letter from the editor - imf elibrary · 2019-12-11 · to attract private investment to its oil...

Letter from the Editor

aie of the most important challenges developing countries will face over thenext few decades is soaring demand for energy, as economic growth boostsenergy consumption. Substantial new investment will be needed in the energysector to keep up with demand, or growth could falter.

Although the developing countries have attracted the interest of internationalinvestors—they offer large new markets, the world's greatest concentration ofproven oil and gas reserves, and an improved business environment followingsuccessful economic reforms—finance for energy projects is scarce. Manyinvestors and financiers find the commercial and political risks intimidating. Inhis article, "Financing Oil and Gas Projects in Developing Countries," HosseinRazavi outlines steps that governments and project sponsors can take to miti-gate these risks, and describes what the World Bank is doing to facilitate theflow of funds.

Charles McPherson's article, "Policy Reform in Russia's Oil Sector," focuses onthe problems faced by one particular country. The world's largest producer of oilin 1990, Russia has experienced an annual decline in oil production of 7-14 per-cent per year over the past five years. Investments on the order of $50 billion to$60 billion will be necessary to halt this decline; even bigger amounts will beneeded if the oil sector is to live up to its potential. Clearly, funding on this scalewill have to come from the private sector. To attract private investment to its oilsector, Russia will need to stay the course with policy reforms. Although consid-erable progress has already been made under very difficult conditions, muchremains to be done, particularly in connection with reforming the price struc-ture, tax system, and legal framework of the energy sector.

Environmental damage is often the unintended by-product of increasedenergy production and consumption. But, as Dennis Anderson points out in hisarticle, "Energy and the Environment: Technical and Economic Possibilities,"pollution can be significantly decreased, even with the fivefold increase inenergy consumption that is expected to occur in the developing countries overthe next three decades. A number of instruments are already available for reduc-ing emissions—for example, cleaner fuels, low-polluting technologies, and envi-ronmental policy measures such as taxes and environmental regulations—andpromising new technologies are being developed. Moreover, the cost of less pol-luting environmental practices is low or declining—and is even negative insome cases, if environmental benefits are taken into account. The main barriersare policies and attitudes. Investment in research and development, demonstra-tion of technologies, and education and training are therefore as important inpollution-abatement efforts as the other instruments at the disposal of govern-ments and businesses.

The challenge is indeed daunting. To enjoy sustained economic growth andraise standards of living, developing countries need to pursue policy reformsthat will enable them to attract sizable investments in their energy sectors. Atthe same time, they need to safeguard the environment for future generations.As these articles show, the difficulties are great but not insuperable.

ATTENTION READERS

Finance & Development's English edition, beginning with theMarch 1996 issue, is now available on the World Wide Web at

http://www.worldbank.org/fanddYour comments would be welcome. Please e-mail them to [email protected]

FINANCEDevelopment is published quarterly inEnglish, Arabic, Chinese, French,German, Portuguese, and Spanishby the International Monetary Fundand the International Bank forReconstruction and Development,Washington, DC 20431, USA.

Opinions expressed in articles and othermaterials are those of the authors; theydo not necessarily reflect IMF or WorldBank policy.

Periodicals-class postage is paid atWashington, DC and at additional mail-ing offices. The English edition is printedat Lancaster Press, Lancaster, PA.Postmaster: please send change ofaddress to:

Finance & Development700 19th Street NW,Washington, DC 20431Telephone: (202) 623-8300Fax Number: (202) 623-4738English edition ISSN 0015-1947

Claire LiuksilaEDITOR-IN-CHIEF

Asimina CaminisASSISTANT EDITOR

Paul GleasonASSISTANT EDITOR

Luisa WatsonART EDITOR

June LavinEDITORIAL ASSISTANT

Jessie HamiltonSTAFF ASSISTANT

ADVISORS TO THE EDITOR

Masood AhmedAdrienne CheastyWilliam EasterlyNaheed KirmaniAnne McGuirk

Gobind NankaniPeter J. QuirkLant Pritchett

Orlando RoncesvallesGarry Schinasi

Marcelo SelowskyChristine Wallich

For advertising information contactPaul Gleason

Advertising ManagerTelephone: (202) 623-7081 • Fax: (202) 623-4738

Internet: [email protected]

©International Monetary Fund. Not for Redistribution

http://www.worldbank.org/fandd

Energy PolicyFinancing Oil and Gas Projects in Developing Countries

Policy Reform in Russia's Oil Sector

Energy and the Environment: Technical andEconomic Possibilities

Adjustment and Growth

Hossein Razavi

Charles P. McPherson

Dennis Anderson

How Successful Are IMF-Supported AdjustmentPrograms?

The Response of Investment and Growth toAdjustment Policies

The Quality of Fiscal Adjustment and Growth

Guest Article

2

6

10

Susan Schadler

Erik C. Offerdal

George A. Mackenzie andDavid W.H. Orsmond

14

18

21

The Tobin Tax and Exchange Rate StabilityCounterpoint

Paul Bernd Spahn 24

Why a Two-Tier Tobin Tax Won't Work

World Economy in Transition

Janet G. Stotsky 28

Social Indicators of Development Sulekha Pate/ 30

The Revenue Decline in the Countries of the FormerSoviet Union

The Implicit Pension Debt

IMF's New Data Standards Online

Forget Convergence: Divergence Past, Present,and Future

Secured Transactions: The Power of Collateral

Stock Market Development and Corporate FinanceDecisions

Urbanization: The Challenge for the Next CenturyBooks

Adrienne Cheasty

Cheikh Kane and Robert Palacios

Laura Wallace

Lant Pritchett

Heywood Fleisig

Asli Demirguc-Kunt andVojislav Maksimovic

32

36

39

40

44

47

50

IMF Programmes in Developing Countries (Design and Impact) by Tony KillickRussian Reform/International Money by Yegor Gaidar and Karl Otto PohlNorth-South Trade, Employment and Inequality: Changing Fortunes in a Skill-Driven World

by Adrian WoodCoping with Austerity: Poverty and Inequality in Latin America edited by Nora Lustig

Fiscal Decentralization in Latin America, edited by Ricardo Lopez MurphyThe New Multilateralism in Japan's Foreign Policy by Dennis T. Yasutomo

Mark AllenDonal DonovanAlun Thomas

George A. MackenzieGerd SchwartzMark S. Lutz

515153

535556

© 1996 by the International Monetary Fund and the International Bank for Reconstruction and Development/THE WORLD BANK.All rights reserved. Requests for permission to reproduce articles should be sent to the Editor. Finance & Development will normally givepermission promptly, and without asking a fee, when the intended reproduction is for noncommercial purposes.

^^^^Rwmjj ^HKn^^^^^^B|Hn ^ ^ ^QQHRQra QmnJm^OTRfflm^^^^^^^H

June 1996 • Volume 33 • Number 2

n FINANCE&Development


Financing Oil and Gas Projectsin Developing Countries

In the future, investmentopportunities in the oil andgas sector are likely to be con-centrated in developing coun-tries. Project financing isscarce, however, because ofthe commercial and politicalrisks. What can be done tomitigate the risks and attractfunding?

T HE BIGGEST increases in de-mand for oil and gas are occurringin the developing world, which isalso where most of the world's

proven oil and gas reserves are located.International energy companies—investors, equipment suppliers, contrac-tors, and consulting firms—are thereforeshifting their attention from Europe andNorth America to developing countries,which are likely to offer more businessopportunities in the oil and gas sector inthe future.

Although many new projects are beingformulated, most do not take off because ofthe difficulties of securing sufficient financ-ing. Project sponsors are being forced todesign more flexible and innovative financ-ing packages involving a range of partnersfrom both the public and private sectors.Still, the commercial and political risksoften discourage potential partners. In aneffort to facilitate the flow of funds, theWorld Bank recently revised its strategy in

H O S S E I N R A Z A V I

the oil and gas sector, giving greateremphasis to helping governments and pri-vate companies manage and mitigate pro-ject risks.

A changing sectorDuring the past five years, the political

and economic environments in the develop-ing countries have changed drastically, ashas the international oil and gas industry.These changes have had a profound effecton the hydrocarbon sector.

The role of the state has beenredefined. Recognizing that the state doesnot make a good entrepreneur, many gov-ernments have redefined their role as poli-cymaker and regulator. They are giving afreer rein to the private sector and lettingmarket forces determine the most efficientways of supplying commodities and ser-vices. This trend, although global, has hadthe most dramatic results in the former cen-trally planned economies.

The international petroleum mar-ket has changed markedly. In the1970s and 1980s, there was substantial con-cern about the security of petroleum sup-plies and the danger of rising prices. Someof these risks still exist, but crude oil andpetroleum products are now viewed as com-modities that should be supplied throughthe most cost-effective channels. Theemphasis is on procuring petroleum prod-ucts in the international market, and devel-opment of domestic resources is seen asjustified only when oil and gas can be pro-duced and marketed at internationally com-petitive prices.

Environmental concerns are nowprominent. Deterioration of the environ-ment has become one of the primaryconcerns of the international community.

Environmental issues are of particular con-cern with respect to the hydrocarbon sectorin developing countries.

First, oil and gas projects often havepotential environmental and safety risks,which have to be investigated and man-aged. In industrial countries, projects aredesigned and implemented in accordancewith clear and transparent standards, but,in most developing countries, there are noenvironmental standards for the oil and gassector. In the past, major oil companiesapplied in-house standards and acted ascustodians for environmental concerns, butthis is changing as many small privatecompanies begin to operate in the hydro-carbon sector.

Second, existing oil and gas facilities inmany developing countries are operating atsub-optimum standards, causing damageto the local and global environment. Oilspills and gas leakages, which need to becleaned up as rapidly as possible, are ofparticular concern.

Natural gas has become the fuelof choice. Partly because of environ-mental concerns and partly because ofeconomic and efficiency considerations,natural gas has become a popular fuel indeveloping countries. Outside of the coun-tries of the former Soviet Union, the use ofnatural gas in developing countries wasquite limited until recently. In the past fiveyears, gas consumption has increased by 6percent annually. A large share of the gasconsumed is used for power generation, asefficiency of gas-based combined cycleplants has increased significantly.

Business opportunitiesTotal world oil consumption is projected

to grow by about 36 percent between 1995

Hossein Razavi,an Iranian national, is Chief of the Oil and Gas Division in the World Bank's Industry and Energy Department.

2 Finance & Development /June 1996


and 2010. Most of this growth will be indeveloping countries, where demand isexpected to increase in almost all sectors(Chart 1). In contrast, the growth of oil con-sumption in the industrial countries islikely to be limited to the transportationsector. Over the same period, total worldgas consumption is also projected to growby about 36 percent. The growth of gasconsumption in developing and industrialcountries alike will be due largely to theexpansion of gas-based power generation;the largest additions to gas demand willtherefore be in developing countries, wheremost of the expansion of power-generatingcapacity is taking place.

