Macroeconomic Policy and Poverty Reduction

Macroeconomic Policyand Poverty Reduction

Brian AmesWard Brown

Shanta DevarajanAlejandro Izquierdo

INTERNATIONAL MONETARY FUND

WORLD BANK

©International Monetary Fund. Not for Redistribution

©2001 International Monetary Fund

ISBN: 1-58906-017-2

Production: IMF Graphics Section

Cover design: Sanaa ElaroussiCover photographs: (from left to right) Marcel Crozet/WHO,

Padraic Hughes/IMF, Curt Carnemark/World BankTypesetting: Julio R. Prego

To order additional copies of this publication or acatalog of other IMF publications, please contact:

International Monetary Fund, Publication Services700 19th Street, N.W., Washington, D.C. 20431, U.S.A.

Tel.: (202) 623-7430 Telefax: (202) 623-7201E-mail: [email protected]: http://www.imf.org


http://www.imf.org

Contents

1. Introduction 1

2. The Links Between Macroeconomic Policy andPoverty Reduction: Growth Matters . 4

Macroeconomic Stability Is Necessary for Growth . . . 5Macroeconomic Instability Hurts the Poor 7Composition and Distribution of Growth

Also Matter 8Implications for Macroeconomic Policy 10

3. Macroeconomic Stability and Economic Growth 13Sources of Instability 14Stabilization 14Elements of Macroeconomic Stability 17

4. Growth-Oriented Macroeconomic Policies andPoverty Outcomes 19

Financing Poverty Reduction Strategies 19Fiscal Policy 26Monetary and Exchange Rate Policies 33Policies to Insulate the Poor Against Shocks 40

References 57

i i i


This pamphlet is based on Volume 2 of the Poverty Reduction StrategySourcebook, which is available at http://www.worldbank.org/poverty/strate-gies/sourctoc.htm. The sourcebook is a guide to assist countries in devel-oping and strengthening their poverty reduction strategies. It wasprepared by the staffs of the World Bank and the IMF.


http://www.worldbank.org/poverty/strategies/sourctoc.htm


Introduction

Poverty is a multidimensional problem that goes beyondeconomics to include, among other things, social, political,and cultural issues (see Box 1). Therefore, solutions topoverty cannot be based exclusively on economic policies,but require a comprehensive set of well-coordinated mea-sures. Indeed, this is the foundation for the rationale under-lying comprehensive poverty reduction strategies.1 So whyfocus on macroeconomic issues? Because economic growthis the single most important factor influencing poverty, andmacroeconomic stability is essential for high and sustainablerates of growth.2 Hence, macroeconomic stability should bea key component of any poverty reduction strategy.

Macroeconomic stability by itself, however, does not en-sure high rates of economic growth. In most cases, sustainedhigh rates of growth also depend upon key structural mea-

1 There has been an emerging consensus on how to make actions at thecountry level, and the support of development partners, more effective inbringing about sustainable poverty reduction. This consensus indicates aneed for poverty reduction strategies that are country-driven, with broadparticipation of civil society, elected officials, key donors, and relevant in-ternational finance institutions; outcome-oriented; and developed from anunderstanding of the nature and determinants of poverty. Under the newframework, the country-led strategy would be presented in a Poverty Re-duction Strategy Paper (PRSP), which is expected to become a key instru-ment for a country's relations with the donor community.

2 Macroeconomic stability is a situation where key economic relation-ships are broadly in balance and sustainable.

1

1


Box 1. Definition and Measurement of Poverty

The World Bank's 2000 World Development Report definespoverty as an unacceptable deprivation in human well-being thatcan comprise both physiological and social deprivation. Physio-logical deprivation involves the non-fulfillment of basic materialor biological needs, including inadequate nutrition, health, edu-cation, and shelter. A person can be considered poor if he orshe is unable to secure the goods and services to meet these ba-sic material needs. The concept of physiological deprivation isthus closely related to, but can extend beyond, low monetary in-come and consumption levels. Social deprivation widens theconcept of deprivation to include risk, vulnerability, lack of au-tonomy, powerlessness, and lack of self-respect. Given that coun-tries' definitions of deprivation often go beyond physiologicaldeprivation and sometimes give greater weight to social depriva-tion, local populations (including poor communities) should beengaged in the dialogue that leads to the most appropriate defi-nition of poverty in a country.

sures, such as regulatory reform, privatization, civil service

reform, improved governance, trade liberalization, and

banking sector reform, many of which are discussed at

length in the Poverty Reduction Strategy Sourcebook, published

by the World Bank.3 Moreover, growth alone is not sufficient

for poverty reduction. Growth associated with progressive

distributional changes will have a greater impact on poverty

than growth that leaves distribution unchanged. Hence,

policies that improve the distribution of income and assets

within a society, such as land tenure reform, pro-poor public

expenditure, and measures to increase the poor's access to

3 The sourcebook is available at http://www.worldbank.org/poverty/strategies/sourctoc.htm.

2




financial markets, will also form essential elements of acountry's poverty reduction strategy.4

To safeguard macroeconomic stability, the governmentbudget, including the country's poverty reduction strategies,must be financed in a sustainable, noninflationary manner.The formulation and integration of a country's macroeco-nomic policy and poverty reduction strategy are iterativeprocesses. Poverty reduction strategies need first to be artic-ulated (i.e., objectives and policies specified), then costed,and finally financed within the overall budget in a noninfla-tionary manner. The amount of finance, much of which willbe on concessional terms, is, however, not necessarily fixedduring this process: if credible poverty reduction strategiescannot be financed from available resources, World Bankand IMF staff should and will actively assist countries intheir efforts to raise additional financial support from thedonor community. Nonetheless, in situations where financ-ing gaps remain, a country would have to revisit the inter-mediate objectives of their strategy and reexamine theirpriorities. Except in cases where macroeconomic imbalancesare severe, there will usually be some scope for flexibility insetting short-term macroeconomic targets. However, the ob-jective of macroeconomic stability should not becompromised.

4 These points are reflected in the design of programs supported by theIMF's Poverty Reduction and Growth Facility (PRGF), which are derivedfrom a country's own poverty reduction strategy. The strategy itself shouldbe based upon fully integrated macroeconomic, structural, and social poli-cies. See Key Features of IMF Poverty Reduction and Growth Facility (PRGF) Sup-ported Programs, August 16, 2000, available at http://www.imf.org/external/np/prgf/2000/eng/key.htm.

3


http://www.imf.org/external/np/prgf/2000/eng/key.htm


The Links Between Macroeconomic Policyand Poverty Reduction: Growth Matters

Economic growth is the single most important factor influenc-ing poverty. Numerous statistical studies have found a strongassociation between national per capita income and nationalpoverty indicators, using both income and nonincome mea-sures of poverty.5 One recent study consisting of 80 coun-tries covering four decades found that, on average, theincome of the bottom one-fifth of the population rose one-for-one with the overall growth of the economy as definedby per capita GDP (Dollar and Kraay, 2000). Moreover, thestudy found that the effect of growth on the income of thepoor was on average no different in poor countries than inrich countries, that the poverty-growth relationship had notchanged in recent years, and that policy-induced growth wasas good for the poor as it was for the overall population.Another study that looked at 143 growth episodes alsofound that the "growth effect" dominated, with the "distri-bution effect" being important in only a minority of cases(White and Anderson, forthcoming). These studies, how-

5 Examples include the relationship between infant mortality rates andper capita income, the ratio of female to male literacy and per capita in-come, and average consumption and the incidence of income poverty. Inall three cases, national poverty indicators improved as per capita incomerose. See the discussion in the World Bank's World Development Report,2000.

4

2


ever, establish association, but not causation. In fact, thecausality could well go the other way. In such cases, povertyreduction could in fact be necessary to implement stablemacroeconomic policies or to achieve higher growth.

Studies show that capital accumulation by the private sec-tor drives growth.6 Therefore, a key objective of a country'spoverty reduction strategy should be to establish conditionsthat facilitate private sector investment. No magic bullet canguarantee increased rates of private sector investment. In-stead, in addition to a sustainable and stable set of macroe-conomic policies, a country's poverty reduction policyagenda should, in most cases, extend across a variety of pol-icy areas, including privatization, trade liberalization, bank-ing and financial sector reforms, labor markets, theregulatory environment, and the judicial system. Theagenda will certainly include increased and more efficientpublic investment in a country's health, education, andother priority social service sectors.7

Macroeconomic Stability Is Necessary for GrowthMacroeconomic stability is the cornerstone of any success-

ful effort to increase private sector development and eco-nomic growth (see Box 2). Cross-country regressions using alarge sample of countries suggest that growth, investment,and productivity are positively correlated with macroeco-

6 Devarajan, Swaroop, and Zou (1997) and Devarajan, Easterly, andPack (forthcoming).

7 There is little empirical evidence, however, that public sector capitalexpenditure has a positive impact on growth, reflecting the tendency forsuch investment in the past to be wasteful or inefficient. This does notmean public investment is unimportant—only that efficiency considera-tions must be central in any public investment program. See Easterly andRebelo (1993), Devarajan, Swaroop, and Zou (1997).

5


Box 2. Macro economic Stability

Macroeconomic stability exists when key economic relation-ships are in balance—for example, between domestic demandand output, the balance of payments, fiscal revenues and expen-diture, and savings and investment. These relationships, how-ever, need not necessarily be in exact balance. Imbalances suchas fiscal and current account deficits or surpluses are perfectlycompatible with economic stability provided that they can be fi-nanced in a sustainable manner.

There is no unique set of thresholds for each macroeconomicvariable between stability and instability. Rather, there is a con-tinuum of various combinations of levels of key macroeconomicvariables (e.g., growth, inflation, fiscal deficit, current accountdeficit, international reserves) that could indicate macroeco-nomic instability. While it may be relatively easy to identify acountry in a state of macroeconomic instability (e.g., large cur-rent account deficits financed by short-term borrowing, highand rising levels of public debt, double-digit inflation rates, andstagnant or declining GDP) or stability (e.g., current accountand fiscal balances consistent with low and declining debt levels,inflation in the low single digits, and rising per capita GDP),there is a substantial "gray area" in between where countries en-joy a degree of stability, but where macroeconomic performancecould clearly be improved.

Finally, macroeconomic stability depends not only on themacroeconomic management of an economy, but also on thestructure of key markets and sectors. To enhance macroeco-nomic stability, countries need to support macroeconomic pol-icy with structural reforms that strengthen and improve thefunctioning of these markets and sectors.

nomic stability (Easterly and Kraay, 1999). Although it is dif-ficult to prove the direction of causation, these results con-firm that macroeconomic instability has generally been associatedwith poor growth performance. Without macroeconomic stabil-ity, domestic and foreign investors will stay away and re-

6


sources will be diverted elsewhere. In fact, econometric evi-dence of investment behavior indicates that in addition toconventional factors (i.e., past growth of economic activity,real interest rates, and private sector credit), private invest-ment is significantly and negatively influenced by uncer-tainty and macroeconomic instability (see, for example,Ramey and Ramey, 1995).

