment, Vol. 24, No. 9, September,pp. 1395–1401.

Buehrer, T. and F. di Mauro. 1993. AComputable General Equilibrium Model ofNepal. Report No. 60. ADB, Manila.

Cashin, P., Manro, P. and Sahay, R. 2001.Macroeconomic Policies and PovertyReduction: Some Cross-CountryEvidence. Finance and Development, Vol. 38,No. 2, June, pp. 46–49.

Coase, R. 1937. The Nature of the Firm.In The Firm, the Market, and the Law. Editedby Ronald H. Coase. Chicago: Universityof Chicago Press, pp. 187–213.

Collier, P. 1999. Consensus Building,Knowledge, and Conditionality. Paperpresented at the International

Symposium on Global Finance andDevelopment, Tokyo, March.

———. 2001. Consensus Building,Knowledge, and Conditionality. Paperpresented at the World Bank AnnualConference on Development Economics,2000.

Collier, P., P. Guillaumont, S. Guillaumont,and J.W. Gunning. 1997. RedesigningConditionality. World Development,Vol. 25, No. 7, September, pp. 1399–1407.

Collier, P., and J.W. Gunning. 1999. TheIMF’s Role in Structural Adjustment.Economic Journal, Vol. 109. November.

Corbo, V. and P. Rojas. 1992. World BankSupported Adjustment Programs:Country Performance and Effectiveness.In Adjustment Lending Revisited. Edited byV. Corbo, S. Fischer, and S. Webb.Washington DC: World Bank.

DAI Consultants. 1998. Philippines GrainsSector Development Program Project. FinalReport.

de Janvry, A., A. Fargeix, and E. Sadoulet.1990. The Political Economy ofStabilization Programs: Growth, Welfare,and Sustainability. Department ofEconomics, University of California,Berkeley.

de Maio, L., F. Stewart, and R. van derHoeven. 1999. Computable GeneralEquilibrium Models, Adjustment and thePoor in Africa. World Development, Vol. 27,No. 3.

152


Deloitte and Touche Consulting Group PteLtd. 1997. Report of Electricity TariffStudy on Long Run Marginal Cost. 2vols. Singapore.

Demery, L., and D. Demery. 1991. Povertyand Macroeconomic Policy in Malaysia.World Development, Vol. 19, No. 11.

Demery, L. et al., eds. 1993. Understandingthe Social Effects of Policy Reform. WashingtonDC: World Bank.

Dervis, K., J. de Melo, and S. Robinson.1982. General Equilibrium Models forDevelopment Policy. Washington DC:World Bank.

Dollar, D. 1992. Outward-OrientedDeveloping Countries: Do They ReallyGrow More Rapidly? Economic Developmentand Cultural Change, Vol. 40.

Dollar, D. and J. Svensson. 1998. WhatExplains the Success or Failure of StructuralAdjustment Programs? Washington DC:World Bank.

Dornbusch, R., S. Fischer, and R. Startz,1999. Macroeconomics (internationaledition). Singapore: McGraw-Hill.

Edmonds, C., and Yao X. 2001. Report ofthe Working Group on Poverty ImpactAssessmentfor Projects. Draft. ADB, Manila.

Elson, D. and Cajatay, N. 2000. The SocialContext of Macroeconomic Policies. InWorld Development, Vol. 28, No. 7 ,pp. 1347–1364.

Evans, D. 1999. Program Lending:Measuring Development Impact.

Preliminary report to the Economics andDevelopment Resource Center. ADB,Manila.

Ferroni, M., and Grootaert, C. 1993. TheSocial Dimensions of Policy Reform:Concepts, Data, Analysis. In Understandingthe Social Effects of Policy Reform. Edited byL. Demery et al. Washington DC: WorldBank.

Fujimura, M., and R. Bolt 2001.Consultation Mission to World Bank andIDB (back-to-office-report). Economicsand Research Department. ADB, Manila,24–28 September.

Garuda, G. 2000. The DistributionalEffects of IMF Programs: A Cross-Country Analysis. World Development, Vol.28, No 6.