Hydrocarbon reserves are also concen-trated in developing countries (Chart 1).Only 5 percent of total proven oil reservesand 9 percent of proven gas reserves are inthe industrial countries; the remainder arein developing countries.

This concentration of market prospectsand oil and gas reserves, combined with

recent economic reforms, has stimulatedsubstantial interest in the developing coun-tries' hydrocarbon sector. As businessopportunities in the oil and gas sectors ofindustrial countries diminish, energy com-panies are shifting their attention to theenergy investment requirements of develop-ing countries. But the traditional system offinancing oil and gas projects in developingcountries has been dismantled, and a newsystem is not yet in place, resulting in ashortage of financing.

Until the 1970s, most petroleum projectsin developing countries were financed bythe international oil companies, throughinternal cash generation. This situationchanged during the 1970s, when govern-ments became heavily involved in thepetroleum sector to ensure better control oftheir reserves and, in the case of petroleum-importing countries, to quell concernsregarding the security of oil supply.Consequently, funds for oil and gas projectscame from government budgets and official

borrowing, as well as from international oilcompanies. In the early 1990s, emphasisshifted again toward private sector financ-ing, as most governments began limitingtheir involvement in, and budgetary contri-butions to, the oil and gas sector. However,the international oil companies havebecome less willing to finance these pro-jects on their own and have begun toinclude a wide range of partners in projectsfor a variety of reasons, including a politi-cal need for local participation and a desireto share project risks. As a result, fundingof oil and gas projects has become quitecomplex, involving public as well as privateinvestors and financiers.

Project risks are bottlenecksProject risks are normally classified

under two general categories—commercialand political. Commercial risks (e.g., costoverruns, delays, and shortfalls in projectrevenues caused by uncertain sales andprices) are all considered to be under the

ChamOil and gas: demand and resources

Most of the increase in demand is occurring in developing countries

Finance & Development /June 1996 3

Sourece Wold Bankestimates 1 Newly indepenent states of the former soviet Union


Chart 2

World Bank lending for oil and gas projects(1975-99)

Source: World Bank data.Note: Technical assistance refers to loans aimed at improving institutional efficiency, restructuring, etc.

control of project sponsors, while politicalrisks (e.g., expropriation of assets, civilunrest, and foreign exchange inconvertibil-ity) are not. With conventional projectfinancing methods, project sponsorsassume and manage the commercial risksand buy insurance against political risks.

In many developing countries, there isanother dimension to political risk that ismore difficult to handle. The lack of well-established legal, institutional, and regula-tory systems and policies makes it possiblefor governments to take unpredictableactions that could substantially affect costsand revenue streams—particularly if, forexample, governments control the domesticprices of oil and gas, or decide to changethe terms of oil and gas taxes and royalties.This risk is the biggest deterrent to privateinvestment in the oil and gas sector ofdeveloping countries. Even in countrieswhere governments have taken steps toestablish a stable framework and clarifypolicies, project sponsors (and financiers)may not have full confidence that the newbusiness environment will remain un-changed and that the government will ful-fill its obligations fairly and consistently.

Project risks are allocated to the differentparties involved through numerous agree-ments and contracts included in the secu-rity package. These documents are aimedat protecting the interests of the sponsorsand, more often, at providing comfort tolenders that risks will be managed to a rea-sonable extent. From a lender's point ofview, three questions need to be answered.First, can the project be constructed andcommissioned within the planned scheduleand budget? Second, can the project gener-ate the projected net revenue? Third, canthe net revenue be allocated and paid back

to the lenders and investors according tothe project agreement? In connection withthese questions, lenders want to know whowould be responsible for damages in theevent the project fails in any of these areas.

The issue of political risk should beaddressed at the outset of project prepara-tion. Most investors and financiers are con-vinced that commercial risks can beeffectively addressed when the time comes,but they feel that political risks cannot becontrolled by anyone. Thus, they do nottake a proposal seriously until they receivesome assurance that political risks aremanageable. Political risks can be miti-gated through a variety of measures,including different forms of guarantees andthe involvement of certain types of part-ners—for example, a key state entity orpowerful local individuals and companies.Formal guarantees can be provided by hostgovernments and by multilateral and bilat-eral agencies. Often, rather than choosingone form of comfort over another, sponsorswill try to combine them to get the mostcomprehensive coverage at the lowest pos-sible cost.

Commercial risks can be mitigatedthrough two distinct avenues. First, spon-sors need to reach an agreement with thegovernment of the host country or with gov-ernment entities about some aspects ofmarketing the project's output. The govern-ment's role varies depending on the countryand the type of project. For gas projects, thegovernment's role is substantial becausemost of the output is bought by a stateentity or is sold at prices regulated by thestate. Therefore, project sponsors need tosecure take-or-pay or throughput agree-ments with the state entities. The govern-ment needs to guarantee the credibility of

the state entities or to provide assurancethat it will permit any necessary increase inenergy prices. Securing government guar-antees and agreements takes a relativelylong time, particularly in countries that lackclear precedents. The second avenue in-volves negotiating with contractors, equip-ment suppliers, fuel suppliers, operatingcompanies, and so on, to determine theirwillingness to compensate for damages ifthey fail to fulfill their obligations. Althoughtechnically complex, this process is nor-mally accomplished efficiently because it isdriven by commercial incentives.

The role of the World BankIn 1995, the World Bank re-examined its

oil and gas lending strategy in consultationwith member countries, representatives ofthe international petroleum industry, andother sources of finance. The strategy wasrevised to take account of the changes thathad taken place over the previous 5 to 10years and to enable the Bank to providemember countries with the most effectiveassistance. In accordance with the newstrategy, which puts substantial emphasison helping developing countries to mitigateproject risks and enabling governments toserve as effective regulators, the Bank willsupport the creation of open and competi-tive markets; encourage protection ofhealth, safety, and the environment; andserve both as a magnet for private capitaland as a lender of last resort. The Bank'snew agenda, designed in cooperation withother World Bank Group members—theInternational Finance Corporation (IFC)and the Multilateral Investment GuaranteeAgency (MIGA)—includes technical assis-tance, lending, and guarantees as follows:

• Helping countries to establish legal



and regulatory frameworks that facilitateprivate investment and enhance efficiency.

• Assisting in the restructuring of publichydrocarbon companies through corpora-tization, commercialization, and privatiza-tion.

• Identifying more efficient and benignfuels and promoting the substitution ofgas for coal and oil when this would beboth more efficient and environmentallybeneficial. As a recent study in India hasshown, liberalized access to modern fuelscan—directly or indirectly—help poorhouseholds move up the "energy ladder" tocleaner, more efficient fuels for cookingthan wood fuels and agricultural residuesfor cooking.

• Assisting governments in the environ-mental cleanup of existing oil and gas facil-ities and in establishing standards andinstitutions required for monitoring the en-vironmental impacts of oil and gas projects.

• Facilitating international trade projects(mainly gas pipelines but also liquefied nat-ural gas projects and oil pipelines). Gaspipeline projects may particularly benefitfrom Bank support because, as investmentswith long payback periods, no alternativeuses, and often uncertain local markets,they are seen by private investors as rela-tively risky. World Bank participation as afacilitator is warranted in projects that arevery complex and that require direct partic-ipation by the state or a state company.

• Financing urgent, economically soundprojects in oil development, processing,transmission, and distribution—but only inthe absence of sufficient private sectorresources.

• Providing guarantees to cover risks forimportant and environmentally vital pro-jects. Since September 1994, the Bank hasbeen offering two different types of guaran-tees as a way of "leveraging" private invest-ments in key projects: (1) a partial riskguarantee covering governmental nonper-formance of contractual obligations in aproject (such as selling inputs to the pro-ject, buying end-products, or making cer-tain related investment); and (2) a partialcredit guarantee that typically extendsmaturities beyond what creditors wouldotherwise provide, for example, by guaran-teeing late-dated repayments.

The level of World Bank lending to theoil and gas sector has fluctuated substan-tially over the past two decades in responseto changes in market conditions as well asto shifts in the Bank's own policy. After theoil crisis of the 1970s, the Bank began toplay a prominent role in the oil and gas sec-tor, assisting member countries in develop-

ing their indigenous energy resources.Bank lending, initially concentrated inexploration and development of hydrocar-bon resources, climbed to $1 billion in 1983.This rapid expansion caused concern thatthe Bank might pre-empt the private sector.The Bank therefore imposed limitations onits lending for oil exploration and produc-tion. These limitations, combined with aperception that future oil demand would beweak, caused lending to drop sharply (it fellto $300 million in 1986). By 1990, the Bankwas again active in the hydrocarbon sector,but this time the emphasis was on promot-ing private sector involvement and sup-porting the development of natural gas as asubstitute for coal and oil.

Bank lending in the oil and gas sector isexpected to reach about $1 billion yearlyduring the second half of the 1990s (Chart2). However, the sectoral composition isexpected to shift away from upstreamindustries to infrastructure, reflecting theBank's view that the private sector willinvest in upstream projects if the infra-structure for delivering the output is inplace. The Bank is also increasing itsemphasis on technical assistance, whichwill be aimed at facilitating sector restruc-turing, privatization, private sector devel-opment, and the establishment ofenvironmental standards and monitoringinstitutions.

RecommendationsSponsors of oil and gas projects in devel-

oping countries often find themselves in aseemingly never-ending process whiledesigning the ownership structure, securitypackage, and financing plan. They must tryto achieve conflicting objectives—minimiz-ing the risk and financing costs while maxi-mizing the likelihood of successful andtimely project implementation. To manageproject risks, sponsors try to get more play-ers involved. There is a strong tendency toinvolve local partners—in the hydrocarbonsector, these are normally state oil and gascompanies. Despite the great differencesbetween state-owned and privately ownedcompanies, recently there has been a sur-prising convergence in the way these enti-ties seek to finance projects in developingcountries. State-owned companies, whichhave traditionally financed their projectsthrough government budgets or officialgovernment-sponsored borrowing, haverecently turned to commercial sources offinance, such as commercial bank loans, pri-vate bond placements, and sales of equity instock markets. And private investors nowseek to incorporate official sources of

finance with government sponsorship intotheir projects—even going so far as to formjoint ventures with state entities.

The trend is clearly toward using the fullrange of financial tools available for oil andgas projects in developing countries. Thus,with a larger number of project partnersand a wider range of financial instruments,project preparation has become much morecomplex. A critical question while design-ing the ownership structure is the role ofgovernment or state entities in the project.Often, some ownership by state entities pro-vides access to a variety of sources of offi-cial funding, but, in practice, most officialfinanciers hesitate to support full stateownership. The appropriate degree of stateparticipation varies, depending on the typeof project and the country's business envi-ronment. For upstream oil and gas projectsand refineries, the role of the state should beminimized. For infrastructure projects, alarger role for the state is normallyjustified. With an appropriate ownershipstructure, sponsors could receive supportfrom multilateral institutions such as theWorld Bank, which now offer a menu ofinstruments that are more flexible than inthe past. The different types of assistanceoffered by the World Bank can be combinedto facilitate the mobilization of funds from avariety of sources.