Macroeconomic Instability Hurts the PoorIn addition to low (and sometimes even negative) growth

rates, other aspects of macroeconomic instability can place aheavy burden on the poor. Inflation, for example, is a re-gressive and arbitrary tax, the burden of which is typicallyborne disproportionately by those in lower income brackets.The reason is twofold. First, the poor tend to hold most oftheir financial assets in the form of cash rather than in in-terest-bearing assets. Second, they are generally less ablethan are the better off to protect the real value of their in-comes and assets from inflation. In consequence, pricejumps generally erode the real wages and assets of the poormore than those of the non-poor. Moreover, beyond certainthresholds, inflation also curbs output growth, an effect thatwill impact even those among the poor who infrequently usemoney for economic transactions.8 In addition, low outputgrowth that is typically associated with instability can have alonger-term impact on poverty (a phenomenon known as"hysteresis"). This phenomenon typically operates throughshocks to the human capital of the poor. In Africa, for in-stance, there is evidence that children from poor families

8 Empirical evidence confirms a strong negative relationship between in-flation and economic growth at all but the lowest levels of inflation. SeeFischer (1993), Bruno and Easterly (1998), Ghosh and Phillips (1998),and Sarel (1996).

7


drop out of school during crises. Similarly, studies for LatinAmerican countries suggest that adverse terms-of-tradeshocks explain part of the decline of schooling attainment(see, for example, Behrman, Duryea, and Szeleky, 1999).

Composition and Distribution of Growth Also MatterAlthough economic growth is the engine of poverty reduc-

tion, it works more effectively in some situations than in oth-ers.9 Two key factors that appear to determine the impact ofgrowth on poverty are the distributional patterns and the sec-toral composition of growth.

If the benefits of growth are translated into poverty reduc-tion through the existing distribution of income, then moreequal societies will be more efficient transformers of growth intopoverty reduction. A number of empirical studies have foundthat the responsiveness of income poverty to growth in-creases significantly as inequality is lowered.10 This is alsosupported by a recent cross-country study that found thatthe more equal the distribution of income in a country, thegreater the impact of growth on the number of people inpoverty (Ravallion, 1997). Others have suggested thatgreater equity comes at the expense of lower growth andthat there is a trade-off between growth and equity when itcomes to poverty reduction.11 A large number of recent em-

9 For any given increment in per capita income, the impact on povertywill depend on how that increment is distributed across the population.While growth is almost always accompanied by a reduction in incomepoverty, and negative growth is accompanied by an increase in poverty, forany given growth rate the impact on poverty can vary substantially.

10 Ravallion (1997), Datt and Ravallion (1992), and Kakwani (1993).11 To the extent that people with high income save a larger proportion

of their income than do those with low income, policies that redistributeincome in favor of the lower-income population may impede savings and,to the extent that such savings are channeled into productive investment,long-term growth.

8


pirical studies, however, have found that there is not neces-sarily such a trade-off12 and that equity in its various dimen-sions is growth enhancing.13

The sectoral composition of growth can determine the im-pact that growth will have on poverty. Conventional wisdomhas been that growth in sectors of the economy where thepoor are concentrated will have a greater impact on reduc-ing poverty than growth in other sectors—indeed, this is al-most a tautology. For example, it is often argued that incountries where most of the poor live in rural areas, agricul-tural growth reduces poverty because it generates incomefor poor farmers and increases the demand for goods andservices that can easily be produced by the poor.14 Variouscountry-specific and cross-country studies have shown thatgrowth in the agricultural and tertiary sectors has had a ma-jor effect on reducing poverty, while growth in manufactur-

12 This refers to developing economies, where often income (andwealth) inequality is particularly acute. In general, there is likely to be apoint beyond which greater equity is incompatible with adequate laborand enterprise incentives, which, in turn, would be detrimental to growth.See Alesina and Rodrik (1994); Benabou (1996); Birdsall and Londono(1997); Deninger and Squire (1998); Perotti (1992, 1993, and 1996); andPersson and Tabellini (1994).

13 By increasing the human capital of the poor, redistributive policiescan increase the productivity of the workforce, thereby enhancing growth.Others have argued that there is also a political economy channel aswell—in countries with greater income equality there is greater politicalsupport for public policies that are more conducive to growth. See Alesinaand Rodrik (1994), and Persson and Tabellini (1994). For empirical sup-port for this effect, see Deininger (1999); Thomas and Wang (1998);Klasen (1999); and Dollar and Gatti (1999). For dissenting views, seeForbes (2000) and Li , Xie, and Zou (1999).

14 It is also often argued that if growth results in the expansion of low-skilled employment, then the poor are more likely to be the beneficiariesof the growth. One recent cross-country study (Fallon and Hon, 1999)found that the more labor-intensive the growth pattern, the faster the de-cline in the incidence of poverty.

9


ing has not.15 This reinforces the case for duty-free access toindustrial country markets for agricultural exports from low-income countries. The links may be more complex over thelong run, however. While faster growth in agriculture mayaddress rural poverty in the short-term, reliance on agricul-tural activity may also intensify output variability, which, inturn, would contribute to increasing rather than decreasingpoverty. A more diversified economy with a vibrant manufac-turing sector might offer the best chances for a sustainableimprovement in living standards in the long run.

Implications for Macroeconomic PolicyWhat are the implications of these empirical findings for

macroeconomic policy? First, in light of the importance ofgrowth for poverty reduction, and of macroeconomic stabil-ity for growth, the broad objective of macroeconomic policyshould be the establishment, or strengthening, of macroeco-nomic stability. Policymakers should therefore define a set ofattainable macroeconomic targets (i.e., growth, inflation, ex-ternal debt, and net international reserves) with the objec-tive of maintaining macroeconomic stability, and pursuemacroeconomic policies (fiscal, monetary, and exchangerate) consistent with those targets. In cases where macroeco-nomic imbalances are less severe, a range of possible targetsmay be consistent with the objective of stabilization. Precisetargets can then be set within that range, in accordance withthe goals and priorities in the country's poverty reductionstrategy (see the section on fiscal policy later in thispamphlet).

15 Datt and Ravallion (1998), Thorbecke and Jung (1996), Timmer(1997), and Bourguignon and Morrisson (1998).

10


Second, most developing countries will likely have substan-tial scope for enhancing the quality of growth, that is, thedegree to which the poor share in the fruits of such growth,through policies aimed at improving income distribution.These policies (e.g., land tenure reform, changes inmarginal and average tax rates, increases in pro-poor socialspending, etc.) often are politically charged, and usually re-quire supporting structural and governance reforms thatwould empower the poor to demand resources and/or en-sure that resources intended for them are not diverted toother groups of the population. As these topics pertainmore broadly to political economy, rather than exclusively tomacroeconomics, they are beyond the scope of this pam-phlet. But they reinforce the point that economic growthalone is not sufficient for poverty reduction and that com-plementary redistributional policies may be needed to en-sure that the poor benefit from growth.

Finally, while issues regarding the composition of growthalso go beyond strict macroeconomics, several general policyobservations can be made. There is a general consensus thatpolicies that introduce distortions in order to influencegrowth in a particular sector can hamper overall growth.The industrial policies pursued by many African developingcountries in the 1960s have long been discredited (WorldBank, 1982). Instead, strategies for sector specific growthshould focus on removing distortions that impede growth ina particular sector. In addition, policymakers should imple-ment policies that will empower the poor and create theconditions that would permit them to move into new as wellas existing areas of opportunity, thereby allowing them tobetter share in the fruits of economic growth. The objectivesof such policies should include creating a stable environ-ment and level playing field conducive to private sector in-

11


vestment and broad-based economic growth; removing thecultural, social, and economic constraints that prevent thepoor from making full use of their existing asset base andaccessing markets; and increasing the human capital base ofthe poor through the provision of basic health and educa-tion services. Using these policies, and the redistributivepolicies described above, policymakers can target "pro-poor"growth—that is, they can attempt to maximize the beneficialimpact of sustained economic growth on poverty reduction.

12


Macroeconomic Stability andEconomic Growth

Broadly speaking, two considerations underlie macroeco-nomic policy recommendations. First, there needs to be anassessment of the appropriate policy stance to adopt in a given setof circumstances (i.e., should fiscal and/or monetary policybe tightened or loosened?). Second, there is the choice of spe-cific macroeconomic policy instruments that would be beneficialfor a country to adopt (e.g., the use of a nominal anchor, avalue-added tax (VAT), etc.). In practice, these two consider-ations are closely linked. Adjusting a policy stance is oftendone via the adoption of a new instrument (or the modifica-tion of an existing one). More important, both considera-tions are essential to efforts to enhance an economy'sstability.

The specific stance must fit each country's particular situa-tion. These situations can be put into three broad classes:(1) instability/disequilibrium; (2) stabilization (e.g., transi-tion from instability to stability); and (3) stability/steady eco-nomic growth. This Section briefly discusses howmacroeconomic policies can contribute to stability. Forcountries that enjoy stable macroeconomic conditions, thereis somewhat greater flexibility in the choice of appropriatestance for macroeconomic policy. The central issue for thesecountries will be to ensure that the financing of their

13

3


poverty reduction strategies does not jeopardize macroeco-nomic stability, which will be discussed in the last section ofthis pamphlet.

Sources of InstabilityThere are two main sources of economic instability,

namely exogenous shocks and inappropriate policies. Exoge-nous shocks (e.g., terms of trade shocks, natural disasters, re-versals in capital flows, etc.) can throw an economy intodisequilibrium and require compensatory action. For exam-ple, many low income countries have a narrow export base,often centered on one or two key commodities. Shocks tothe world price of these commodities can therefore have astrong impact on the country's income. Even diversifiedeconomies, however, are routinely hit by exogenous shocks,although, reflecting their greater diversification, shocks usu-ally need to be particularly large or long-lasting to destabi-lize such an economy. Alternatively, a disequilibrium can be"self-induced" by poor macroeconomic management. For ex-ample, an excessively loose fiscal stance can increase aggre-gate demand for goods and services, which places pressureon the country's external balance of payments as well as onthe domestic price level. At times, economic crises are theresult of both external shocks and poor management.

StabilizationIn most cases, addressing instability (i.e., stabilization) will

require policy adjustment) whereby a government introducesnew measures (possibly combined with new policy targets)in response to the change in circumstances.16 Adjustment

16 In certain cases, the return to a steady growth state may also requirestructural reform and measures to improve the functioning of markets.

14


will typically be necessary if the source of instability is a per-manent (i.e., systemic) external shock or the result of ear-lier, inappropriate macroeconomic policies. However, if thesource of instability can be clearly identified as a temporaryshock (e.g., a one-time event) then it may be appropriatefor a country to accommodate it.17 Identifying whether aparticular shock is temporary or is likely to persist is easiersaid than done. Since there is often a considerable degreeof uncertainty surrounding such a judgment, it is usuallywise to err somewhat on the side of caution by assumingthat the shock will largely persist and by basing the corre-sponding policy response on the appropriate adjustment.

In most circumstances where adjustment is necessary, bothmonetary (or exchange rate) and fiscal instruments willhave to be used. In particular, successful adjustment to apermanent unfavorable shock that worsens the balance ofpayments will often require a sustained tightening of the fis-cal stance, as this is the most immediate and effective way toincrease domestic savings and to reduce domestic demand—two objectives typically at the center of stabilizationprograms.