Gilbert, J. and T. Wahl. 2000. AppliedGeneral Equilibrium Assessments ofTrade Liberalization in China. PaperPresented at the Workshop on China’sAccession to the WTO: An Overview ofRecent Analyses, Beijing.

Goldstein, M. and P. Montiel. 1986.Evaluating Fund Stabilisation Programswith Multi-Country Data. IMF StaffPapers, Vol. 29, No. 2.

Harrigan, J., R. Loader, and C. Thirtle.1992. FAO Training Materials forAgricultural Planning, No. 31. Rome.

Iqbar, Z. and R. Siddiqui. 1999. Impact ofFiscal Adjustment on IncomeDistribution in Pakistan, PakistanDevelopment Review, Vol. 38, No. 1.



RE

FE

RE

NC

ES

IMF. 1999. Poverty Reduction StrategyPapers: Operational Issues. IMF,Washington, DC.

———. 2001. Conditionality in Fund-Supported Programs: Policy Issues. PolicyDevelopment and Review Department.IMF, Washington, DC.

Jensen, M. and W. Meckling. 1976. Theoryof the Firm: Managerial Behavior, AgencyCost, and Ownership Structure. Journalof Financial Economics, Vol. 3, No. 4,pp. 305–360.

Johnson, J.H., and S.S. Wasty. 1993.Borrower Ownership of Adjustment Programsand the Political Economy of Reform. WorldBank Discussion Paper, No. 199. WorldBank, Washington, DC.

Kanbur, R. 1990. Projects versus PolicyReform. Paper presented at the WorldBank Annual Conference on DevelopmentEconomics, Washington, DC.

Kasper, W. 1999. Structural Change,Growth and “Social Justice”: An Essay;Structural Adjustment: Exploring thePolicy Issues. Workshop Proceedings.AusInfo, Canberra, pp. 125–61.

Kehoe, P. and T. Kehoe. 1994. A Primeron Static Applied General EquilibriumModels. Quarterly Review—Federal ReserveBank of Minneapolis, Vol. 18, No. 1.

Khan, H. 1999. Sectoral Growth andPoverty Alleviation: A MultiplierDecomposition Technique Applied toSouth Africa. World Development, Vol. 27,No. 3.

Khan, M. 1990. The Macro EconomicEffects of Fund-Supported AdjustmentPrograms. IMF Staff Papers, Vol. 37, No 2.

Kherallah, M. et al. (undated.) NewInstitutional Economics (NIE): What’sNew and What Does It Mean for IFPRI?

Klein, P. 1999. New Institutional Economics.Available: http://encyclo.findlaw.com/0530book.pdf.

Killick, T. 1993. The Adaptive Economy:Adjustment Policies in Small, Low-IncomeCountries. Economic DevelopmentInstitute. Washington DC: World Bank.

———. 1998. Aid and the Political Economyof Policy Changes. London: Routledge.

———, ed. 1995. The Flexible Economy:Causes and Consequences of the Adaptability ofNational Economies. London and New York:Routledge and Overseas DevelopmentInstitute.

Knapman, B. and C. Saldanha. 1999.Reforms in the Pacific: An Assessmentof the ADB’s Assistance for ReformPrograms in the Pacific. Office of PacificOperations. ADB, Manila.

Krugman, P. and Obstfeld, M. 2000.International Economics: Theory and Policy.Fifth edition. Addison-Wesley.

Kydd, J. 1998. Agricultural Policy Analysis.Wye College External Programme.University of London.

Lee, P.U. 1998. Report on Impact onPoverty of Philippine Power SectorRestructuring, University of Asia and thePacific. May.

154


Little, I.M.D. 1982. Economic Development:Theory, Policy and International Relations. NewYork: Basic Books Inc.

Lord, M. 2001. Macroeconomic Policiesfor Poverty Reduction in Cambodia.Report to Programs Department West.ADB, Manila. May.

McMillan, M., D. Rodrik, and K. HornWelch. 2002. When Economic Reform GoesWrong: Cashews in Mozambique. NationalBureau of Economic Research WorkingPaper No. w9117. August.

Miller, M., and C. Upton. 1986.Macroeconomics: A Neoclassical Introduction.University of Chicago Press.