Governments can further facilitateinvestment in the oil and gas sector byestablishing clear regulatory and fiscalregimes. Often the risk-reward profile of aproject can be substantially improved byclarifying the rules of the game and assur-ing project sponsors and lenders of the sta-bility of relevant policies. Governmentsmay also benefit by studying the practicesof other countries with regard to fiscal sys-tems and the parameters of regulatoryregimes. This is another area in which theWorld Bank and other multilateral institu-tions can offer valuable assistance based ontheir cross-country experience. IF&D

For information on obtaining funds frommultilateral, bilateral, and commercialfinanciers, see Hossein Razavi, 1996, FinancingEnergy Projects in Emerging Economies (Tulsa,Oklahoma: PennWett Books).



Policy Reform in Russia's Oil SectorC H A R L E S P. M c P H E R S O N

Despite difficult circum-stances, policy reform inRussia's oil sector has beenimpressive. Nevertheless,further reforms and invest-ments in excess of $60 billionover the next 10 years will beneeded to help the sectorrealize its potential

R USSIA'S OIL sector is critical tothe country's overall economicrecovery. Oil accounts for 5 per-cent of Russia's GDP, 20 percent of

its foreign exchange earnings, and 10 per-cent of its fiscal revenues. Over the past fiveyears, oil sector performance has fallen farbelow its potential. The world's largest oilproducer in 1990, Russia has now slippedto third place, with production declining atannual rates of 7-14 percent. Investmentson the order of $50 billion to $60 billion willbe required over the next 10 years just tostabilize production, and even larger sumswill be needed to bring about a recovery.Funding at these levels will not be forth-

coming unless the private sector, both for-eign and domestic, can be engaged. Privatesector commitments will depend cruciallyon policy reform in the oil sector.

Achieving policy reform in Russia haspresented major challenges. Many reformconcepts were entirely new to Russia fouryears ago, especially those based on marketprinciples or the need for legal trans-parency. The number and range of partici-pants involved has also made the reformprocess highly complex (see table). Againstthis background, Russia's achievements inthe oil sector have been impressive. Notsurprisingly, however, a number of criticalissues are still outstanding.

Oil pricesAt the beginning of the reform period

(1992), Russia's oil prices were administra-tively set at less than 5 percent of world lev-els. A move to market prices was urgentlyrequired to better allocate resources in bothproduction and consumption. It was alsorecognized that appropriate pricing wouldreduce Russia's disproportionately highconsumption of fossil fuels with beneficialeffects on the environment.

These arguments seem straightforwardenough but they ran into serious obstacles,among them charges that oil price liberal-ization would fuel inflation, cause unaccept-

able hardship for consumers, and createwindfall profits for the "oil barons." Similararguments against price reform were com-monly heard in the industrial countries fol-lowing the leap in world oil prices in the1970s.

There were at least two additionalsources of resistance to oil price liberaliza-tion in Russia. First, complete price reformdepended—and still depends—on the unre-stricted access of Russian oil to world mar-kets. Domestic policymakers have ex-pressed concern over the social and politi-cal consequences of a diversion of crude oilsupplies away from the domestic market infavor of exports. Second, any significantdifference between domestic and worldprice levels translates into a favorable envi-ronment for arbitrage and establishesvested interests in blocking reform.

These considerations prevented oil frombeing included in Russia's general price lib-eralization of January 1992. Nevertheless, aprocess of steady reform got under way.Within 18 months, all administrative con-trols, limitations on profit margins, andsales allocations for the domestic marketwere largely ended. Once prices had beenliberalized on the domestic market, thereformers turned their attention to tradeliberalization. Oil export quotas were abol-ished in early 1995, and oil export duties,

Charles P. McPherson,a Canadian national, is Principal Energy Economist in the Infrastructure, Energy, and Environment Division of the Europe and Central Asia RegionalOffice of the World Bank.



Oil sector reform involves many participants

Price reform

President's officeMinistry of Fuel and PowerMinistry of FinanceMinistry of EconomyMinistry of Foreign Economic

RelationsDuma Economic PolicyCommittee

Regional governmentProducersConsumersWorld Bank and IMFBilateral lenders

Tax reform

President's officeMinistry of Fuel and PowerMinistry of FinanceMinistry of EconomyDuma Economic PolicyCommittee

Regional governmentProducersWorld Bank and IMFBilateral lenders

Legal reform

President's officeMinistry of Fuel and PowerState Committee on GeologyMinistry of Environment andNatural Resources

Duma Industry and EnergyCommittee

Duma Ecology and NaturalResources Committee

Duma Economic PolicyCommittee

Federation CouncilRegional governmentProducersWorld Bank and IMFBilateral lenders

Enterprise reform

President's officeMinistry of Fuel and PowerMinistry of EconomyState Property AgencyRegional governmentProducersWorld Bank and IMF

Foreign direct investment

President's officeMinistry of Fuel and PowerState Committee on GeologyRegional governmentProducersWorld Bank and IMFBilateral lenders

Source: World Bank.

which were driving a wedge betweendomestic and international prices, werereduced. By the end of 1995, domesticprices were at parity with internationalprices, net of the export duty, and wereequivalent to 70 percent of world prices inabsolute terms (Chart 1). The governmentannounced a 50 percent reduction in the oilexport duty, effective April 1,1996, and hasmade a commitment to abolish the remain-ing duty by July 1,1996.

Although offset to a degree by paralleltax increases, higher domestic prices haveimproved the industry's cash flow andincentives, and contributed to some re-newed rehabilitation of low-cost productionfacilities. Domestic consumption of oil and

Chart 1

Russian crude oil pricesconverging to world levels 1

Source: World Bank.1 Crude oil price at border.

oil products has declined sharply, not onlybecause of industrial restructuring anddeclining incomes but also, in part, becauseof the price increases.

Oil taxationTax reform remains one of the most

pressing oil sector policy issues. The heartof the debate lies in the tension between theurgent need for revenue to reduce the bud-get deficit and the equally compelling con-cern that increased oil taxation not "kill thegoose that lays the golden eggs." It shouldbe possible to reconcile these two objectivesby putting in place a tax system that is flex-ible enough to encourage a wide range of oilproduction and development projects, whilecollecting a significant and progressiveshare of the economic rents from the sector.

Russia's present oil tax system attemptsto satisfy both fiscal and incentive objec-tives by varying the incidence of an excisetax and certain revenue taxes as a functionof the estimated profitability or cost of anoil producer's operations. The governmenthas experienced problems with the applica-tion of this system, however, and a criticalreview of its efficiency is planned.

Russia is already well on its way towardintroducing new tax arrangements formajor oil development projects that are nowpending. To get such projects off theground, the government has negotiated aseries of production sharing agreements(PSAs) with foreign and domestic investorson an individual project basis. These agree-ments contain attractive, "state-of-the-art"fiscal provisions and will become effectivewhen acceptable enabling legislation isenacted. The PSA formula is relatively

simple, calling for a modest royalty or rev-enue tax, a corporate profits tax, and anadditional profits-based tax (or productionshare) that would escalate with the actualprofitability of a project.

The formula will apply only to projectsto be carried out under PSAs. There is noreason, however, why the same concept,with some modification, should not be moregenerally applied. A shift in this direction isone of the options the government is likelyto explore over the next 12 to 18 months.

One of the advantages of the new systemis its greater reliance on profits-based—asopposed to revenue-based—taxation. Theprincipal concern of the government inassessing any possible change to the exist-ing system is that the fiscal contribution ofthe oil sector might be jeopardized. In par-allel with its review of tax design, the gov-ernment is committed to taking much-needed steps to strengthen administrationand enhance tax compliance.

LegislationA legal framework setting out clearly the

rights and duties of the government andcontractors, and covering key issues suchas ownership and taxation is a sine quanon for investment in the oil sector of anymajor oil-producing region. While somesaw the urgency of establishing such aframework, many Russians, including lead-ers in the oil industry, did not at first recog-nize the importance of the necessarylegislation. During the Soviet era, theindustry had operated through administra-tive fiat and the lobbying of powerfulcontacts. Now, acceptable framework legis-lation is widely regarded as essential, but



questions remain concerning what form thelegislation should take:

• Should legislation cover all resources,or focus specifically on petroleum?

• Should the legal regime be license-based, raising the specter of unilateralwithdrawal or revision of licenses by thegovernment, or should it be a contract-based regime that is mutually binding onthe government and the investor?

• Should the parliament approve allcontracts and amendments, or shouldreasonable authority be delegated to thegovernment?

• How should federal and regional inter-ests and authority be balanced in such diffi-cult areas as approving licenses andcontracts, controlling operations, safe-guarding the environment, and sharing taxrevenues?

• What special privileges, if any, shouldbe granted to foreign investors?

These conceptual issues and most of thestandard provisions of petroleum legisla-tion and contracts represented largelyuncharted water for the Russian reformers.Not surprisingly, progress has come in fitsand starts. Overlapping and sometimescompeting legal drafting initiatives werecommon, as were serious inconsistenciesand conflicts among the laws and draftsaffecting petroleum operations.

An omnibus law, the Law onUnderground Resources, covering all natu-ral resources was enacted in 1992, but itcontained no specific provisions forpetroleum and established a license-based,rather than a contract-based, regime.During 1992-95, considerable effort wasmade to draft more than half a dozen lawsor regulatory packages, involving severalgovernment ministries, parliamentary com-mittees, and industry lobbies, and somecoherence began to emerge. Over the sameperiod, foreign and domestic investorsmade great strides in negotiating PSAscontaining provisions that were standardinternationally for the petroleum industry.The Law on PSAs, enacted in December1995, represents a very important steptoward an acceptable legal framework formajor PSA investments in the oil sector.Several key issues relating to assurances ofa contract-based regime, contract stability,governing law, and arbitration, are stillunresolved, but the government is workingactively to address them, drafting clarify-ing regulations and, as required, legislativeamendments.

Enterprise reformViewed against the background of the

legacy of central planning and a completeabsence of accountability, transparency,and competition, Russia's achievements inthe areas of institutional and enterprisereform in the oil sector have been consider-able. Legislation and decrees have, withreasonable clarity, defined the roles of gov-ernment at both the national (Moscow) andregional levels. The government hasincreasingly distanced itself from interfer-ence in the commercial operations of the oilindustry and recently decided against thecreation of a national oil company.