Adjustment policies may contribute to a temporary con-traction of economic activity, but this contingency shouldnot be used to argue against implementing adjustment poli-cies altogether, as the alternative may be worse. Attemptingto sustain aggregate demand through unsustainable policieswill almost certainly aggravate the long-run cost of a shock,and could even fail in the short run to the extent that it un-

17 Broadly speaking, this means leaving the underlying stance ofmacroeconomic policy unchanged (or, in some cases, the stance may beadjusted temporarily to mitigate the impact of the shock) and adjustingpolicy targets in a way that takes into account the impact of the shock.However, if such a policy stance cannot be financed in a noninflationaryway, then some adjustment will also be necessary.

15


dermines confidence. In the long run, greater benefits tothe poor are to be had as a result of the restoration ofmacroeconomic stability. The appropriate policies to protectthe poor during adjustment are to maintain, or even in-crease, social expenditures and to adopt, where feasible,compensatory measures that would insulate or offset tempo-rary adverse impacts to the fullest extent possible.18 This isbest done by devoting resources to the establishment of ef-fective social safety nets,19 as an enduring part of a country'spoverty reduction strategy, rather than as a response to cri-sis. Countries that lack such resources/safety nets could beforced to either subject their poor to the short-term adverseeffects of stabilization or to delay the pace with whichmacroeconomic adjustment proceeds (and put off the corre-sponding long-term benefits to economic growth andpoverty reduction).

Countries in macroeconomic crisis typically have littlechoice but to stabilize quickly, but for countries in the"gray" area of partial stability, finding the right pace mayprove difficult. In some cases, a lack of financing will drivethe pace of stabilization. Where financing is not a con-straint, however, policymakers will need to assess and care-fully weigh various factors on a case-by-case basis in choosingthe most appropriate pace of stabilization.

18 Indeed, a key feature of programs supported by the IMF's Poverty Re-duction and Growth Facility (PRGF) is to assess the distributional impactof key macroeconomic policies and to put in place countervailing mea-sures needed to protect the poor. See Key Features of IMF Poverty Reductionand Growth Facility (PRGF) Supported Programs, August 16, 2000 athttp://www.imf.org/external/np/prgf/2000/eng/key.htm.

19 Social safety nets are designed to mitigate possible adverse effects ofreform measures on the poor. These instruments include temporary ar-rangements, as well as existing social protection measures reformed andadapted for this purpose, such as limited food subsidies, social security ar-rangements for dealing with various life cycle and other contingencies,and targeted public works. See Chu and Gupta (1998).

16



Elements of Macroeconomic StabilityMacroeconomic policies influence and contribute to the

attainment of rapid, sustainable economic growth aimed atpoverty reduction in a variety of ways. By pursuing soundeconomic policies, policymakers send clear signals to theprivate sector. The extent to which policymakers are able toestablish a track record of policy implementation will influenceprivate sector confidence, which will, in turn, impact uponinvestment, economic growth, and poverty outcomes.

Prudent macroeconomic policies can result in low and sta-ble inflation. Inflation hurts the poor by lowering growth andby redistributing real incomes and wealth to the detrimentof those in society least able to defend their economic inter-ests. High inflation can also introduce high volatility in rela-tive prices and make investment a risky decision. Unlessinflation starts at very high levels, rapid disinflation can alsohave short-run output costs, which need to be weighedagainst the costs of continuing inflation.

By moving toward debt sustainability, policymakers will helpcreate the conditions for steady and continuous progress ongrowth and poverty reduction by removing uncertainty as towhether a government will be able to service new debt. Bykeeping domestic and external debt at levels that can be ser-viced in a sustainable manner without unduly squeezingnondebt expenditure, policymakers can also ensure that ad-equate domestic resources are available to finance essentialsocial programs.

Inappropriate exchange rate policies distort the compositionof growth by influencing the price of tradable versus non-tradable goods. Household survey data for a number ofcountries indicate that the poor tend to consume higheramounts of nontradable goods while generating relativelymore of their income from tradable goods (Sahn, Dorosh,

17


and Younger, 1997). Hence, in addition to distorting tradeand inhibiting growth, an overly appreciated exchange ratecan impair the relative incomes and purchasing power ofthe poor.

By building and maintaining an adequate level of net interna-tional reserves, a country can weather a temporary shockwithout having to reduce essential pro-poor spending. Exter-nal shocks can be particularly detrimental to the poor be-cause they can lower real wages, increase unemployment,reduce nonlabor income, and limit private and net govern-ment transfers. The level of "adequate" reserves depends onthe choice of exchange rate regime.

18


Growth-Oriented Macroeconomic Policiesand Poverty Outcomes

Since the emphasis of this pamphlet is on the role ofmacroeconomic policy in supporting a country's poverty re-duction strategy, the discussion of macroeconomic policiesin this section focuses on countries that have broadlyachieved macroeconomic stability. Recent data indicate thatmany developing countries are presently in a state ofmacroeconomic stability (see Tables 1-3 at the end of thispamphlet). When formulating a country's poverty reductionstrategy, policymakers will need to assess and determinewhat is the most appropriate combination of key macroeco-nomic targets that would preserve macroeconomic stabilityin their particular circumstance. Three key issues are dis-cussed in this section: (1) how to finance poverty-reducingspending in a way that doesn't endanger macroeconomicstability; (2) what specific policies can be adopted to im-prove macroeconomic performance; and (3) policies to pro-tect the poor from domestic and external shocks.

Financing Poverty Reduction Strategies

Once a country has developed a comprehensive and fullycosted draft of its poverty reduction strategy, it will need toensure that the strategy can be pursued and financed in amanner that does not jeopardize its macroeconomic stability

19

4


and growth objectives.20 To do so, policymakers need to in-tegrate their poverty reduction and macroeconomic strate-gies into a consistent framework. The following paragraphspresent a conceptual framework that could be useful to poli-cymakers in determining whether their poverty reductionstrategy is consistent with their macroeconomic objectives.

Given that it is difficult to determine beforehand what thegrowth target should be, policymakers may wish to considerdeveloping alternative macroeconomic scenarios that takeinto consideration possible variations in the rate of eco-nomic growth. Such scenarios could be usefully discussedwith stakeholders and development partners with a view toassessing the impact of lower-than-projected economicgrowth on key macroeconomic targets and poverty outcomesand to developing appropriate contingencies. The mostlikely or "base case" scenario would then be used as the ba-sis for carrying out an initial attempt aimed at integratingthe macroeconomic and poverty reduction strategies into aconsistent framework. Once this has been accomplished,similar exercises could be carried out regarding the othercontingency scenarios for reference during the implementa-tion stage of the strategy.

Figure 1 shows the various macroeconomic linkages andconstraints within a country and highlights the main trade-offs facing policymakers. The starting point is the initial ar-ticulation of the country's poverty reduction strategy, based ondiscussions with representatives of the government, stake-holders, and development partners. Ideally, these discussionswill have resulted in the development of a comprehensive

20 Even if the strategy can be fully financed with concessional resources,policymakers will need to assess the degree to which poverty-reducingspending may place pressure on the price of nontraded goods and therebythreaten stability.

20


action plan that identifies priority sectoral policies to bepursued in support of poverty reduction, including in theareas of education, health, and rural infrastructure. Giventhat poverty is multidimensional, the action plan will alsolikely include priority measures with regard to governance,structural reform, and other relevant areas, each of whichmay have budgetary implications.

The first step will be to provide a full costing of the envisagedpoverty reduction strategy. A comprehensive system for budgetformulation of poverty reduction strategies requires the de-velopment of Medium-Term Expenditure Frameworks(MTEF), which currently exist in only a limited number ofcountries (e.g., Ghana and Uganda). Details regarding howsuch costing exercises can be carried out are presented inChapter 5 of the Poverty Reduction Strategy Sourcebook, "PublicSpending for Poverty Reduction"21

The second step involves an assessment of the govern-ment's spending program with regard to priority spending,nondiscretionary spending, and discretionary nonpriorityspending. In doing so, policymakers should consider thescope for reallocating existing government spending intopriority areas and away from nonproductive, nonpriorityspending, as well as from areas where a rationale for publicintervention does not exist.

The third step involves an assessment of domestic andexternal sources of budget finance. This would include a re-view of (1) the existing tax and nontax revenue base, in-cluding the effect of any changes in the tax systemenvisaged under the poverty reduction strategy; (2) thescope for financing public spending through net domestic

21 The Sourcebook can be found at http://www.worldbank.org/poverty/strategies/sourctoc.htm.

21




borrowing in light of the need to maintain macroeconomicstability and to ensure adequate availability of credit to theprivate sector in support of private sector development andeconomic growth; and (3) the scope for external financing(e.g., grants, net external borrowing, and debt relief) that isrealistic and sustainable under the present circumstances.

Once policymakers have carried out these assessments,they can then determine whether the desired poverty reduc-tion strategy can be financed in a manner consistent withthe country's growth and stability objectives. In this regard,it is important to note that there are no rigid, pre-deter-mined limits regarding a country's fiscal stance (such as, forexample, "the budget deficit must not be more than 'x' per-cent of GDP"). Rather, arriving at an appropriate, integratedpoverty reduction and macroeconomic framework will re-quire juggling a large number of parameters and weighingthe trade-offs between multiple objectives. The linkages inFigure 1 are meant to illustrate that this is an iterative pro-cess. In this regard, quantitative frameworks that could assistpolicymakers in assessing the distributional implications oftheir macroeconomic policies would be particularly useful.Such frameworks, however, are presently only at a nascentstage of development (see Box 3).

If there remains an imbalance between spending and ex-pected financing that could jeopardize the country'smacroeconomic growth and stability objective, one optionwould be to ascertain the extent to which additional exter-nal financing may be available. But, as discussed earlier, poli-cymakers would need to assess the extent to whichaccommodating such expenditure could place pressure onthe price of nontraded goods and jeopardize stability. Sincethe development of a poverty reduction strategy involves aparticipatory process that includes the country's develop-

22


Box 3. Quantitative Frameworks for Assessing the DistributionalImpact of Macroeconomic Policies

In developing poverty reduction strategies, policymakerswould benefit from a quantitative framework that they coulduse to assess the distributional impact of the macroeconomicpolicy options under consideration. Such a framework would beuseful because the links between macroeconomic policies andpoverty are complex. A quantitative framework that identifiesthe critical relationships on which the outcome depends couldtherefore assist countries in assessing these trade-offs.

What would be some of the desirable characteristics of such aquantitative framework? First, the framework should be capableof identifying some of the critical trade-offs in poverty-reducingmacroeconomic policies. For example, how do the costs (interms of poverty) of higher spending (and higher fiscal deficits)compare with the benefits of targeting that spending on thepoor? Second, the framework should be consistent with eco-nomic theory on the one hand, and with basic data availability,such as national accounts and household income and expendi-ture surveys, on the other. Otherwise, the frameworks will notbe able to foster a dialogue between conflicting parties on theseissues. Third, and most important, the framework should besimple enough that government officials can use it on theirdesktop computers. This means that it should not make unduedemands on data, and it should be based on readily availablesoftware, such as Microsoft Excel™.

World Bank staff is presently developing alternative quantita-tive frameworks that could be used to evaluate some of themacroeconomic aspects of poverty reduction strategies.1 It is ex-pected that other possible quantitative frameworks will be devel-oped over time that could assist country teams in this regard.