Munasinghe, M. and J. Warford. 1982.Electricity Pricing: Theory and Case Studies.World Bank. Johns Hopkins UniversityPress.

Mundle, S. 1998. Financing HumanDevelopment: Some Lessons fromAdvanced Asian Countries. WorldDevelopment, Vol. 26, No. 4, pp. 659–672.

Nelson, E. 1998. A Review of AsianDevelopment Bank Practices forAssessing the Impact of Program Loanson Poverty, DAI.

Pearson, S. 1989. The Policy Analysis Matrixfor Agricultural Development . CornellUniversity Press.

Productivity Commission. 2001.Structural Adjustment Key Policy Issues.Commission Research Paper. AusInfo,Canberra.

Quibria, M.G. 2001. Growth and Poverty:Lessons from the East Asian MiracleRevisited. Preliminary draft. ADBInstitute, Tokyo.

Rao, M.G. 1998. Accommodating PublicExpenditure Policies: The Case of FastGrowing Asian Economies. WorldDevelopment, Vol. 26, No. 4, pp. 673–694.

Reed, G. 1996. The Use of CGE Modellingin the Analysis of Trade Policy Reform.Center for Research in EconomicDevelopment and International Trade,University of Nottingham, UK.

Robinson, S. 1989. Multisectoral Models.In Handbook for Development Economics.Edited by H.B. Chenery, and T.N.Srinivasan. Vol. 2. Amsterdam: North-Holland.

Rodrik, D. 1996. Understanding EconomicPolicy Reform. Journal of EconomicLiterature, Vol. 34, March, pp. 9–41.

———. 1999. Institutions for High-Quality Growth: What they are and Howto Acquire Them. Paper presented at theIMF Conference on Second-GenerationReforms. October.

Roumasset, J. 2000. Black Hole Security.Available: http://www2.soc.hawaii.edu/econ/workingpapers/005abstract.html

Working Paper No. 00-05. Paper preparedfor the PRAEO meetings of theInternational Western EconomicAssociation, Sidney, January.



RE

FE

RE

NC

ES

Sachs, J. and A. Warner. 1995. EconomicReform and the Process of GlobalIntegration. Brookings Papers on EconomicActivity, No. 118.

Sadoulet, E. and A. de Janvry. 1995.Quantitative Development Policy Analysis.Baltimore: Johns Hopkins UniversityPress.

Sahay, P., P. Cashin, and P. Mauro. 2001.Macroeconomic Policies and Poverty: A State ofPlay and a Research Agenda. Washington DC:World Bank.

Sahn, D., P. Dorosh, and S. Younger. 1999.A Reply to de Maio, Stewart and van derHoeven. World Development, Vol. 27, No. 3.

Saldanha, C. and J. Whittle. 1998. Using theLogical Framework for Sector Analysis andProject Design: A User’s Guide. Manila.

Scarf, H. and J. Shoven. 1984. AppliedGeneral Equilibrium Analysis. New York:Cambridge University Press.

Schiavo-Campo, S. and P. Sundaram.2001. To Serve and Preserve: Improving PublicAdministration in a Competitive World. Manila:ADB.

Sen, A. 1998. Human Development andFinancial Conservation. World Development,Vol. 26, No. 4, pp. 733–742.

Shoven, J.B. and J. Whalley. 1992. ApplyingGeneral Equilibrium . Cambridge:Cambridge University Press.

Squire L. 1989. Project Evaluation inTheory and Practice. In Handbook ofDevelopment Economics. Edited by Hollis

Chenery and T.N. Srinivasan.Amsterdam: Elsevier.

Stiglitz, J.E. 1998. Towards a NewParadigm for Development: Strategies,Policies and Processes. Prebisch Lectureat UNCTAD.

———. 1999. The World Bank at theMillennium. Economic Journal, Vol. 109,November, pp. f577–f597.

———. 2000. Challenges in the Analysisof the Role of Institutions in EconomicDevelopment. Keynote address at theVilla Borsig Workshop Series 2000: TheInstitutional Foundations of a MarketEconomy, sponsored by DSE.

Streeten, P. 1987. What Price Food?Macmillan Press.