As early as November 1992, a presi-dential decree established a blueprint forenterprise reform: all oil enterprises were tobe given independent legal (joint stockcompany) status as quickly as possible; upto 12 vertically integrated oil companieswould be formed; and the government'scontrolling shareholdings would be sold offin three years. These goals have all beenlargely met. The one cloud still hangingover this chapter of the reform story isthe lack of transparency and the hastewith which major oil enterprise privatiza-tions were handled in late 1995. It is gener-ally accepted that with more carefulmanagement, and more openness andtime, these sales could have realized muchlarger revenues for the government and

Chart 2

Russian-foreign joint ventureshave performed well1

Source: World Bank.1 Crude oil production.

attracted more experienced, market-oriented investors.

With independent legal status and theprospect of privatization, enterprises havebegun to restructure, to put their opera-tions on a commercial basis. The need togain access to international capital marketsand to attract share capital has led mostcompanies to conduct independent legal,financial, tax, and reserve audits, therebypromoting good corporate governance.Recent gains notwithstanding, tax arrearsare a major, urgent problem for companiesand the government alike.

Finally, competition, supported by legis-lation, appears to be developing well at theproducing level and even at the refiningand distribution levels of the "oil chain."Local monopolies are likely to be short-lived. Natural monopolies in oil pipelinetransport are still state-owned and will besubject to regulation of rates and rules ofaccess. The potential for private ownershipof new transport projects is being seriouslydiscussed.

Foreign direct investmentForeign private investment is playing,

and will continue to play, a critical role inachieving stabilization and recovery in theoil sector. What it will take to attract for-eign capital is largely reflected in thereform agenda discussed above: interna-tional prices, export access, an internation-ally competitive tax and legal framework,restricted government intervention, andwell-run, financially sound Russian coun-terpart companies. Attracting foreign capi-tal will also depend on a welcoming oraccommodating Russian attitude towardsuch investment.

Russians have generally been deeplyskeptical of foreign involvement in theireconomy. In good part, this is due to nation-alist sentiments, which have deep historicalroots, as well as to professional pride.Foreign direct investment has been viewedas a necessary evil, providing essentialbridge finance during economic transition,at the expense of giving away part of thenational heritage. In fact, Russia stands togain a great deal from the involvement offoreign capital, as do other major oil regionsin the world. These gains include sharedfinancial and technical risk, accelerateddevelopment, and shared technical andmanagerial ideas and practices. Theseadvantages have become increasinglyapparent to the larger Russian firms thathave now been exposed to major interna-tional firms for several years. Their foreignpartners hold an average of less than

Finance & Development /June 19968


45 percent of international deals becausethey recognize the benefits of joint ventures.

Through 1995, approximately $1 billionhas been invested in foreign joint-ventureprojects in Russia under special incentiveschemes providing exemptions from themore onerous features of the existing taxsystem. These projects have performedwell relative to the rest of the sector (Chart2), but, because they have been limited inscope and number, represent only a smallfraction of total production.

Russian focus has now shifted to thenegotiation of megadeals with foreigninvestors in the context of PSAs. As part ofthis process, many important issuesbetween the government and foreigninvestors, and between Russian and foreignjoint ventures, have been resolved or atleast clearly identified. Ten of these pro-jects, worth an estimated $60 billion in newinvestments, have been negotiated with 14different international companies. Theirimplementation depends on the completion

of an enabling legal framework. Once theseprojects are fully operational and the bene-fits of cooperation become more tangible,Russian resistance to foreign investment, atleast in the oil sector, could erode rapidly.

Next stepsGiven the difficult circumstances,

Russian progress on oil sector reform hasbeen substantial. Prices have risen from lessthan 5 percent of world levels to 70 percent;an acceptable tax package has been devel-oped, which was negotiated in the contextof PSAs; and the government has commit-ted itself to seriously considering broadertax reform. Competing drafts of petroleumlegislation have begun to coalesce, and thereis growing consensus on the required legalreforms. The recently enacted Law on PSAswas a major achievement. A rational insti-tutional framework for the oil sector nowexists, with minimal state interference, andenterprises have rapidly become more com-mercial. Despite the fact that many

Russians remain ambivalent about foreigninvestment in the oil sector, truly world-scale projects are now pending on a foreignjoint-venture basis, and more are planned.

These achievements notwithstanding, anumber of critical obstacles to reformremain: export duties need to be abolishedin order to complete price reform; PSA fis-cal provisions should be further definedand confirmed by legislation; broader taxreform for the oil sector needs to beurgently addressed; regulations and legisla-tive amendments will be required to com-plement the PSA law and unlock thebillions of dollars in pending PSA invest-ments; and further progress on commercial-ization, audits, and tax compliance isrequired to bring oil sector enterprises upto international standards.

The Russian reform context remainsvery complex and, if anything, this com-plexity is increasing as the number ofactive participants grows. Nevertheless, thepotential payoff to reform is enormous. F&DJ

FINANCINGENERGY PROJECTSIN E M E R G I N G E C O N O M I E S

C h i e f , O i l a n d G a s D i v i s i o n , W o r l d B a n k

This book provides first-hand information and analysis regarding how multilateral, bilateral, and commercial financiers makedecisions about oil, gas, and electric power projects in developing countries. It describes the intricacies of public and private financingof energy projects and provides guidance in the preparation of acceptable project packages for upstream and downstream oil and gasinvestments as well as power generation, transmission, and distribution projects.

Contents: Part I: The Setting, includes chapters on fundamentals of project financing and challenges specific to financing projects in developing countries.Part II: Getting to Know the Financiers, includes chapters on multilateral institutions, regional development banks, bilateral agencies (export credit and aid agencies), and

commercial funds (domestic and international capital markets). Part III: Designing an Acceptable Project Package, includes chapters on establishing economic and financialviability, structuring the financing package and addressing environmental standards. Annexes: include coordinates for the finance and aid agencies, sources of information for

preparation of energy projects, and guidelines for preparation of environmental assessments.

peniWell.^ PUBLISHING COMPANY

0-87814-469-2 • 285 Pages • January 1996 • $74.95

TO ORDER, FAX: 202-614-0200Export orders: Add a 20% handling fee plus $15.U.S. orders: Shipping, handling and sales tax where applicable willbe added. Canadian orders: 7% GST will be added.

To expedite order,please refer to the codebelow when ordering.

PBAV01



Energy and the Environment:Technical and Economic Possibilities

D E N N I S A N D E R S O N

Total energy consumption indeveloping countries isexpected to soar over the nextfew decades. But this does notnecessarily bode ill for theenvironment. Technologicaladvances are making itpossible to reduce pollutionconsiderably, even as energyuse increases.

D EVELOPING countries will soonbe the world's largest markets forenergy. Their total energy con-sumption today is only half that of

the rich countries—and their per capitaenergy consumption a mere one-tenth ofwhat it is in the rich countries—but it isdoubling every 15 years and is expected toincrease fivefold over the next threedecades or so in the course of economicgrowth. This projection assumes signifi-cant improvements in energy efficiency,without which the increase could be higheryet. Moreover, even with an increase of thismagnitude, per capita energy consumptionin the developing countries would still berelatively low—allowing for populationgrowth, it would be less than one-fourth ofper capita energy consumption in theindustrial countries today.

Will developing countries be able toincrease energy use while reducing pollu-tion? From a technical and an economic per-spective, the answer is yes, if theenvironmental policies required are put inplace. It is possible to reduce pollution byfactors of 10 or more in the most seriouscases, even if energy consumption levelsrise fivefold. Furthermore, developing coun-tries would find themselves better off botheconomically and environmentally.

A basic identityLet us recall a basic identity relating pol-

lution to energy use:Emissions of pollutants = [energy use] x

[emissions per unit of energy use]

In every country, opportunities to reduceenergy use through efficiency improve-ments can always be found. Decreasingthe subsidies—common in many coun-tries—for fossil fuels and electricity pro-duction would reduce both physical andeconomic waste, as would innovations thatimprove the efficiency with which energy isused in factories, commercial establish-ments, homes, electricity production, andtransport. In electricity production alone,the amount of fossil fuels needed to gener-ate a kilowatt-hour has declined by 90 per-cent over the past hundred years, almostentirely because of technical advances thatincreased the thermal efficiency of powerstations. Yet demand for electricity doubledevery decade for more than 70 years, partlybecause energy efficiency also reducedcosts and prices. The demand for energy indeveloping countries is highly income elas-

tic: per capita income elasticities for theconsumption of electricity, for example, arecurrently 2.0 or higher. Also, more than 2billion people are still without (or areunable to afford) electricity, oil, or gas fordomestic purposes. As incomes rise indeveloping countries, energy use can beexpected to increase appreciably, even withcontinued gains in efficiency.

Hence we need to turn our attention tothe second term on the right-hand side ofthe equation—emissions per unit of energyuse. The evidence that these can be reducedis encouraging. Data on the emissionsintensities of technically proven low-pollut-ing practices relative to the emissions inten-sities of practices still in widespread use indeveloping countries are presented in thetable. The pollution indexes show howlarge the scope for improvement is in fourkey areas:

• Household fuels. The pollution fromsmoke and the damage to natural resources(e.g., soil depletion and erosion) caused bythe use of fuelwood and dung for cooking.

• Electricity production. Particulatematter emissions and acid deposition fromthe use of coal.

• Motor vehicles. A range of tailpipeemissions from diesel and gasoline engines.

• All fossil fuels. Carbon dioxide (CO2)emissions and climate change.

The striking feature of the low-pollutingpractices listed is that emissions (or ero-sion, in the case of soils), with the partialexception of nitrogen oxides (NO*), can bereduced to very low levels, often to one-hun-dredth or less of emissions levels resultingfrom polluting practices. Even if energy use

Dennis Anderson,a UK national, is Senior Adviser in the World Bank's Industry and Energy Department.

10 Finance & Development / June 1996


Low-polluting technologies are available

Relative pollution intensities of polluting and low-polluting practices forselected activities and pollutants

(polluting practice = 100)

Source and type of emissionsor environmental damage

Index per unit of outputLow-

Polluting pollutingLow-polluting

practices

Household fuelsSmoke from firewood, dung 100

Soil erosion (sediment yield) 100

Electricity productionPaniculate matter 100Carbon monoxide (CO) 100Sulfur dioxide (SO2) 100Nitrogen oxides (NOX) 100

Motor vehicles: diesel enginesParticulate matter 100Sulfur dioxide (SOZ) 100

Motor vehicles: gasoline enginesLead 100Carbon monoxide (CO) 100Nitrogen oxides (NOX) 100Volatile organic compounds 100

0.0 Gas, kerosene<1 to 5 Stoves with flues<1 to 5 Agroforestry; erosion-prevention

practices such as contouring,mulching, use of vetiver grass,"no till" agriculture

<0.1<0.1

Oto <55 to 10

<105

05

205

Natural gas; clean coal technologies;scrubbers; low-sulfur fuels;low NOX combusion methods;emission control catalysts

Clean fuels and particulate trapsLow-sulfur fuels

Unleaded and reformulated fuels;catalytic converters

All fossil fuelsCarbon dioxide (CO2) 100 <01 Renewable energy sources

Sources:Energy-related pollution. Smith (1988) for household fuels. For electricity production, gasoline and

diesel engines, CO2 emissions, and marine pollution (oil), a review of technologies and evidence ofpollution abatement are provided in Anderson (1991), drawing on Organization for EconomicCooperation and Development (1989), Asian Development Bank (1991), and Bates and Moore(1992). The possible negative figure for CO2 emissions could be realized by using biomass—espe-cially wood—as an energy source in a renewable way, since this would be associated with anincrease of carbon storage.