1 See Agenor and others (2000). In developing this particular framework,the authors opted for a modular approach that allows different models tobe incorporated as alternative sub-components of the overall framework.

23


Figure 1: Financing Poverty Reduction Strategies in a Sustainable Manner

POVERTYREDUCTIONSTRATEGY

Governance Structural

GrowthTarget

Credit toGovernment

Macro Sectoral Other

InflationTarget

ReservesTarget

FISCAL ACCOUNTS

BUDGET FINANCING BUDGET EXPENDITURE

Tax Revenues(Step 3a)

Priority Spending(Step 1)

New DomesticBorrowing(Step 3b)

Discretionary nonpriority spending

(Step 2a)

24Private SectorDevelopment

BANK CREDIT

Private SectorCredit

Money

Target=

+

+


External Financing Spending(Step 3c)

—Grants—Net borrowing—Debt relief

NondiscretionarySpending(Step 2b)

Macroeconomic objective andpoverty reduction are now integratedinto a consistent framework

Imbalance?

No Yes

Reconsider:(Step 4)

• Macro targets• Poverty goals• Revenue effort• Spending program• Domestic financing• External financing

25

NFATarget


ment partners, the case for additional donor support can beexamined. To the extent possible, donors should be encour-aged to make medium-term aid commitments in support ofa country's poverty reduction strategy so that the countrycan have confidence as it begins new spending programsthat these activities can be sustained.22

If the desired poverty reduction program cannot befinanced in a manner consistent with the country's eco-nomic stability and growth objectives, then policymakers willneed to reconsider the parameters discussed above. Keyquestions would include: Is there further scope for domesticrevenue mobilization? Can discretionary nonpriority spend-ing be cut back more? Is there scope for cutting back cer-tain priority spending without undermining the povertyreduction objective? Can the domestic financing target berelaxed without jeopardizing macroeconomic stability or pri-vate sector development objectives? Can the macroeconomictargets be modified in a manner that would not underminethe interrelated objectives of rapid economic growth, lowand stable inflation, and poverty reduction? The answers tothese questions will determine the extent to which the de-sired poverty reduction programs can be pursued in the cur-rent period.

Fiscal PolicyFiscal policy can have a direct impact on the poor, both

through the government's overall fiscal stance and throughthe distributional implications of tax policy and public

22 Ensuring there is appropriate flexibility in fiscal targets and support-ing authorities' efforts to secure commitments of higher donor flows whenwarranted are key features of the IMF's PRGF-supported programs. See KeyFeatures of IMF Poverty Reduction and Growth Facility (PRGF) Supported Pro-grams, August 16, 2000, at http://www.inf.org/external/np/prgf/2000/eng/key.htm.

26


http://www.inf.org/external/np/prgf/2000/eng/key.htm

http://www.inf.org/external/np/prgf/2000/eng/key.htm

spending. Structural fiscal reforms in budget and treasurymanagement, public administration, governance, trans-parency, and accountability can also benefit the poor interms of more efficient and better targeted use of public re-sources. As indicated above, there is no rigid, pre-deter-mined limit on what would be an appropriate fiscal deficit.An assessment would need to be based on the particular cir-cumstances facing the country, its medium-term macroeco-nomic outlook, and the scope for external budgetaryassistance. The terms on which external assistance is avail-able are also important. There is a strong case, for instance,for allowing higher grants to translate into higher spendingand deficits, to the extent that those grants can reasonablybe expected to continue in the future, and provided thatthe resources can be used effectively.

With regard to the composition of public expenditure,policymakers will need to assess not only the appropriate-ness of the proposed poverty reduction spending program,but also of planned nondiscretionary, and discretionary non-priority, spending. In so doing, they will need to take intoparticular consideration the distributional and growth im-pact of spending in each area and place due emphasis onspending programs that are pro-poor (e.g., certain programsin health, education, and infrastructure) and on the effi-cient delivery of essential public services (e.g., public health,public education, social welfare, etc.). In examining theseexpenditures, policymakers should evaluate the extent towhich government intervention in general, and publicspending in particular, can be justified on grounds of mar-ket failure and/or redistribution.

Policymakers must also ask themselves whether the envis-aged public goods or services can be delivered efficiently(e.g., targeted at the intended beneficiaries) and, if not,

27


whether appropriate mechanisms and/or incentives can beput in place to ensure such efficient delivery. Countriesshould begin by assessing in a frank manner their adminis-trative capacity at both the national and subnational levelsto deliver well-targeted, essential public services in supportof poverty reduction. In this regard, policymakers shouldconsider the extent to which both technical assistance andthe private sector can play a role in improving the deliveryof these services.

In the context of medium-term budget planning, policy-makers should consider the scope for reallocating existinggovernment spending into priority areas23 and away fromnonproductive spending, including areas where a rationalefor public intervention does not exist. Operation and main-tenance expenditure tied to capital spending should also bereviewed with a critical eye. The quality of public expendi-ture could be assessed in the context of a public expendi-ture review with the assistance of multilateral and/orbilateral donors. Policymakers could then assess the newpoverty reduction projects and activities that have beenidentified in the context of the poverty reduction strategyand integrate them into the preliminary spending program.In so doing, they should attempt to rank the poverty pro-grams in order of relative importance in line with the coun-try's social and economic priorities, the marketfailure/redistribution criteria identified above, and thecountry's absorptive capacity in the light of existing institu-tional and administrative constraints. If spending cuts aredeemed necessary in the context of the integrated povertyreduction/macroeconomic framework, policymakers shouldrefer back to the ranking of the spending program based on

23 "Priority areas" are defined as those activities identified as crucial forpoverty reduction.

28


the relative importance and priority assigned to eachactivity.

A key aspect of any poverty reduction strategy will be anassessment of the impact of the present tax and nontax sys-tem on the poor. An important medium-term objective formany developing countries will be to raise domestic revenuelevels with a view to providing additional revenue in supportof their poverty reduction strategies.24 The existing revenuebase should be reviewed relative to its capacity to providefor the poverty spending requirements from nonbank do-mestic financing. Revenues should be raised in as economi-cally neutral a manner as possible, while taking intoconsideration equity concerns and administrative capacities(see Box 4).

In a developing country , taking account of allocationaleffects means that the tax system in particular should notattempt to affect savings and investment—experience indi-cates that aggregate savings and investment tend to be in-sensitive to taxes, with the result that the tax system typicallyonly affects the allocation of those aggregates across alterna-tive forms. As regards equity, the tax system should be as-sessed with respect to its direct and indirect impact on thepoor. It is difficult to have a tax system that is both efficientand progressive, particularly in those countries without awell-developed tax administration. Therefore, governmentsshould seek to determine a distribution of tax burdens seenas broadly fair rather than use the tax system to achieve adrastic income redistribution.

Tax policy should aim at moving toward a system of easilyadministered taxes with broad bases and moderate marginalrates. To the extent that some revenue provisions may be

24 For a discussion of tax policy and developing countries, see Tanzi andZee (2000).

29


Box 4. Tax Policy

The best tax systems typically include most or all of the follow-ing elements:

• A broad-based consumption tax, such as a VAT, preferably witha single rate, minimal exemptions, and a threshold to ex-clude smaller enterprises from taxation. The VAT generallyshould extend through the retail sector, and should applyequally to domestic production and imported goods andservices. The VAT should cover agricultural products andinputs, subject to the threshold, which will exclude smallfarmers.

• Excise taxes should apply to petroleum products, alcohol,and tobacco; should be collected at the point of productionor import; and should apply equally to domestic productionand imports.

• Taxes on international trade should play a minimal role. Im-port tariffs should have a low average rate and a limited dis-persion of rates, to reduce arbitrary and excessive rates ofprotection. Exemptions should be kept at a minimum andnontariff barriers should be avoided altogether. Exportersshould have duties rebated on imported inputs used forproducing exports and export duties should generally beavoided.

• The personal income tax should be characterized by only a fewbrackets and a moderate top marginal rate, by limited per-sonal exemptions and deductions, by a standard exemptionthat excludes persons with low incomes, and by extensiveuse of final withholding.

• The corporate income tax should be levied at one moderaterate. Depreciation allowances should be uniform across sec-tors, and there should be minimal use of tax incentivesother than permitting net operating losses to be carried for-ward for some reasonable period of time.

The use of a simplified regime for small businesses and the in-formal sector may complement these major taxes. Real propertytaxes may also be used if they can be administered appropri-ately, though this may be difficult in developing countries.

30


regressive, they should be offset through the expendituresystem (e.g., transitory, well-targeted food subsidies couldoffset the impact of a broad-based consumption tax andcushion the adverse impact of adjustment policies on thepoor). Finally, where revenue systems are being adminis-tered by a civil service that is highly constrained in terms ofhuman resources, technical support, and funding, countriesshould rely heavily on final withholding, and keep to the ab-solute minimum any exemptions, special provisions, or mul-tiple rates.

The scope for domestic budgetary financing will dependon a number of factors, including the sustainable rate ofmonetary growth, the credit requirements of the private sec-tor, the relative productivity of public investment, and thedesired target for net international reserves. Sacrificing lowinflation (through faster monetary growth) to finance addi-tional expenditure is generally not an effective means to re-duce poverty because the poor are most vulnerable to priceincreases. At the same time, since private sector develop-ment stands at the center of any poverty reduction strategy,governments need to take into account the extent to whichpublic sector borrowing "crowds out" the private sector's ac-cess to credit, thereby undermining the country's growthand inflation objectives. At times, public sector borrowingcan also "crowd in" private sector investment by putting inplace critical infrastructure necessary for private enterpriseto flourish. Given that at any point in time there is a finiteamount of credit available in an economy, policymakersmust therefore assess the relative productivity of public in-vestment versus private investment and determine theamount of domestic budgetary financing that would be con-sistent with the need to maintain low inflation and supportsustainable economic growth.

31


The amount and type of available external resources tofinance the budget will vary depending on the particular cir-cumstances facing the country. Countries that have access toexternal grants need to consider what amount is availableand sustainable under the present circumstances. The sameis true in the case of external debt, but policymakers alsoneed to determine whether the terms on such borrowingare appropriate and whether the added debt burden is sus-tainable. To the extent that a country is benefiting from, ormay benefit from, external debt relief under the enhancedHeavily Indebted Poor Countries (HIPC) Initiative, net re-source flows—flows that are predictable over the mediumterm—will be freed up to finance poverty-related budgetaryexpenditure. Domestic debt reduction could also representa viable use of additional concessional foreign assistance,since it would both free up government resources to be di-rected at priority poverty expenditure, as well as free up ad-ditional domestic credit for use by the private sector.

There may be a limit to the amount of additional externalfinancing that a country would deem to be appropriate,however. For example, there may be absorptive capacity con-straints that could drive up domestic wages and prices, aswell as appreciate the exchange rate and render the coun-try's exports less competitive, thereby threatening both sta-bility and growth. The extent of such pressures will dependon how much of the additional aid is spent on imports ver-sus domestic nontraded goods and services. There may alsobe uncertainty regarding aid flows, especially over themedium term, as well as considerations regarding long-termdependency on external official aid. In the absence ofmedium-term commitments of aid, policymakers may there-fore wish to be cautious in assuming what levels of assistancewould be forthcoming in the future.