———. 1989. A Survey of the Issues andOptions. In Structural Adjustment andAgriculture: Theory and Practice in Africa andLatin America. Edited by SimonCommander. London: OverseasDevelopment Institute.

Tarp, F. 1993. Stabilisation and StructuralAdjustment: Macroeconomic Frameworks forAnalysing the Crisis in Sub-Saharan Africa.London and New York: Routledge.

Thorbecke, E. 1991. Adjustment, Growthand Income Distribution in Indonesia.World Development, Vol. 19, No. 11.

———. 2001. Poverty Analysis andMeasurement within a GeneralEquilibrium Framework. Paper deliveredat Asia Pacific Forum on Poverty, Manila,February.

156


Thissen, M. 1998. A Classification of EmpiricalCGE Modelling. SOM Research Report99C01, University of Groningen, TheNetherlands.

Tsutsumida, T. and T. Wada. 1997.Incomplete Contracts, Transaction Cost Theoryand its Empirical Evidence. Available:www.iptp.go.jp/reserch_e/monthly/m-s e r c h / p o s t a l / 1 9 9 7 / n o 1 0 3 /monthly-1032.html.

Vane, H.R. and J.L. Thompson. 1985. AnIntroduction to Macroeconomic Policy. Secondedition. Harvester Press.

Vargas, E. D. Schreiner, G. Tembo, and D.Marcouiller. (undated.) Computable GeneralEquilibrium Modeling for Regional Analysis.The Web Book of Regional Science,Regional Research Institute, WestVirginia University. Available:www.rri.wvu.edu/WebBook/ Schreiner/Part%201.htm.

Warford, J. 1997. Marginal OpportunityCost Pricing for Municipal WaterSupply, Economy and EnvironmentalProgram for Southeast Asia.International Development ResearchCenter, Ottawa.

Williamson, O. 1981. The Economics ofOrganization: The Transaction CostApproach, American Journal of Sociology,Vol. 87, pp. 548–577.

World Bank. 1990. World Development Report1990. New York: Oxford University Press.

———. 1993. The East Asian Miracle. NewYork: Oxford University Press.

———. 1998. Programmatic and EmergencyAdjustment Lending: World Bank Guidelines.R98-249. October. Washington DC:World Bank.

———. 2000a. Annual Review of DevelopmentEffectiveness 2000. Operations EvaluationDepartment. Washington DC: WorldBank.

———. 2000b. World Development Report2000/2001: Attacking Poverty. OxfordUniversity Press for the World Bank.

———. 2000c. Fixing ESW: Where AreWe? Discussion draft. Operations Policyand Strategy Group. World Bank,Washington, DC.

———. 2001a. Integrated Macro-economic Model for Poverty Analysis(IMMPA) project. Available: http//:www1.worldbank.org/wbiep/macro-p r o g r a m / m e m / r i g h t _ I M M P A _m o d e l i n g _ f r a m e w o r k /index_immpa_modeling_framework.htm.

———. 2001b. Adjustment LendingRetrospective. Operations Policy andCountry Services. World Bank,Washington, DC.

———. 2001c. Poverty Reduction and theWorld Bank: Progress in Fiscal 2000 and2001. Poverty Reduction Group. WorldBank, Washington, DC.

———. 2002a. World Development Report2002: Building Institutions for Markets.Oxford University Press for the WorldBank.



RE

FE

RE

NC

ES

———. 2002b. A User’s Guide to Povertyand Social Impact Analysis. Draft.Poverty Reduction Group and SocialDevelopment Department. World Bank,Washington, DC. 19 April.

Zezza, A. and L. Lambi. 2001. Meso-Economic Filters Along the Policy Chain:Understanding the Links between PolicyReforms and Rural Poverty in LatinAmerica. Paper prepared for theAgriculture and Economic DevelopmentAnalysis Division (ESA) of the Food andAgricultural Organization (FAO) of theUnited Nations.

Documents

Economic Analysis of Policy-based Operations: Key Dimensions · Economic Analysis of Policy-Based Operations: Key Dimensions 1 O VERVIEW macroeconomic policy framework. A macro-meso-micro