Soil erosion. Site-specific evidence (for over 200 cases) of erosion with and without the erosion con-trol methods noted is presented in Doolette and Magrath (1990). The decreases in erosion ratesshown here are based on their data for 11 cases in Taiwan Province of China and correspond to thebest agronomic (but not necessarily the most expensive) practices.

1 Can be negative if biomass (wood) fuels are used.

expands fivefold, therefore, a 90 percentreduction or more is technically feasible inmost cases, and a substantial reduction ispossible in all cases.

Household fuelsThe World Bank's World Development

Report 1992: Development and tlieEnvironment (WDR 1992) noted that stud-ies of smoke from the use of fuelwood anddung ("biofuels") for cooking in rural areas". . . have found particulate matter levelswhich regularly exceed by several ordersof magnitude the safe levels of WHO[World Health Organization] guidelines—[SJmoke contributes to acute respiratoryinfections that cause an estimated 4 milliondeaths annually among infants and chil-dren. Recurrent episodes of such infectionsshow up in adults as chronic bronchitis and

emphysema, eventually contributing toheart failure."

Nearly 2 billion people are dependent onthese fuels. The report's conclusion—thatindoor air pollution is one of the mostsevere environmental problems facing low-income developing countries—has not beendisputed.

Indoor air pollution could be almostentirely eliminated by substituting gas,kerosene, or electricity for the fuels nowused in cooking. Comparisons of fuel useacross countries and over time show that asteady transition to cleaner fuels occurs asincomes rise and as industries expand. Bythe time per capita annual income in agiven country has risen to around $1,500,the transition to modern fuels is almostcomplete. However, no fewer than 70 devel-oping countries with populations totaling

3.5 billion have incomes well below thislevel and are likely to be dependent on bio-fuels for cooking for some time. Other mea-sures thus need to be taken in the interim. Ithas been found, for example, that improvedwood stoves with flues not only raiseenergy efficiency—typically, by 30-50 per-cent—but also reduce indoor pollution—bya factor of 20 to 100—to levels well withinWHO guidelines.

What can be done about the other envi-ronmental impacts associated with thewidespread use of biofuels, such as soil ero-sion, the loss of soil nutrients, and defor-estation? As indicated in the table, changesin agricultural practices show immensepromise, not only for accelerating afforesta-tion and taking pressures off forests andwoodlands but also for raising agriculturalproductivity by reducing soil erosion—infact, by offering ways to regenerate top-soil—and improving the nutrient and mois-ture content of soils. Examples areagroforestry practices, contouring, terrac-ing, bunding of fields (that is, creatingembankments in them to control the flow ofwater), and the planting of vetiver grass.

Electricity productionControls of particulate matter emissions

through electrostatic precipitators were pio-neered more than 40 years ago in industrialcountries and have been widely adopted innew plant since the late 1950s. Emissionshave been reduced by two to three orders ofmagnitude (factors of ten) relative to con-ventional coal boilers with mechanicalcontrols.

In the 1960s and 1970s, the problem ofacid deposition became more widely recog-nized. Once again, there was an innovativeresponse, this time in the form of flue gasdesulfurization, a switch to low-sulfur coalsand gas, and, more recently, developmentsin combustion technologies—often knownas "clean coal" technologies—such as flu-idized bed and coal gasification. Theremarkable growth of commercially provenworld gas reserves has also opened up newopportunities for a very low-polluting andefficient means of electricity generation.Abatement levels of 95 percent for sulfurdioxide (S02) are now feasible—and even of100 percent, if gas is available. For NOX,abatements of around 90 percent are feasi-ble using catalysts and by changing boilerdesigns to reduce combustion temperatures.

Motor vehiclesIn terms of the most harmful tailpipe

emissions of local pollutants—particulates,lead, carbon monoxide, volatile organic



compounds, and nitrous oxides—the story is roughly the same asfor coal-fired plants for electricitygeneration. The rich countrieshave already greatly reducedemissions through the technolo-gies now in use, mainly unleadedand reformulated gasolines andtailpipe controls. Urban air pollu-tion in developing countries couldbe substantially reduced throughapplication of these technologies.

There is, however, a major qual-ification—the effects of trafficcongestion. Vehicle emission stan-dards are usually based on "sta-tionary tests." But emissions riserapidly when vehicles accelerate,stand in traffic, and make frequent stopsand starts—in short, when there is trafficcongestion. In addition, the accumulation ofpollutants in the atmosphere is the sourceof the damage, rather than the emissionsper se. Thus, the picture is not quite as rosyas the data in the table suggest. However,we do know that pollution from lead infuels can be eliminated; there has been amajor improvement in industrial countries,compared with most developing coun-tries, where reformulated gasolinesand tailpipe controls are not inwidespread use. Progress in reducingurban pollution from motor vehicleswill therefore depend greatly on trafficmanagement, congestion pricing, andurban transport policies, as well as onthe introduction of reformulated fuelsand tailpipe controls.

Global warmingMore than 90 percent of the world's

primary commercial energy demands aremet by fossil fuels, 7 percent by nuclearpower, and 3 percent by hydroelectricity.Noncommercial traditional fuels—fuel-wood, crop wastes, and animal dung—addanother 10-15 percent to primary energysupplies. On current trends, the share offossil fuels in total energy consumptionseems likely to rise further for four reasons:the growth of commercially proven re-serves; technical progress in extracting andusing fossil fuels, and, thus, lower costs; thesubstitution of fossil fuels for wood anddung for cooking; and the relatively highcosts and environmental problems associ-ated with nuclear power. Will global warm-ing change the situation? Are alternativesto fossil fuels emerging should there be aneed for them?

Despite the uncertainties about globalwarming and its possible consequences, we

Chart 1Global carbon emissions under

two scenarios

Source: Anderson (1991), (1994).

do know that it will not be possible to pre-vent the accumulation of carbon in theatmosphere unless noncarbon (or non-net-carbon-emitting) alternatives become avail-able. Improving energy efficiency will helpand is important for economic as well asenvironmental reasons, but it will not pre-vent carbon accumulations from growingexponentially or indefinitely, so long ascarbon emissions from the burning of fossil

"The main barriers topollution abatement arepolicies and attitudes."

fuels exceed 2-3 billion tons per year—thecurrent estimate of the "natural" net rate ofabsorption of carbon by the earth's oceansand land masses. Presently, the rate ofemissions is around 6 billion tons per year,and emissions are growing almost in directproportion to world energy demand; it isconceivable that emissions will exceed10 billion tons in 20 years, and 20 billiontons in 50 years—and this would be in anenergy-efficient world.

Alternative technologies are emerging,however (Chart 1). These technologies arebased on direct solar energy—primarilyphotovoltaics and solar-thermal schemesfor power generation—or on other renew-able energy sources such as wind andbiomass. (See K. Ahmed and D. Anderson,"Where We Stand With RenewableEnergy," Finance & Development, June1993; Jennings, 1995; and World EnergyCouncil, 1993.) To the list should be addedgeothermal resources, which hold consider-able promise. These technologies have anumber of attractive features:

• Potential for furtherdevelopment. Although al-ready proven, they are a newindustry and fertile ground forinnovation and discovery. Costscontinue to decline, and there is arapidly growing market for solarand solar-derived technologies indeveloping countries. Privateindustry is especially active intheir development.

• Modularity. All of thesetechnologies can be designed forsmall or large-scale uses.

• Short lead times.Installation takes months, ratherthan years.

• Low land requirements(except for biomass). It has been estimatedthat the developing countries could meet allof their current and future energy needswith solar energy, using an area amountingto only 5 percent or less of the land nowbeing used for crops agriculture. Moreover,unused areas are often the best locations.

A hurdle still to be overcome is the costof storage. In the case of solar and windenergy, it may be necessary to produce

hydrogen through electrolysis and touse fuel cells or combustion to recon-vert the hydrogen back into usefulenergy. Alternatively, solar energy maybe stored in other ways: electrically byusing batteries (several advanced bat-tery technologies are under develop-ment but are still expensive);kinetically (using ultra-high-speed fly-

wheel devices with low-friction bearings);thermally (for example, using molten saltsor even metals or heated bricks);thermochemically (using the high tempera-tures of solar concentrators to create syn-thetic gases); hydraulically (by pumpingwater into reservoirs); by the storage ofcompressed air, which can later be used todrive turbines; or in the form of biomass.

All of these options are under activeresearch, and all are known to work. Forexample, the Weizmann Institute in Israelhas successfully demonstrated an approachin which a synthetic gas is created frommethane (CH4) and carbon dioxide (C02)using solar heat. The gas can be storeduntil needed, when it can be desynthesizedin the presence of a catalyst to give off thestored (solar) energy for use in electricitygeneration; the residuals are the originalconstituent gases (CH4 and C02). The cyclecan be repeated indefinitely.

These developments in solar energy andrelated storage technologies show muchcommercial promise. However, most of the



world's energy requirements willbe met by fossil fuels for someyears to come. If renewable energywere to become the primaryenergy source—for environmentalor commercial reasons, orboth—the transition would takedecades. In this period, the optionsfor using fossil fuels more effi-ciently and in less polluting wayswill be of much importance.

Declining costsAt the time the WDR 1992 was

published, the costs of the low-polluting practices examined inthe report were large in absoluteterms but quite small in relativeterms. For vehicle fuels and emis-sions controls, for example, theywere in the range of 5 to 15 centsper gallon of fuel used, includingthe annualized capital costs ofequipment divided by annual fuel con-sumption. The report showed that reduc-tion of particulate matter emissions andacid deposition from electricity generationwould increase supply costs by about 5-10percent, but that these costs could be fullyoffset by the gains in efficiency from thenew technologies or if gas were available.For the supply of fuelwood through agro-forestry, field studies consistently find thatfinancial rates of return to farmers whointroduce more sustainable practices are 15percent or more. With respect to the globalwarming problem, it seems that the costs ofthe noncarbon technologies will be far fromprohibitive and may even lead to a (pleas-ant) economic surprise because of technicaladvances in solar and solar-derived energy.In general, the costs of addressing environ-mental problems related to energy produc-tion and use have declined since 1992.