32


Monetary and Exchange Rate Policies

Monetary and exchange rate policies can affect the poorprimarily through three channels: inflation, output, and thereal exchange rate. As mentioned above, inflation hurts thepoor because it acts as a regressive tax and curbs growth.Fluctuations in output clearly have a direct impact upon theincomes of the poor, and monetary and exchange rate poli-cies affect these fluctuations in two ways: first, changes inthe money supply can have a short-run effect on real vari-ables such as the real interest rate,25 which in turn affectoutput; and second, a country's chosen exchange rateregime can buffer, or amplify, exogenous shocks. Finally, thereal exchange rate can affect the poor in two ways.26 First, itinfluences a country's external competitiveness and henceits growth rate. Second, a change in the real exchange rate(through, for example, a devaluation of the nominal rate)can have a direct impact on the poor.27

25 The real interest rate represents the real cost of borrowing—that is,the cost in terms of goods—and is approximately equal to the nominal in-terest rate minus the expected rate of inflation.

26 The real exchange rate represents the relative price of a basket ofgoods in two countries. It is commonly measured by multiplying the nomi-nal exchange rate by the ratio of consumer price indices in the two coun-tries. If the real exchange rate appreciates, the basket of goods becomesmore expensive in the home country. This can happen if either the homecurrency appreciates, or if the home country's prices rise relative to thoseof the foreign country.

27 For example, as indicated earlier, recent studies have shown that insome countries, the income of the poor is more associated with tradablegoods and consumption with nontradable goods than the income and con-sumption patterns of other income groups. In these countries, this impliesthat a depreciation or devaluation of the domestic currency would makethe country's exports more attractive and stimulate demand for tradablegoods. Since the poor's incomes are tied to the production and export oftradables, this would, in turn, increase their income while the cost of theirconsumption of nontradables would remain unchanged.

33


Given that monetary and exchange rate policies affect thepoor through their impact on inflation, output, and the realexchange rate, it might seem, at first glance, that such poli-cies should therefore be used to target all three of thesevariables. However, although monetary and exchange ratepolicies may affect the poor through all of these channels,the monetary authorities cannot necessarily control the sizeand nature of the resulting impact. For example, changes inthe money supply may affect output and employment in theshort run, but they do so in a way that is at best uncertainand imperfectly understood. As a result, monetary authori-ties are typically unable to exploit this impact systematically.Similarly, monetary and exchange rate policies are unable tomanipulate the real exchange rate beyond a short period oftime. Therefore, actively using these policies to pursue aparticular short-run exchange rate goal, which may be in-consistent with underlying economic fundamentals, couldintroduce instability.

Monetary and exchange rate policies should target thosevariables over which they have the most control, namely thelong-run impact of inflation on the rate of growth. Broadlyspeaking, this can be achieved by setting one objective formonetary and exchange rate policies: the attainment andmaintenance of a low and stable rate of inflation. In prac-tice this means (1) choosing, and firmly committing to, aninflation rate target within the context of the overall povertyreduction strategy and the associated macroeconomic frame-work; (2) adopting the required policies to achieve the tar-get; and (3) not using monetary and exchange rate policiesto pursue, overtly or otherwise, additional or alternative ob-jectives. Formulated and implemented in this way, monetaryand exchange rate policies can form the basis for a stablemacroeconomic environment.

34


Improving Inflation Performance

In some cases, it may be desirable to target a lower rate ofinflation. What policies can help meet this objective? Ulti-mately, this question has to be answered on a case-by-casebasis. However, policymakers should consider two generalpolicies that are essential parts of any effort to improve in-flation performance: strong and sustained fiscal adjustment;and the use of a nominal anchor and other measures (e.g.,inflation targeting) to enhance policy credibility.

Fiscal Adjustment

A loose fiscal stance can put upward pressure on pricesthrough two channels: aggregate demand and financing.Such a fiscal stance increases the demand for domesticgoods, which, in the absence of a corresponding increase insupply, puts upward pressure on their prices. It can also in-crease demand for imports, putting downward pressure onthe value of the domestic currency and, hence, (in a flexibleexchange rate regime) upward pressure on the prices of im-ported goods. Further, if the fiscal stance is financed byprinting money, this expands the money supply and tends toincrease inflation.

In theory, if inflationary pressures from the fiscal stanceare being transmitted exclusively through the financingchannel, then inflationary pressures could be reduced with-out fiscal adjustment if alternative (sustainable) sources offinancing, such as external financing, are available. In prac-tice, however, some fiscal adjustment is typically also neces-sary because either the amount of alternative finance isinsufficient and/or the fiscal stance is also putting upwardpressure on prices through the aggregate demand channel.Indeed, evidence shows that successful disinflation episodeshave typically been accompanied by sizable and sustained

35


fiscal adjustment (Phillips, 1999). Therefore, countries thatwish to target a significantly lower rate of inflation need toensure that the corresponding fiscal adjustment is adequate.

Credibility and Nominal Anchors

Setting policy targets is important. Consistently achievingthose targets is equally important. When targets under a pol-icy are systematically missed, the policy loses credibility. If apolicy lacks credibility, the private sector does not believethat the authorities are truly committed to their policy tar-gets, and hence does not fully factor the authorities' targetsinto its inflation expectations, for instance when settingwage bargains. This can result in an inflation bias—that is,higher inflation outcomes brought on solely by the lack ofpolicy credibility itself.

Credibility can sometimes be enhanced by imposing re-strictions on policy (i.e., limiting the degree of discretion ofthe monetary authorities), or by adopting specific institu-tional arrangements. For example, the adoption of a fixedexchange rate regime involves a commitment to exchangedomestic currency for foreign currencies at a predefinedrate. This imposes an automatic discipline upon domesticmonetary policy. In effect, control over monetary policy issurrendered to the central bank of the country whose cur-rency has been chosen as the peg—typically a low inflationcountry—which, in turn, imparts credibility to the domesticpolicy objective of achieving low inflation.

More generally, evidence shows that inflation performancehas been better in countries using a nominal anchor(Phillips, 1999). Using a nominal anchor involves specifyingand committing to a predetermined path for a nominal vari-able—such as the exchange rate (i.e., the fixed exchangerate discussed above is a nominal anchor) or a money aggre-

36


gate—that is to a certain degree under the control of theauthorities.28 If the variable threatens to deviate from its tar-geted path the authorities take corrective action.29 In thisway, inflation, and inflationary expectations, can beanchored.

In some countries, fixed exchange rate regimes haveclearly been effective in establishing and maintaining lowinflation. More generally, there is empirical evidence that in-flation performance has been better in countries runningfixed exchange rate regimes (see, for example, Ghosh andothers, 1999). However, the choice of a fixed exchange ratehas to be based on broader considerations than simply itsmerits as a nominal anchor. In particular, the underlyingstructural features of an economy need to be supportive ofa fixed regime broadly speaking (for example, the degree ofprice rigidity, the nature of its predominant exogenousshocks, the degree of political support, etc.—these issues arediscussed below). Adopting a fixed exchange regime toserve only temporarily as a nominal anchor can be risky. Exit-ing a fixed regime once inflation performance is satisfactorycan be difficult. Moreover, if a country's economic condi-tions are not supportive, or political support for the policy

28 Other nominal variables can also serve as anchors. What is essential isthat the variable targeted be nominal, and not real, since real variablescannot provide an anchor for nominal prices. For example, countries thathave targeted the real exchange rate have generally had worse inflationperformance than other countries. See Phillips (1999).

29 The two most commonly used nominal anchors are a fixed exchangerate and a money aggregate (such as reserve money or broad money).Under a fixed exchange rate regime, whenever the market rate threatensto depart from the predetermined rate, the monetary authorities buy orsell foreign exchange for the domestic currency to ensure that the ex-change rate remains fixed. Similarly, under a monetary anchor the mone-tary authorities specify a predetermined path for a monetary aggregate,and tighten or loosen the monetary stance when the aggregate threatensto depart from that path.

37


insufficient, the peg could come under considerable pres-sure, which may, in the end, force a costly abandonment ofthe regime and undermine the original objective of stabiliz-ing inflation.

Both types of nominal anchors restrict the use of mone-tary instruments.30 A standard critique has been that, al-though the use of a nominal anchor may improve inflationperformance, it comes at the cost of reducing the discretionof the authorities to respond to short-run shocks. In prac-tice this trade-off may not be significant, however. Even ifthe monetary authorities have full discretion,31 as discussedabove, their ability to influence short-run output movementssystematically is limited. Moreover, their ability to exercisediscretion is likely to be limited by the need to preserve, orenhance, policy credibility.

Inflation targeting has been adopted as the monetaryregime in an increasing number of industrialized and devel-oping countries in recent years. It is typically and preferablyassociated with a flexible exchange rate system. Inflation tar-geting sets an inflation target for the central bank and givesthe responsibility for achieving the target to the centralbank. To enhance accountability, credibility, and efficiency,the central bank in an inflation targeting regime is gener-ally required to be extremely transparent about its opera-tions, explaining its decisions to the public, publishing, inmost cases, a regular inflation report.

In the long run, however, only policies to which the au-thorities are fully committed can be credible. Imposing re-

30 Under a fixed exchange rate, the monetary authorities give up con-trol of the money supply. Under a monetary anchor, the authorities can-not pursue an exchange rate target.

31 If there are no explicit policy targets, the monetary authorities havefull discretion. This differs from the concept of independence of the mon-etary authorities.

38


strictions on policy when the necessary policy commitmentis absent (or even when the private sector erroneously sus-pects a lack of commitment) can have disastrous results. Forexample, the private sector's belief that a country's authori-ties are not committed to defending its fixed exchange ratemay lead to a speculative attack on the peg. Although de-vices may be used to accelerate the attainment of a policy'scredibility, there is no substitute for commitment to the pol-icy, as demonstrated through sustained adherence to a pru-dent macroeconomic stance.

External Shocks and the Choice of Exchange Rate Regime

The choice of exchange rate regime—fixed or flexible—depends crucially on the nature of the economic shocks thataffect the economy, as well as the structural features of theeconomy, which may either mitigate or amplify these shocks.Choosing a fixed exchange rate regime when these underly-ing features of the economy are not supportive leaves acountry more exposed to the possibility of an external crisis,which can result in the ultimate abandonment of the peg.In addition, shocks to output can have a strong impact onthe poor. Since different exchange rate regimes have differ-ent insulating properties vis-a-vis certain types of shocks,choosing the regime that best insulates the economy willserve to moderate fluctuations in output, and thereby bestserve the poor.

For example, if the predominant source of disturbance toan economy is shocks to the terms of trade, a flexibleexchange rate regime may be best because the nominalexchange rate is free to adjust in response to the shock andbring the real exchange rate to its new equilibrium (see, forexample, Devarajan and Rodrik, 1992). Alternatively, if do-mestic monetary shocks predominate, such as shocks to the

39


demand for money, output may be best insulated by a fixedexchange rate that allows these shocks to be absorbed byfluctuations in international reserves. Of course, one of thechallenges facing the policymaker is to identify which shocksare in fact predominant in a particular economy.

The structural features of the economy may also affect theimpact a particular shock has on the economy, as well as theinsulating properties of exchange rate regimes. For exam-ple, if an economy is characterized by a significant degree ofnominal wage rigidity, wages will not fully adjust (at least inthe short run) in response to small real shocks, and hencethe effect of those shocks on output will be amplified. Inthese circumstances, even if domestic monetary shocks areimportant, a flexible exchange rate regime may well bepreferable (in contrast to the conclusions above). Anotherimportant structural feature is the degree of an economy'sopenness. Typically the more open an economy is, thegreater is its exposure to external shocks. This would arguegenerally in favor of a flexible exchange rate regime. How-ever, if an open economy is sufficiently diversified (i.e., ittrades a wide range of goods and services) and if its pricesare sufficiently flexible, then a fixed exchange rate may bepreferable because the volatility of flexible exchange ratesmay impede international trade, and thus lower external de-mand (although the evidence on this is mixed). In conclu-sion, these various pros and cons of fixed versus flexibleexchange rate regimes need to be carefully assessed andweighed on a case-by-case basis—again, there is no universal"right answer."

Policies to Insulate the Poor Against Shocks

Given that the poor are adversely affected by macroeco-nomic shocks, what should governments do about it? The

40


question can be divided into two parts: How should eco-nomic policy be designed to cushion the impact of shockson the poor, in particular during times of crisis and/or ad-justment? What specific policies can governments undertaketo insulate the poor from the consequences of shocks by re-moving existing distortive policies?

Social Safety Nets

Sound macroeconomic policies will help a country to re-duce its exposure to macroeconomic shocks, but there is nocost-effective policy that will insure against all possibleshocks. It is therefore crucial to have social safety nets inplace to ensure that poor households are able to maintainminimum consumption levels and access to basic social ser-vices during periods of crisis. Social safety net measures arealso necessary to protect the poor from shocks imposed onthem during periods of economic reform and adjustment.32

Safety nets include public work programs, limited food sub-sidies, transfers to compensate for income loss, social funds,fee waivers, and scholarships for essential services such aseducation and health. The specific mix of measures will de-pend on the particular characteristics of the poor and theirvulnerability to shocks and should be well-targeted and de-signed in most cases to provide temporary support.

32 Reform programs should be designed with the poor and vulnerablein mind. The mix and sequencing of reform measures should be designedto minimize the hardships brought about by the program. The appropri-ate mix and sequencing cannot, however, ensure that the adverse effectswill be removed entirely and, hence, social safety nets are needed to miti-gate possible short-run adverse effects on the poor. In some cases, it maybe appropriate to delay reforms until adequate safety net measures can beput in place. See Chu and Gupta (1998).

41


Equally important, the resources allocated to social safetynets should be protected during economic crises and/or ad-justment, when fiscal tightening may be necessary. Govern-ments should have budgetary guidelines approved by theirlegislatures that prioritize and protect poverty-related pro-grams during periods of crisis and provide a clear course ofaction that ensures access of the poor to basic social servicesduring periods of austerity (see Lustig, forthcoming). As willbe discussed below, countercyclical fiscal policies can alsoensure the availability of funds for financing safety nets dur-ing crises.

Another important factor to consider is that safety netsshould already be operating before economies get hit byshocks so that they can be effective in times of distress (fora more detailed account, see World Bank, 2000). However, ifa shock occurs before appropriate safety nets have been de-veloped, then "second-best" social protection policies may benecessary. For instance, food subsidies have been found tobe inefficient and often benefiting the non-poor, and mostreform programs call for their reduction or even elimina-tion. However, after a severe shock such as the 1997-98 EastAsian financial crisis, when countries like Indonesia lackedcomprehensive safety nets, existing food subsidies wereprobably the only means of preventing widespread malnutri-tion and starvation. In the context of a country's reformprocess, however, these subsidies should be replaced withbetter targeted and less distorting transfers to the poor.

Removing Market Distortions and Distortive Policies

In addition to pursuing favorable economic policies andputting in place appropriate social safety nets, there are spe-cific structural reforms that governments can undertake toinsulate the poor from the adverse consequences of shocks.

42


Most of these have to do with addressing the mechanismsthrough which macroeconomic shocks are transmitted tothe poor. (see Box 5).

To the extent that asset market distortions prevent thepoor from saving and insulating themselves against shocks,policies to remove these distortions can be valuable.33 Forinstance, foreign exchange controls can force the poor tohold their assets in domestic currency, whose value typicallydeclines with adverse shocks. Relaxing these controls in awell-managed fashion could give the poor access to safer as-sets, such as foreign currency, that could protect them fromdevaluations, a typical outcome following negative shocks.34

Similarly, severe financial repression, such as controlled in-terest rates, can impede the poor's ability to save.35 If prop-erly managed, financial liberalization policies can thereforehave the additional benefit of increasing self-insurance forthe poor.

Policies that increase borrower information and relax bar-riers to access to credit markets can help the poor reduceconsumption volatility, since credit availability makes themless dependent on current income. Also, to the extent thatcollateralized credit allocation amplifies the effects of nega-

33 Contrary to what some may believe, the poor do save, to smooth con-sumption over time, as well as to guard against adverse shocks. For a re-cent analysis, see Deaton and Paxson (2000).

34 Also, capital controls that drive a wedge between domestic and worldreal interest rates make it possible to extract an inflation tax, which espe-cially hurts the poor.

35 For many countries, domestic asset holdings of the poor are mainlycomposed of currency, so it would seem that this channel is not relevant.But this may just reflect that low controlled interest rates provide a disin-centive to save in bank deposits. Removing financial distortions could shiftthe allocation of domestic assets in favor of deposits and, to the extentthat market interest rates account for expected inflation, insulate thepoor's savings from inflation.

43


Box 5. How Shocks Harm the Poor: Transmission ChannelsCredit markets, as well as safe asset markets for appropriate

saving, are major instruments for coping with income volatility.Distortions in these markets curtail the ability of the poor to fol-low consumption smoothing patterns. Government behavior inresponse to shocks is also a major determinant of the effects ofthese shocks on the poor. Financial sector behavior can also am-plify the effects of shocks.

• Distortions in asset markets. To self-insure against shocks,agents need to be able to save in assets whose value does notfall when they are needed to compensate for a fall in in-come. Such saving instruments are typically composed of for-eign assets, domestic financial assets, and domestic realassets. To the extent that governments impose controls onthese asset markets, it impedes the ability of the poor to usethese savings instruments, and channels their savings intoless effective instruments.1

• Access to and structure of credit markets. Access to credit marketsis extremely limited for the poor to buffer the effects ofshocks, in part as a consequence of inadequate borrower in-formation available to credit institutions. The structure ofcredit markets can also affect the poor indirectly: Firms mayfind that access to credit is typically collateralized. There-fore, shocks that drastically reduce the value of collateral willalso reduce firms' access to credit, amplifying the effect ofshocks when firms become credit constrained. Severe down-turns may dramatically cut employment of the poor and,

1For example, in Ethiopia, livestock prices (often the poor's only asset)fall during a drought because all farmers are selling their cattle to compen-sate for the bad harvest.

tive shocks by reducing small- and medium-sized firms' ac-cess to credit when asset prices fall (Kiyotaki and Moore,1977, and Izquierdo, 1999), policies promoting better finan-cial-sector credit allocation mechanisms based on projectprofitability and borrower information could reduce the inci-

44


hence, their welfare, especially since the poor are usuallyemployed by small firms that depend on collateralizedcredit.Procyclical fiscal policy. During booms, some governments havedecreased export commodity taxes, reduced revenue collec-tion effort, or heavily increased expenditures, amplifying theeffects of positive shocks. These policies can make adjust-ment more severe during busts, since the resulting expendi-ture cuts following a negative shock have sometimes fallenon social programs and transfers to the poor—-just whenthey are needed the most. Examples of these policies can betraced back to the experience of Cote d'Ivoire in 1976—79,and Colombia in 1975-1980. It has also been argued thatsynchronizing tax policies with shocks tends to obscure theinformation content of prices on which agents make theirsaving decisions, leading in many cases to lower saving ratesthan needed to buffer future adverse shocks (Collier, Gun-ning, and associates 1999).Financial sector vulnerability and transmission to other sectors.Shocks can also be amplified through the banking system.For example, a negative terms of trade shock can adverselyaffect bank liquidity by reducing demand for domestic de-posits, forcing banks to curtail credit roll-over, spreading theshock throughout the economy (IADB, 1995, and Haus-mann, 1999). Similarly, if a sudden stop in capital flows ren-ders many nontradable goods producers bankrupt becauseof big swings in relative prices, this may in turn create finan-cial turmoil as loans become nonperforming, spreading theeffect of the shock across the financial system (Calvo, 1998).Bankruptcies in the nontradable sector may translate intounemployment of the urban poor.

dence of this particular transmission channel and its indirecteffects on the poor (i.e., lower employment opportunities).36

36 Collateralization may be initially the only way for small firms to gainaccess to credit markets, but its amplification effects should not be under-stated. Instead, policies that reduce informational problems (i.e., the rea-son for collateralization) should be implemented.

45


Finally, and most important, governments can do a lot toreduce the pro-cyclical nature of their fiscal policies by sav-ing rather than spending windfalls following positive shocksand ideally using those savings as a buffer for expendituresagainst negative shocks. A cautious approach would be forthe government to "treat every favorable shock as temporaryand every adverse one as permanent," although judgmentwould also depend on, among other things, the availabilityof financing (Little, and others, 1993). However, even thisrule of thumb may not be enough. Governments need tofind ways of "tying their hands" to resist the pressure tospend windfall revenues (Devarajan, 1999). For example,when the source of revenue is publicly owned, such as oil orother natural resource, it may be appropriate to save thewindfall revenues abroad, with strict rules on how much ofit can be repatriated. Countries such as Colombia, Chile,and Botswana have tried variants of this strategy, with bene-fits not just for overall macroeconomic management, butalso for protecting the poor during adverse shocks, sincesaved funds during good times can be applied to financingof safety nets during crisis.

46


Tables

The following three tables show macroeconomic data,such as GDP growth, for a range of developing countries.The tables reveal that many developing countries are in astate of macroeconomic stability.