For electricity generation, estimates ofthe marginal costs of abating various pollu-tants are shown in Chart 2. The negativecosts on the left side of the curve representthe economic benefits that would arise frompursuing the "win-win" option of energyefficiency—mainly by eliminating subsi-dies. To the right are points representingthe marginal costs of turning to the low-polluting options. As can be seen, very highlevels of abatement can be achieved usingthe new technologies—probably at a nega-tive overall cost, once the economic benefitsof efficiency are taken into account.

PoliciesThe main barriers to pollution abatement

are policies and attitudes, not economics or

Chart 2Developing countries: marginal costs

of pollution abatement inelectricity production

the availability or costs of low-pollutingtechnologies. The good news is that ifappropriate policies are introduced, therewill be a response from business and con-sumers, as in the past. The main policyinstruments are familiar: environmentaltaxes on the main pollutants or sources ofpollution; environmental laws and regula-tions, traditionally the instruments mostfavored by governments; and, for local pol-lution, negotiated arrangements, backed bylocal laws and institutions, between the pol-luting and polluted parties.

Three other policy instruments tend toreceive less attention, but can be just asimportant in the long term as environmen-tal taxes and laws. First are public and pri-vate research and development (R&D)programs—all the developments describedabove would not have taken place withoutsuch programs. Second, technologies needto be demonstrated before they become full-blown marketable propositions, in the pre-sent case stimulated by environmentaltaxes or laws. In this respect, tax incentivesand investment grants in recognition of thepositive externalities of innovation have avaluable role to play. Third, investment ineducation and training is crucial. Most peo-ple in business and finance are familiarwith the technologies and practices alreadybeing used, and very often with some of theemerging options. However, significantinvestments are required to retrain staffand produce new generations of engineersand business leaders familiar with, andwilling to invest in, new technologies andpractices.

There is every reason to believe that

developing countries will be ableto increase their energy use in thefuture while greatly reducing pol-lution. Energy is an economic"good," not an economic "bad."Technologies and practices capa-ble of addressing the environmen-tal problems arising from energyuse are either already available orin development. What is neededis broader recognition of theirimmense potential for abatingpollution and supportive environ-mental policies based on eco-nomic principles. I F&D I

ReferencesDennis Anderson, 1991, "Energy and the

environment: an economic perspective on recenttechnological developments and policies," SpecialBriefing Paper No. 1 (Edinburgh: Wealth ofNations Foundation).

, 1994, "Cost-effectiveness in address-ing the CO2 Problem, with special reference tothe investments of the Global EnvironmentFacility," Annual Review of Energy and theEnvironment, Vol. 19, pp. 423-55.

Asian Development Bank, 1991, Environ-mental Considerations in Energy Development(Manila).

Robin W. Bates and Edwin A. Moore, 1992,"Commercial Energy Efficiency and theEnvironment" (unpublished background paperfor the World Bank's World DevelopmentReport 1992;.

J.B. Dooktte and William Magrath, eds.,1990, Watershed Development in Asia:Strategies and Technologies, World BankTechnical Paper No. 127 (Washington).

John S. Jennings, 1995, "Future SustainableEnergy Supply," Address to the Sixteenth WorldEnergy Council Congress, Tokyo (London: RoyalDutch Shell Group of Companies).

Organization for Economic Cooperation andDevelopment, 1989, Energy and theEnvironment: Policy Overview (Paris).

Kirk Smith, 1988, "Air Pollution: AssessingTotal Exposure in Developing Countries,"Environment, Vol. 30 (10), pp. 16-35.

World Bank, Industry and EnergyDepartment, 1996, Rural Energy and EconomicDevelopment (forthcoming).

World Energy Council, 1993, Energy forTomorrow's World (New York: St. Martin'sPress).


Source: Anderson (1994).


How Successful Are IMF-SupportedAdjustment Programs?

S U S A N S C H A D L E R

How successful have countriesbeen during adjustmentprograms? While most haveseen a quick turnaround intheir external accounts, asharper focus on medium-term sustainability, firmnominal anchors, and bettercoordination of fiscal andmonetary policy would helpto raise growth and stamp outinflation.

T HE 1980s were marked by severestrains in the international econ-omy. External financing slowed toa trickle for many countries, while

primary commodity prices dropped sharply.As a result, IMF lending rose to unprece-dented levels during the late 1980s andearly 1990s. Moreover, conditions in thecountries borrowing from the IMF were, byhistorical standards, unusually weak: fiscaland external imbalances were large; outputwas often falling; frequently, inflation washigh. In most countries, these weaknesseswere not simply cyclical but resulted fromdeep-seated structural distortions.

These circumstances posed considerablechallenges for the design of "conditional-ity"—the policies that a country agrees tofollow as a condition for borrowing fromthe IMF. Conditionality had to be pushed

beyond the traditional reliance on demandrestraint. For countries that were address-ing large debt overhangs or entrenchedstructural distortions rather than cyclicalweaknesses in the balance of payments,reliance on demand restraint alone for theneeded adjustment would have required anunacceptably deep depression of domesticdemand. Thus, supply-side policies aimedat bolstering growth—such as reducing therole of the government and opening theeconomy to external competition—becamean important part of conditionality. Also,conditionality began to focus more onthe sustainability of policies over themedium term.

These issues were examined in a 1995IMF study (see references) of the condi-tionality attached to 45 IMF lendingarrangements approved between mid-1988and mid-1991 for 36 middle-income coun-tries. The study concluded that the con-ceptual approach to designing adjustmentprograms was sound, that most countriesadhered reasonably well to their policyprograms, and that most aspects ofmacroeconomic performance improved.The most striking gains were on the exter-nal accounts; developments in the keydomestic targets—inflation, investment,and growth—were less impressive. Thesepatterns prompted an examination of waysto strengthen the design of programs.

The setting and the strategyEconomic conditions at the outset of the

arrangements were, almost without excep-tion, dire. Most countries had debilitatingunderlying problems—severe institutionalweaknesses, fiscal indiscipline, and weakexternal competitiveness. In these circum-

stances, many had been pushed to a crisisby a sudden outside disturbance, such as adeterioration in the terms of trade. Othershad seen a gradual deterioration in theexternal accounts, erosion of externalreserves, slow growth, inadequate savingsand investment, and rising inflation (seechart). Typically, a country's first responsewas to borrow abroad, without addressingunderlying problems. Countries usuallysought IMF support as a last resort.

Although the countries shared severeexternal financing constraints, initial condi-tions did vary. Almost half of the countrieshad had at least two recent borrowingarrangements with the IMF; for these coun-tries, trade imbalances had been narrowedin preceding arrangements, and remainingexternal problems largely reflected debtoverhangs. In countries that were justbeginning the adjustment process, externaldifficulties more frequently reflected anacute imbalance between what theyabsorbed and what they produced. In manyLatin American and Eastern Europeancountries, excess demand showed up inhigh inflation, but in many African, Asian,and Middle Eastern countries, inflationwas low. The Central European countriesbrought a new dimension to the traditionaldiversity of initial conditions: in general,their need for, and commitment to, struc-tural reform surpassed that in other coun-tries. At the same time, the collapse of theirprincipal export markets and a deteriora-tion in their terms of trade made theiradjustment unusually daunting.

The strategy for addressing macroeco-nomic imbalances had three prongs:(1) reining in domestic demand throughfiscal and credit restraint; (2) implementing

Susan Schadler,a US national, is Senior Advisor in the IMF's Policy Development and Review Department.



Initial conditions variedCountries with IMF arrangements and other developing countries 1

structural reforms to promote a supplyresponse and improve the efficiency ofresource use; and (3) securing externalfinancing to support the program (andoften to clear external arrears). While thisbroad approach was suitable for virtuallyall countries confronting unsustainableexternal positions, the diversity of specificcircumstances meant that the degree of

reliance on each prong needed to be attunedto each country's circumstances.

Program developmentsIn most countries, macroeconomic per-

formance improved during the arrange-ment and, in many, continued to do soafterward. Improvements were large in theexternal sector. In countries that started

their arrangements in the midst of acutebalance of payments crises (mostly thecountries that had not had recent arrange-ments with the IMF), a marked improve-ment occurred quickly. In countries withless acute difficulties, underlying problemswere addressed. Official reserves rose tomore comfortable levels; about half of thecountries that had had external arrearscleared them; some countries (particularlythose that had had recent arrangements)benefited from large increases in capitalinflows; and current accounts convergedtoward positions that could be financed bynormal capital inflows.

There were disappointments, however,even in external developments—particu-larly in sub-Saharan Africa and CentralEurope, capital inflows failed to material-ize as projected; several countries did notfully resolve their debt overhangs andarrears; and, in most countries, strongexport growth early in the program fadedtoward the end of the period reviewed.

Developments in the domestic economywere less impressive. A few countries (theformer Czechoslovakia, Mexico, andPoland) sustained dramatic reductions ofinflation from very high initial rates, butmany continued to experience the moder-ately high inflation rates they started with;a few (Algeria, Hungary, Jamaica, andRomania) even saw inflation accelerate. Formost countries outside Central Europe,there was some strengthening of growthand, on average, an increase in savingsratios. Still, no country shifted to a dis-tinctly more rapid pace of growth backedby higher savings. Most worrisome wasthat few countries saw an increase in totalinvestment as a share of GDP. On thebrighter side, however, private investmentrates rose, on average, as public investmentrates fell, and the productivity of invest-ment—as measured by the incrementalcapital-output ratio—rose.

Policy focusIn light of these developments, the

review focused on the characteristics ofpolicies supported by the arrangementsand how they could be improved. Thereview examined whether policies weretailored to the particular needs of eacheconomy, whether there was too muchemphasis on reaching external objectivesor a subordination of domestic objectives,and whether there was adequate focus onthe sustainability of policies over themedium term.

Government budgets. Fiscal adjust-ment strategies were tailored, in several


Sources: IMF Executive Board documents; and World Economic Outlook, various issues.1 Figures for other developing countries are a weighted average of calendar year data with weights

corresponding to the share of each calendar year in t-1, t-2, t-3 of countries with arrangements, excludingCentral Europe.

2 Ratio to exports of goods and services before debt rescheduling.3 Argentina, Brazil, Costa Rica, Cote d'lvoire, Ecuador, Haiti, Jamaica, Madagascar, Mali, Mexico,

Morocco, the Philippines, Uruguay, Yugoslavia, and Zaire.4 Algeria, the Congo, Egypt, Gabon, Nigeria, and Tunisia.5 Bulgaria, the former Czechoslovakia, Hungary, Poland, and Romania.6 Cameroon, El Salvador, Guatemala, Honduras, Jordan, Pakistan, Papua New Guinea, Trinidad and

Tobago, and Venezuela.


crucial respects, to the specific nature ofinitial imbalances. Fiscal adjustment, onaverage, was targeted to come aboutequally from revenue increases and expen-diture restraint. Planned expenditure cuts,however, were largest in the countrieswhere the ratio of expenditure to GDP wasthe highest, while the opposite was true forrevenue. Targeted changes in overalldeficits were greatest in the countries withthe largest initial imbalances.

In fact, fiscal adjustment was, on aver-age, close to that targeted. With availabledata, however, and the short period consid-ered, it is impossible to account preciselyfor the separate roles of discretionary pol-icy change, exogenous influences, andcyclical factors. On average, spendingrestraint was somewhat greater thanplanned, and revenue increases wereslightly short of target. Proportionately,restraint of capital spending was greaterthan that of current spending.

These observations suggest that fiscaladjustment strategies were sensitive tocountries' initial conditions and policieswere largely implemented as planned.However, programs were often not pre-sented in the context of an explicit medium-term analysis of the potential for revenuemobilization, expenditure needs, andfinancing. It was, therefore, not possibleto answer one critical question: howmuch of the gap between actual and sus-tainable fiscal deficits was closed over theprogram period?

One recommendation of the study is thatprograms should be based more explicitlyon medium-term fiscal frameworks thatconsider the medium-term targets formoney creation and inflation, the intertem-poral dynamics of the government debtratio, and the impact of fiscal adjustmenton the savings-investment balance of theeconomy. Also, where data permit, pro-grams should take fuller account of theeffect of current decisions on future fiscaltrends or prospects for investment andgrowth—for example, where protectingcapital spending is important, where socialexpenditures have borne a large share ofcuts, or where the efficiency of revenuemeasures is questionable.

Financial programs. The studyfound that targets for money supplygrowth were exceeded in two-thirds of theprogram years. About as often, inflationtargets were missed by sizable margins.Most missed money targets did not reflecthigher-than-targeted credit growth, butrather higher-than-targeted inflows of for-eign exchange.

This echoes the earlier observation thatprograms more consistently achievedexternal objectives than inflation targets.The typical structure of financial pro-grams—ceilings on the net domestic assetsof the monetary authority or banking sys-tem and floors on the net internationalreserves of the monetary authority—ishighly effective in achieving reserve targetsbut inherently less so in restrainingmoney creation and reducing inflation.When inflation expectations are entrenchedand nominal exchange rate appreciation isresisted, capital inflows can accommodatehigher-than-targeted inflation. To helpredress this limitation, programs shouldspecify benchmarks for an appropriatemonetary aggregate. Deviations from suchbenchmarks would signal dangers for infla-tion targets and the need for prompt policyresponses, guided by the causes of the devi-ations and by the short-term trade-offsbetween building reserves, controllinginflation, and preserving competitiveness.

Financial market reform. Financialmarket liberalization, particularly the de-control of interest rates and the introduc-tion of auctions for government paper,progressed quite remarkably. As a result,real interest rates often rose from negativeto positive levels—sometimes even to levelsthat could be considered excessive. In gen-eral, high real rates fell back to moderatelylow levels within one or two years.

The study considered ways to avoid highreal interest rates even for short periods.Essentially, these involve ensuring that, asfinancial market liberalization and shifts infiscal financing from bank to market bor-rowing create a market-clearing role forinterest rates, the persistence of other dis-tortions does not put excessive pressure oninterest rates. Two issues are particularlyimportant. First, programs need to focus onimproving prudential controls, bank super-vision, and the competitive structure offinancial markets to minimize oligopolisticbehavior and reduce pressure on interestrates from adverse selection by banks.

Second, programs must ensure adequatefiscal adjustment to avoid excessive re-liance on tight credit for achieving inflationtargets. Judging the appropriate mix offinancial policies—even ex post—is diffi-cult. Typical indicators of insufficient fiscaladjustment and excessive reliance on creditrestraint—persistently high real interestrates, large capital inflows, or weak privateinvestment—can occur for reasons unre-lated to excessive credit restraint. Few ofthe countries reviewed showed clear evi-dence of excessive reliance on credit

restraint. Nevertheless, further work on thedevelopment of indicators is needed.

Exchange rate as an anchor.Almost half of the countries reviewed usedthe exchange rate as a nominal anchor—that is, a fixed or pre-announced path forthe nominal exchange rate. These were usu-ally countries with inflation rates that wereinitially either very low or very high. Bycontrast, most countries with moderatelyhigh initial inflation (annual rates of 20-50percent) chose to manage their exchangerates more flexibly. These differencesreflected the choice between two possibleshort-term roles for the exchange rate—onthe one hand, anchoring inflation expecta-tions and disciplining policies, and, on theother, maintaining or improving competi-tiveness. In general, the countries that hadan exchange rate anchor reduced inflationor maintained it at low levels. The cost wastypically a loss of price competitiveness asinflation fell or stayed low, but was stillabove that in trading partners. In two coun-tries—Argentina (1989) and Yugoslavia(1988)—an exchange rate peg was not ade-quately supported by other policies andproved unsustainable. Countries that chosea more flexible policy tended to see littledecline or even increases in inflation, but,on average, saw some improvement incompetitiveness.

The experience confirms the results ofother studies that exchange rate anchorsare effective in reducing high, or sustaininglow, inflation when supported by macroeco-nomic discipline and when initial exchangerates are not misaligned. In a first-bestworld, where financial policies are suffi-ciently restrictive, there would be clearrewards to using the exchange rate as ananchor for a period of at least a few monthsto a year. In fact, policies are often notstrong enough and a short-term trade-offbetween inflation and competitivenessarises; often the choice is dictated by theexternal constraint. The trade-off needs tobe kept under review, especially in the faceof adverse movements in the terms of tradeor protracted loss of competitiveness.

When the exchange rate is used moreflexibly, limits to the effectiveness of nomi-nal depreciation to achieve real deprecia-tions must be recognized. Real exchangerate rules—targets for the real exchangerate to guide nominal exchange rate pol-icy—should be avoided. Decisions onwhether to adjust nominal rates shouldtake into account the prospects for neededsupport from fiscal policy and wages.

Structural reforms. All countriespursued structural reforms—typically



fastest and first in exchange, trade, andfinancial systems, and, in Central Europe,price liberalization. Reform of taxes, publicspending, and, particularly, public enter-prises and labor markets was slower. Therewas no evident pattern in the sequencing ofreforms, which appears to have been afunction more of political opportunities andadministrative capabilities than optimalsequencing considerations. In some casesthis may have led to problems, but theywere not especially serious. In general, theanalytical basis for comparing differentsequences for removing distortions isweak, and insisting on specific patterns ofsequencing may often have slowed or evenobstructed structural reform owing todelays in taking politically difficult actions.

The experience suggests that, exceptwhen there are clear risks in a specificpattern of reform, it is best to move asrapidly as possible with reforms when theyare technically and politically feasibleand to channel any resulting tensions intoa consensus for further change. Neverthe-less, the costs of sequences of reforms thatmay prove particularly risky—such asfinancial market liberalization before ade-quate prudential controls and bank super-vision are in place—need to be recognizedand avoided.

Labor markets. Although widespread,labor market rigidities were addressed onlyto a limited degree. In many countries, realwages in the organized sector were not veryresponsive to labor market conditions.Backward-looking indexation (de facto orexplicit) existed in many countries, espe-cially those with high and intermediate lev-els of inflation. Modifying this practice wasessential to lowering inflation, reducing realwages to market-clearing levels, and gird-ing economies against disruptions fromnonfinancial shocks. In most countries, theproblem was recognized, but a political con-sensus for change could not be secured.Poor data and political sensitivities ham-pered an evaluation of the degree to whichlabor market and wage-setting practicesconstrained job creation, investment, andgrowth. Greater efforts are needed to collectwage and employment data, and, wherepossible, to apply analytical techniquescommon in industrial countries for judgingwhether real wages are excessive.

Saving, investment, growthThe study looked at possible causes of

the rather weak development of savingsduring programs. The econometric evi-dence bears out the conclusions of otherstudies that the channels for policies to

influence saving rates directly were limited.Changes in government savings had adirect effect on aggregate savings,although, on average, about half of anysuch change was offset by an oppositemovement in private savings. Othersystematic determinants of private sav-ings—actual and expected growth ratesand the terms of trade—were beyond thedirect control of policies. The influencesthat were controlled most directly by poli-cies—changes in real interest rates andexchange rates, financial reform, and pen-sion reform—proved to have weak effects.Indeed the success of programs in attract-ing capital inflows and raising expectationsabout future growth may, in the short run,even depress private saving ratios.

There are several possible reasons forthe sluggishness of investment andgrowth—slow growth of final demand,deficient savings (domestic or foreign) tofinance investment, restraint on govern-ment investment, and the need to establisha track record to gain investors' confidence.A first perusal of the evidence suggeststhat the last of these—a wait-and-see ten-dency among investors—is likely to havebeen the strongest explanatory factor. Finaldemand generally picked up during theprograms; countries that experienced largeincreases in the availability of foreign sav-ings did not systematically see better out-comes for investment than other countries;and many of the countries that squeezedgovernment investment the most saw thelargest rebounds in private investment. Theexploration of these issues clearly neededto be more exhaustive than was possible inthe study, and follow-up studies wereundertaken on the responses of investmentand growth to adjustment policies.

Were strategies right?In broad terms, the study concluded that

the basic three-pronged approach wasappropriate. The approach did place astrong emphasis on achieving balance ofpayments objectives, but this was inevitablein light of limited access to externalresources and the generally dire balance ofpayments problems that brought countriesto the IMF for support. Most programsappear to have pursued these objectiveswith sensitivity for the ultimate goals of eco-nomic policy—to improve living standardsthrough higher growth. Fostering sustain-able growth is invariably a protracted pro-cess requiring steadfast adherence tosensible policies over a long period. Thus,most programs aimed to secure immediateimprovements in external finances as part

of the process of establishing conditions forgreater efficiency, saving, investment, andgrowth over the longer term.

Still, the study did identify several areaswhere program design could be strength-ened. In general, programs would benefit,data permitting, from comprehensivemedium-term scenarios that spell out con-sistent medium-term prospects for savings,investment, intertemporal debt dynamics,and the current account. This would helpdetermine the adequacy of fiscal adjust-ment, realistic investment and growthgoals, and the need for structural reform tocultivate an environment attractive toinvestment. With respect to financial poli-cies, episodes of very high real interestrates underscore the importance of bankreform and adequate fiscal adjustment soas to avoid excessive reliance on creditrestraint to rein in demand and containinflation. The increasing number of coun-tries for which disinflation is an importantobjective means that more attention needsto be given to establishing a nominalanchor in each program. In many circum-stances, the exchange rate is best suited forthis role, although firm money targets areuseful when uncertainties about the equilib-rium real exchange rate exist. Finally, pro-grams need to extend the good record ofstructural change in exchange, trade, andfinancial systems to labor markets, taxpolicy, and public enterprise finances. iF&Dl

References:Susan Schadler, Franek Rozwadowski,

Siddharth Tiwari, and David O. Robinson,Economic Adjustment in Low-Income Countries:Experience Under the Enhanced StructuralAdjustment Facility, IMF Occasional PaperNo. 106, Washington (1993).