47


Table 1. Real GDP Growth(Annual percentage change)

High-growth countriesMozambiqueSudanVietnamUgandaBhutanIndiaTogoMyanmarLao PDRAngolaEritreaArmeniaEthiopiaCote d'IvoireBeninBangladeshSenegalRwandaNicaraguaNepal

1994

7.54.08.86.48.17.9

16.87.58.21.49.85.43.52.04.43.82.9

-50.23.38.2

1995

4.325.2

9.511.5

6.88.06.86.97.0

11.32.96.96.17.04.65.54.7

34.44.33.5

1996

7.14.09.39.15.57.39.76.46.8

11.76.85.9

10.96.95.55.05.2

15.84.75.3

1997

11.36.78.14.77.85.04.35.77.06.67.93.35.95.95.75.35.0

12.85.15.0

1998

12.05.05.85.67.16.1

-1.05.04.05.03.97.2

-1.04.54.55.15.79.54.02.3

1999

9.04.04.27.87.06.23.0

...4.0

-0.13.05.57.04.35.04.35.15.96.33.4

1994-99Average

8.58.27.67.57.16.86.66.36.26.05.75.75.45.15.04.84.84.74.64.6

48


Burkina FasoMalawiChadGuineaMaliMauritaniaGhanaCambodiaNigerLesothoZimbabwePakistanYemen, Rep. ofCentral African RepublicMongoliaCameroonTanzaniaGeorgiaGambia, TheKenyaMadagascarIndonesiaNigeriaSolomon IslandsSao Tome and Principe

1.2-10.2

10.24.00.94.63.34.04.03.46.83.9

-3.64.92.3

-2.51.4

-11.40.22.60.07.50.15.22.2

4.015.41.04.46.24.64.07.62.64.5

-0.55.17.97.26.33.32.62.40.94.41.78.22.57.02.0

6.09.03.74.63.25.54.67.03.4

10.08.75.02.9

-4.02.45.04.3

10.52.24.12.17.84.33.51.5

4.74.94.14.86.83.24.21.03.38.03.71.28.15.24.05.14.0

11.04.92.13.74.72.7

-0.51.0

6.23.18.14.43.33.24.71.08.3

-5.02.53.34.84.73.55.03.52.94.71.83.9

-13.21.8

-7.02.5

5.24.5

-1.03.75.34.34.44.52.32.51.23.92.23.42.74.44.32.04.21.64.70.21.04.02.5

4.64.54.44.34.34.24.24.24.03.93.73.73.73.63.53.43.42.92.92.82.72.52.12.02.0

49


Table 1 (concluded)

Low-growth countries

Congo, Rep. ofHaitiZambiaUzbekistanGuinea-BissauKyrgyz RepublicCongo, Dem. Rep. ofComorosAzerbaijanBurundiSierra LeoneTajikistanTurkmenistanUkraineMoldova

AfghanistanKorea, Dem. People's Rep. ofLiberiaSomalia

1994

-5.5-8.3-3.4-5.2

3.2-20.1-3.9-5.3

-24.3-3.9

3.5-19.0-17.3-22.9-31.2

. . .

. . .

. . .

. . .

1995

4.04.4

-2.3-0.9

4.4-5.4

0.7-3.9

-11.8-7.3

-26.6-11.8

-7.2-12.2

-1.4. . .

. . .

. . .

. . .

1996

6.32.76.51.74.67.1

-0.9-0.4

1.3-8.428.7-4.4-6.7

-10.0-7.8

. . .

. . .

. . .

. . .

1997

-2.41.43.42.55.49.9

-5.70.05.80.4

-17.62.3

-11.3-3.0

1.3. . .

. . .

. . .

. . .

1998

3.63.1

-2.04.4

-28.11.83.00.0

10.04.8

-0.88.25.0

-1.7-8.6

. . .

. . .

. . .

. . .

1999

-0.7...

1.3...

8.7......

1.010.0-1.0-8.1

...

...-2.0-5.0

. . .

. . .

. . .

. . .

1994-99Average

0.90.70.60.5

-0.3-1.3-1.4-1.4-1.5-2.6-3.5-4.9-7.5-8.6-8.8

. . .

. . .

. . .

. . .

Source: Country authorities.

50


Table 2. GDP Deflator(Annual percentage change)

Low-inflation countries

Gambia, TheGuineaBangladeshComorosEthiopiaMauritaniaCameroonUgandaCentral African RepublicSenegalNepalIndiaLesothoBurkina FasoNigerEritreaCambodiaBhutanMali

1994

3.82.93.49.42.66.4

11.06.8

22.827.87.49.77.5

27.832.722.28.99.3

27.9

1995

4.05.56.78.0

12.74.4

17.09.4

10.35.96.38.68.99.85.4

11.39.19.8

18.4

1996

2.92.83.82.31.04.65.44.61.80.97.87.98.64.24.72.97.1

11.45.4

1997

5.02.41.03.53.25.52.73.90.82.37.35.68.02.23.12.79.2

14.71.0

1998

1.35.05.33.09.79.91.1

10.71.72.23.38.98.43.13.02.7

17.05.94.0

1999

5.24.08.93.01.91.9

-1.23.01.31.69.25.57.21.63.0

12.65.77.82.2

1994-99Average

3.73.84.94.95.25.56.06.46.56.86.97.78.18.18.79.19.59.89.8

51


Table 2 (concluded)

1994 1995 1996 1997 1998 19991994-99Average

Solomon IslandsPakistanCote d'IvoireNicaraguaTogoVietnamBeninChadCongo, Rep. ofBurundiKenyaRwanda

High-inflation countriesTanzaniaIndonesiaMadagascarHaitiNigeriaSierra LeoneMyanmarYemen, Rep. of

11.512.941.7

7.833.814.533.543.436.8

6.735.218.0

28.27.8

41.635.627.820.122.133.2

7.113.89.6

10.912.219.515.48.6

22.615.311.251.3

28.99.9

45.231.056.063.719.454.8

12.18.42.7

11.65.16.16.7

11.6-3.019.0

8.710.5

22.38.7

17.821.236.9

3.023.039.6

8.113.33.29.2

11.412.14.72.77.5

23.315.515.6

18.512.6

7.316.31.4

16.832.912.9

12.07.83.1

12.92.78.94.24.1

-19.112.110.62.6

19.373.1

8.412.721.627.034.0-9.3

. . .

5.52.7

13.91.15.83.5

-2.124.1

0.65.0

-2.4

11.617.29.8

11.925.0

26.8

10.210.310.511.111.111.211.311.411.512.814.415.9

21.521.621.723.425.925.926.326.3

52


GhanaGuinea-BissauMozambiqueMongoliaZambiaMalawiLao PDRKyrgyz RepublicSao Tome and PrincipeSudanMoldovaTajikistanUkraineAzerbaijanUzbekistanTurkmenistanArmeniaGeorgiaAngolaCongo, Dem. Rep. of

AfghanistanKorea, Dem. People's Rep. ofLiberiaSomalia

30.123.259.466.656.626.27.7

180.973.580.8

278.1236.2953.5

1,428.61,238.61,022.14,107.39,349.22,170.7

26,762.4

. . .

. . .

. . .

. . .

43.244.752.042.536.990.319.742.074.541.138.7

285.0415.5545.7370.9705.7161.2162.7

1,886.1466.4

. . .

. . .

. . .

. . .

39.848.640.933.524.340.413.935.350.8

107.430.5

397.966.226.481.6

1,174.319.640.2

5,421.8613.1

. . .

. . .

. . .

. . .

19.535.511.124.425.913.517.719.3

100.264.912.9

100.218.19.2

70.561.617.7

7.093.6

187.3

. . .

. . .

. . .

. . .

17.07.63.8

11.523.223.284.0

8.537.128.9

8.049.913.2-8.333.213.511.23.4

60.915.0

. . .

. . .

. . .

. . .

11.53.15.22.8

25.646.4

125.4. . .

16.0. . .

37.3. . .. . .

4.8. . .. . .

1.115.0

412.6. . .

. . .

. . .

. . .

. . .

26.927.128.730.232.140.044.757.258.764.667.6

213.8293.3334.4359.0595.4719.7

1,596.31,674.35,608.8

. . .

. . .

. . .

. . .

Source: Country authorities.

53


Table 3. Primary Surplus1

(As a percentage of GDP)

Surplus/low deficit countries

MauritaniaKenyaLesothoCongo, Rep.CameroonZambiaCote d'IvoireSao Tome and PrincipeNigeriaSenegalTanzaniaZimbabweBeninBhutanMyanmarIndonesiaVietnamGambia, TheMadagascar

1994

2.48.38.1

-1.1-2.25.21.4

. . .0.51.60.60.00.8

-0.50.01.1

. . .2.6

-3.0

1995

6.57.36.35.93.23.83.2

. . .8.12.7

-0.5-0.7-0.5

0.10.01.4

-0.6-2.9-1.1

1996

11.46.26.57.05.02.13.8

-0.37.52.11.20.72.12.00.01.1

-0.2-5.3-0.2

1997

8.26.22.14.45.43.53.0

11.33.32.24.7

-0.42.0

-2.10.0

-0.5-0.8-1.4

0.7

1998

8.45.8

-4.52.24.10.12.6

-2.8-6.9

1.72.28.03.33.00.0

-3.3-1.6

2.9-1.8

1994-98Average

7.46.83.73.73.13.02.82.72.52.11.61.51.50.50.00.0

-0.8-0.8-1.1

54


UgandaNigerMaliGuineaBurkina FasoMozambiqueEthiopiaMalawiCentral African RepublicGhanaRwandaTogo

High deficit countriesChadSolomon IslandsBurundiBangladeshHaitiSierra LeonePakistanLao PDRNepalAngolaIndia

-2.3-4.5-2.0-2.0-1.8-4.2-4.3

-16.2-7.6-4.4-7.0-6.2

-3.5-5.5-2.8-4.6-1.8-4.5-6.0

. . .-7.0-8.1-8.8

-1.9-1.5-1.7-1.3-2.2-1.6-1.4

1.0-2.8-2.1

0.0-3.0

-3.4-5.3-3.1-5.3-8.3-7.8-5.9-4.5-6.6

-16.5-7.9

-0.91.40.2

-1.70.3

-1.5-3.2

2.3-1.5-4.4-4.1-3.0

-4.0-4.4-8.3-4.4-7.6-4.4-7.0-5.9-7.5

1.1-7.8

-0.9-1.1-1.1-1.3-2.3-1.3

0.7-5.2-0.9-2.7-1.3

0.2

-2.9-4.9-3.4-4.3-3.1-5.0-6.4-8.1-7.3

. . .

-8.5

0.4-0.4-1.6-1.8-2.1-1.4-2.1

7.00.60.1

-1.9-3.0

-4.10.1

-2.7-4.2-4.0-5.1-5.5-7.2-7.8

. . .

-9.4

-1.1-1.2-1.2-1.6-1.6-2.0-2.1-2.2-2.5-2.7-2.9-3.0

-3.6-4.0-4.1-4.5-5.0-5.4-6.2-6.4-7.2-7.8-8.5

55


Table 3 (concluded)1994-98

1994 1995 1996 1997 1998 Average

NicaraguaMongoliaEritreaAfghanistanCambodiaComorosCongo, Dem. Rep. of

Guinea-BissauKorea, Dem. People's Rep. ofLiberiaSomaliaSudan

Source: Data provided by the authorities.1Negative sign indicates a primary deficit.

56

-11.4-22.8

. . .

. . .

. . .

. . .

. . .

. . .

. . .

. . .

. . .

. . .

-11.0-4.1

. . .

. . .

. . .

. . .

. . .

. . .

. . .

. . .

. . .

. . .

-10.2-8.2

. . .

. . .

. . .

. . .

. . .

. . .

. . .

. . .

. . .

. . .

-7.6-8.6

. . .

. . .

. . .

. . .

. . .

. . .

. . .

. . .

. . .

. . .

-5.0-11.3-27.3

. . .

. . .

. . .

. . .

. . .

. . .

. . .

. . .

. . .

-9.0-11.0-27.3

. . .

. . .

. . .

. . .

. . .

. . .

. . .

. . .

. . .


References

Agenor, Pierre-Richard, Shantayanan Devarajan, William Easterly,Hippolyte Fofack, Delfin Go, Alejandro Izquierdo, Lodovico Piz-zati, 2000, "A Macroeconomic Framework for Poverty ReductionStrategies," Development Research Group and World Bank In-stitute (unpublished; Washington: World Bank).

Alesina, Alberto, and Dani Rodrik, 1994, "Distributive Politics andEconomic Growth," Quarterly Journal of Economics, Vol . 109(May), pp. 465-90.

Balassa, Bela, 1981, "The Newly Industrializing Developing Coun-tries after the Oil Crisis," Weltwirtschaftliches Archiv, Vol. 117,No.1, pp. 142-94.

Behrman, Jere, Suzanne Duryea, and Miguel Szeleky, 1999, "School-ing Investments and Macroeconomic Conditions: A Micro-MacroInvestigation for Latin America and the Caribbean" (unpub-lished; Washington: Inter-American Development Bank).

Benabou, Roland, 1996, "Inequality and Growth," in NBER Macroeco-nomics Annual: Volume II, ed. by Ben Bernanke and Julio Rotem-berg (Cambridge, Mass.: MIT Press).

Birdsall, Nancy, and Juan Luis Londorio, 1997, "Asset InequalityMatters: An Assessment of the World Bank's Approach to PovertyReduction," American Economic Review, Vol. 87(May), pp. 32-37.

Bourguignon, Francois, and Christian Morrisson, 1998, "Inequalityand Development: The Role of Dualism," Journal of DevelopmentEconomics, Vol. 57 (December), pp. 233-57.

Bourguignon, Francois, William H . Branson, and Jaime de Melo,1989, Macroeconomic Adjustment and Income Distribution: A Macro-Micro Simulation Model (Paris: OECD Development Centre).

Bruno, Michael, and William Easterly, 1998, "Inflation Crises andLong-Run Growth," Journal of Monetary Economics, Vol. 41 (Febru-ary), pp. 3-26.

Calvo, Guillermo, 1998, "Capital Flows and Capital-Market Crises:The Simple Economics of Sudden Stops," Journal of Applied Eco-nomics, Vol. 1 (November), pp. 35-54.

57


Chu, Ke-young, and Sanjeev Gupta, eds., 1998, Social Safety Nets: Is-sues and Recent Experiences (Washington: International MonetaryFund).

Collier, Paul, and Jan Willem Gunning, 1999, "Explaining AfricanEconomic Performance," Journal of Economic Literature, Vol. 37(March), pp. 64-111.

_______, and associates, 1999, Trade Shocks in Developing Countries(Oxford: Oxford University Press).

Datt, Gaurav, and Martin Ravallion, 1992, "Growth and Redistribu-tion Components of Changes in Poverty Measures: A Decompo-sition with Applications to Brazil and India in the 1980s,"Journal of Development Economics, V o l . 38 (Apri l) , pp. 275—95.

_______, 1998, "Farm Productivity and Rural Poverty in India," Jour-nal of Development Studies, V o l . 34 (Apri l ) , pp. 62-85.

Deaton, A., and C. Paxson, 2000, "Growth and Saving Among Indi-viduals and Households," Review of Economics and Statistics, V o l . 82(May), pp. 212-25.

Deininger, Klaus, 1999, "Asset Distribution, Inequality, andGrowth," Development Research Group (Washington: WorldBank).

, and Lyn Squire, 1998, "New Ways of Looking at Old Is-sues: Inequality and Growth," Journal of Development EconomicsVol. 57 (December), pp. 259-87.

Devarajan, Shantayanan, 1999, "Cameroon," in Trade Shocks in De-veloping Countries, ed. by Paul Collier and Jan Gunning (Oxford:Oxford University Press).

, William R. Easterly, and Howard Pack, forthcoming "IsInvestment in Africa Too Low or Too High?", Journal of AfricanEconomies.

Devarajan, Shantayanan, and Dani Rodrik, 1992, "Do the Benefitsof Fixed Exchange Rates Outweigh Their Costs? The CFA Zonei n Africa," i n Open Economies: Structural Adjustment and Agricul-ture, ed. by Ian Goldin and L. Alan Winters (Cambridge, NewYork, and Melbourne: Cambridge University Press).

Devarajan, Shantayanan, Vinaya Swaroop, and Heng-fu Zou, 1997,"The Composition of Public Expenditure and Economic Perfor-mance," Journal of Monetary Economics, V o l . 37 (Apri l) ,pp. 313-44.

Dollar, David, and Roberta Gatti, 1999, "Gender Inequality, Incomeand Growth: Are Good Times Good for Women?" Policy Re-

58


search Report on Gender and Development Working Paper Se-ries No. 1. (Washington: World Bank).

Dollar, David, and Aart Kraay, 2000, "Growth Is Good for the Poor,"World Bank Development Research Group (unpublished; Wash-ington, D.C., World Bank).

Easterly, William, and Aart Kraay, 1999, "Small States, Small Prob-lems?" Policy Research Working Paper No. 2139, DevelopmentResearch Group (Washington: World Bank).

Easterly, William, and Sergio Rebelo, 1993, "Fiscal Policy and Eco-nomic Growth: An Empirical Investigation," Journal of MonetaryEconomics, Vol. 32 (December), pp. 417-58.

Fallon, Peter, and Vivian Hon, 1999, "Poverty and Labor-intensiveGrowth." Note prepared for World Development Report 2000/2001(unpublished; Washington: World Bank).

Fischer, Stanley, 1993, "The Role of Macroeconomic Factors inGrowth," Journal of Monetary Economics, V o l . 32 (December),pp. 485-512.

Forbes, Kristin, 2000, "A Reassessment of the Relationship BetweenInequality and Growth," American Economic Review, Vol. 90(September), pp. 869-887.

Ghosh, Atish, Anne-Marie Gulde, Jonathan Ostry, and Holger Wolf,1999, "Does the Nominal Exchange Rate Regime Matter?" (un-published; Washington: International Monetary Fund).

Ghosh, Atish, and Steven Phillips, 1998, "Warning: Inflation May BeHarmful to Your Growth," IMF Staff Papers, International Mone-tary Fund, Vol. 45 (December), pp. 672-710.

Hausmann, Ricardo, 1999, "Managing Terms of Trade Volatility,"World Bank PREM Note No. 18, February (Washington: WorldBank).

Inter-American Development Bank (IADB), 1995 "OvercomingVolatility," Economic and Social Progress in Latin America (Bal-timore: Johns Hopkins University Press).

Izquierdo, Alejandro, 1999, "Credit Constraints and the Asymmet-ric Behavior of Asset Prices and Output under External Shocks,"(Doctoral Dissertation, University of Maryland).

Kakwani, Nanak, 1993, "Poverty and Economic Growth with Appli-cation to Cote d'Ivoire," Review of Income and Wealth, Vol. 39(June) pp. 121-139.

Kiyotaki, Nobuhiro, and John Moore, 1997, "Credit Cycles," Journalof Political Economy, Vol. 105 (April), pp. 211-48.

59


Klasen, Stephan, 1999, "Does Gender Inequality Reduce Growthand Development? Evidence from Cross-Country Regressions,"Policy Research Report on Gender and Development WorkingPaper Series No. 7. (Washington: World Bank).

Li , Hongyi, Danyang Xie, Heng-fu Zou, 1999. "Dynamics of IncomeDistribution," Development Research Group, (unpublished;Washington: World Bank).

Little, I., R. Cooper, W. M. Corden, and S. Rajapatirana, 1993,Boom, Crisis and Adjustment: The Macroeconomic Experience of Develop-ing Countries (Oxford: Oxford University Press).

Lustig, Nora, forthcoming. "Crises and the Poor: Socially Responsi-ble Macroeconomics." Economia, Journal of the Latin Americanand Caribbean Economic Association.

Malmberg Calvo, Christina, 1998, Options for Managing and Fi-nancing Rural Transport Infrastructure, World Bank TechnicalPaper No. 411 (Washington: World Bank).

Masson, Paul, Miguel Savastano, and Sunil Sharma, 1997, "TheScope for Inflation Targeting in Developing Countries," IMFWorking Paper No. 97/130 (Washington: International Mone-tary Fund).

Mitra , Pradeep, 1994, Adjustment in Oil-Importing Developing Countries(Cambridge: Cambridge University Press).

Perotti, Roberto, 1992, "Income Distribution: Politics and Growth,"American Economic Review, V o l . 82 (May), pp. 311-16.

_______, 1993, "Political Equilibrium, Income Distribution, andGrowth." Review of Economic Studies, Vol. 60 (October),pp 755-76.

_______, 1996, "Redistribution and Non-consumption Smoothing inan Open Economy," Review of Economic Studies, Vol. 63 (July),pp 411-33.

Persson, Torsten, and Guido Tabellini, 1994, "Is Inequality Harmfulfor Growth?" American Economic Review, Vol. 84 (June),pp. 600-21.

Phillips, Steven, 1999, "Inflation: The Case for a More Resolute Ap-proach" in Economic Adjustment and Reform in Low-Income Coun-tries: Studies by the Staff of the International Monetary Fund, " ed. byHugh Bredenkamp and Susan Schadler (Washington: Interna-tional Monetary Fund).

60


Ramey, Garey, and Valerie A. Ramey, 1995, "Cross-Country Evi-dence on the Link between Volatility and Growth," American Eco-nomic Review, Vol. 85 (December), pp. 1138—51.

Ravallion, Martin, 1997, "Can High-Inequality Developing CountriesEscape Absolute Poverty?" Policy Research Working PaperNo. 1775 (Washington: World Bank).

Sahn, David, Paul Dorosh, and Stephen Younger, 1997, StructuralAdjustment Reconsidered: Economic Policy and Poverty in Africa, (NewYork: Cambridge University Press).

Sarel, Michael, 1996, "Nonlinear Effects of Inflation on EconomicGrowth," Staff Papers, International Monetary Fund, Vol. 43(March), pp. 199-215.

Tanzi, Vito, and Howell Zee, 2000, "Tax Policy for Emerging Mar-kets: Developing Countries," IMF Working Paper No. 00/35(Washington: International Monetary Fund).

Thomas, Vinod, and Yan Wang, 1998, "Missing Lessons of East Asia:Openness, Education, and the Environment," Latin America andCaribbean Studies: Proceedings series (Washington: WorldBank).

Thorbecke, Erik, and Hong-Sang Jung, 1996, "A Multiplier Decom-position Method to Analyze Poverty Alleviation," Journal of Devel-opment Economics, Vol. 48 (March), pp. 279-300.

Timmer, C. Peter, 1997, "How Well Do the Poor Connect to theGrowth Process?" Consulting Assistance on Economic ReformDiscussion Paper No. 178. (Cambridge, Mass.: Harvard Institutefor International Development).

White, Howard, and Edward Anderson, forthcoming, "Growth Ver-sus Distribution: Does the Pattern of Growth Matter?", Instituteof Development Studies, University of Sussex.

W o r l d Bank, 1982, Accelerated Development in Sub-Saharan Africa(Washington: World Bank).

World Bank, 2000, World Development Report (New York and Wash-ington: Oxford University Press and World Bank).

61


Date post:	19-Oct-2021
Category:	Documents
Upload:	others
View:	3 times
Download:	0 times

Macroeconomic Policy and Poverty Reduction

Documents