Susan Schadler, Adam Bennett, MariaCarkovic, Louis Dicks-Mireaux, Mauro Mecagni,James Morsink, and Miguel Savastano, IMFConditionality Review: Experience Under Stand-By and Extended Arrangements: Part I: KeyIssues and Findings, IMF Occasional PaperNo. 128, Washington (1995).

, IMF Conditionality Review:Experience Under Stand-By and ExtendedArrangements, Part II: Background Papers,IMF Occasional Paper No. 129, Washington(1995).



The Response of Investment and Growth toAdjustment Policies

E R I K C . O F F E R D A L

The basic pillars of economicgrowth—investment and sav-ing—are strongly influencedby the private sector's confi-dence in a country's policies.The timeliness, sustainability,and consistency of policies aretherefore critically important.

A RECENT IMF study asked howadjustment policies could bettercontribute to reinvigoratingmedium-term growth in develop-

ing countries. To answer this question, thestudy examined the influence of macroeco-nomic policies and core structural reformson the investment, saving, and growth per-formances of eight developing countries(Bangladesh, Chile, Ghana, India, Mexico,Morocco, Senegal, and Thailand) since1970. The impetus for this study came froma review of IMF-supported adjustment pro-grams. (See the article by Susan Schadlerin this issue.)

Most of the eight countries sufferedadverse external shocks—higher oil prices,falling commodity export prices, and risingworld interest rates—during the latter halfof the 1970s and the early 1980s. Their ini-tial responses to these shocks typicallyincluded expansionary policies, mainly therunning up of large fiscal deficits, some-times combined with new or intensified

import restrictions and exchange controls.These policies exacerbated, rather thanrelieved, an already difficult situation byfueling inflation or increasing external debtburdens, or both. In many cases, the resultwas a macroeconomic crisis, frequentlyinvolving a severe external financing con-straint that entailed an abrupt shift in netresource flows. The subsequent adjustmentprogram typically had to begin with effortsto restore macroeconomic stability. Theadjustment effort started in the early 1980sand continued up to the present time for allcountries except Chile and India.

The short-term declines in output, in-vestment, and measured productivityresulting from the crisis and early adjust-ment were especially dramatic in Chile(both in 1975 and 1982) and Mexico (in1982-83). Most other countries experienceda more moderate slowdown in growth.Private investment as a share of GDP typi-cally fell for several years, while cuts inpublic investment were part of the fiscalconsolidation. An exception was Ghana,where growth picked up quickly as a severeimport compression was reversed in 1983.

In the medium term, Chile, Thailand,and—to a lesser extent—Ghana succeededin achieving sustained growth rates thatwere higher than those prevailing beforethe implementation of adjustment policies.Both Chile and Thailand also achieved amarked and sustained increase in privateinvestment, supported by higher domesticsavings and capital inflows. In many of theother countries—especially Bangladesh,Mexico, Morocco, and Senegal—therecorded gains in output growth were mo-

dest at best, and in some (India, Mexico, andMorocco) the recovery in private investmentoccurred only after a "pause" of a few years.

Economic stabilityOne approach to answering the question

of whether economic stability matters forgrowth is to compare growth rates for eachof the eight countries with world averagegrowth rates after taking into accountthe effects on growth of key, longer-terminfluences (such as investment and terms-of-trade movements) and of macroeconomicpolicies. This amounts to comparing eachcountry's performance with a control groupthat is modified to take account of thesebroader influences on growth. Of course,all such comparisons involve potentialpitfalls, since it is difficult to predict whatwould have happened in the absence ofadjustment.

This comparison indicates that countriesexperiencing episodes of severe macroeco-nomic instability—Chile in the early 1970s,Ghana prior to 1983, and Mexico in theaftermath of the 1980s debt crisis—typi-cally had growth rates well below the worldaverage during those episodes. Moreover,following implementation of their adjust-ment programs and restoration of macro-economic stability, both Chile and Ghanaachieved growth rates well above the worldaverage. Thus, after controlling for otherinfluences on growth, policies orientedtoward supporting macroeconomic stabil-ity do appear to be beneficial to rapidrecovery of output.

However, there is more to the storythan stability. In Mexico, on the one hand,

Erik C. Offerdal,a Norwegian national, is a Senior Economist in the Policy Review Division of the IMF's Policy Development and Review Department.

18 Finance & Development /'June 1996


significant stabilization was achieved in theaftermath of the debt crisis, yet recordedgrowth rates have barely recovered to theworld average. Thailand, on the other hand,enjoyed prolonged macroeconomic stability,yet it achieved growth rates above theworld average both before and after adjust-ment. Both experiences suggest importantadditional roles for confidence in a coun-try's policies and for structural factors.

Growth-fostering policiesThree aspects of policies—timeliness,

sustainability, and consistency—are crucialfor investor confidence. An inappropriatedesign of adjustment policies may prolongor contribute to macroeconomic instabilityand uncertainty, and thus sap private sectorconfidence.

Delaying adjustment until the point ofcrisis will generally result in a longerand more protracted slowdown ofoutput growth for three principal rea-sons. First, the rapid contractionin domestic demand will leave littletime for resources to be reallocated tocushion the impact on output.Second, policies undertaken in acrisis may well include ad hocmeasures that are not conducive toinvestment and growth. Evidencesuggests that, at least initially, fiscaladjustment in many of the countriesdiscussed was not based on growth-oriented changes in tax and expendituresystems. (See the article by George A.Mackenzie and David W.H. Orsmond in thisissue.) Third, in a crisis environment, pri-vate sector confidence weakens, which, inturn, is likely to reduce the effectiveness ofadjustment policies. For example, a tighten-ing of fiscal policy should be partly offsetby "crowding in" of private investment,since the reduced demand for credit by thepublic sector would entail lower interestrates and more credit being available to theprivate sector. However, expectations ofpolicy reversals may induce investors totake a "wait-and-see" attitude, therebystalling growth.

Progress toward sustainability of eco-nomic policies can foster private sector con-fidence by reducing concerns about policyreversals. One limited yet useful criterionfor gauging sustainability is a fiscal bal-ance that will maintain a constant ratio ofpublic debt to GDP, in the context of lowinflation and market-determined interestrates. Judged by this criterion, Thailand'sdeficit was broadly sustainable from theoutset of adjustment. This perception prob-ably contributed to the relatively rapid

increase in private investment. The otherseven countries all made progress toward asustainable fiscal position. However, in sev-eral of them (notably Morocco), such a posi-tion was reached only after several years;and in others (Ghana and India), theimprovement was later partly reversed,probably delaying the "crowding in" of pri-vate investment. Moreover, the initiallyslow response of private investment to astrong fiscal adjustment, especially inMexico, is probably closely linked to thedebt crisis. The sudden loss of externalfinancing meant that even a sharplyreduced deficit had to be financed largelyfrom domestic sources, which put consider-able upward pressure on interest rates,thereby squeezing private investment.

Consistency between the elements ofmacroeconomic policies is crucial. The mix

"Three aspects ofpolicies—timeliness, sustain-

ability, and consistency—are crucial for investor

confidence."

of fiscal and monetary policy must includestrong enough fiscal adjustment to mini-mize upward pressure on interest ratesand increase private sector access to bankcredit. Moreover, the policy mix shouldbe kept under review to ensure its consis-tency with the overall macroeconomicobjectives, in view of the unpredictableshifts in private sector saving and invest-ment behavior at a time of substantialchanges in macroeconomic and structuralpolicies. In Mexico, even though fiscalpolicy was sustainable when gauged by theabove-mentioned criterion, the externalposition proved to be untenable owing to anunexpectedly sharp fall in private saving.The coordination of fiscal, exchange rate,and wage policies is also essential toestablishing policy credibility and strength-ening external competitiveness. In bothChile and Thailand (after 1984), deprecia-tions of the nominal exchange rate led tosignificant gains in external competitive-ness as fiscal consolidation held domesticprice pressures in check. These improve-ments were supported by the suspensionof compulsory wage indexation in Chile(in 1982) and, in Thailand, by the flexibilityof the labor market, which facilitated thenecessary adjustments in real wages.

Why an investment pause?Most of the eight countries experienced a

considerable decline in private investment,either before or during the early part oftheir adjustment programs, amounting to3-4 percentage points of GDP. Moreover, insome countries (Mexico, Morocco, and, to alesser extent, India) private investmentremained at this more depressed level for2-4 years before beginning a recovery. Inother countries (Bangladesh, Ghana, andSenegal), the recovery started earlier butwas weak and uneven.

An examination of the behavior of invest-ment in the eight countries indicates thatboth the credibility of policies and thedegree of macroeconomic instability anduncertainty are important influences on theforward-looking decisions of privateinvestors. If investors are concerned that a

market environment is highly uncer-tain, they may delay investing untila more positive assessment can bereached. This may lock the economyinto a low-investment, low-growthequilibrium if a sufficient number offirms postpone investments. In con-trast, policies perceived as consis-tent and unlikely to be reversed mayturn expectations around andinduce private investors to go aheadwith their plans, thereby reducing

the costs of adjustment.These influences appear to have been

especially important in Ghana, where boththe lingering impact of foreign exchangecontrols and slippages in financial policiesinduced the 1992-93 investment decline; inMexico, where persistent high inflation wasan important deterrent to real privateinvestment during 1983-87; and in Senegal,where inflation and rising external debtcontributed to holding back investment inthe early 1980s.

Economic policies were also found toinfluence investment through several otherchannels: changes in the level of economicactivity (capturing the contractionaryimpact of adjustment policies and confi-dence effects), as well as real interest ratesand the supply of credit to the private sec-tor (reflecting the effects of the mix of fiscaland monetary policies and financial sectorreforms). In six countries, public invest-ment was found to compete with and crowdout private investment, implying that selec-tive rationalization of public investmentwould help to promote private investment.However, this is an issue where generaliza-tions are difficult, since investment in somecomponents of public infrastructure maywell encourage private investment.



Why a slow saving response?Only Chile and Thailand achieved large

and lasting increases in their saving rates.By contrast, Bangladesh, Ghana, Morocco,and Senegal achieved moderate increasesin saving rates during adjustment, al-though Ghana's gain was short-lived.

In most countries, the increase in savingresulted primarily from higher public sav-ing, although this was partly offset bydecreased private saving. This offset, typi-cally on the order of 50 percent, suggeststhe need to be cautious about projections ofa recovery in overall saving, especially inthe short term. Also, when public saving israised at the expense of current expendi-tures, such as primary education—thatshould more correctly be measured asinvestment in human capital—it may harmgrowth prospects. The impact of otherpolicy-related factors, including real inter-est rates, were, however, found to be small.Large terms-of-trade fluctuations and sup-ply shocks also played a substantial role inGhana, Morocco, and Senegal.

On the one hand, strong fiscal adjust-ment contributed to the saving perfor-mances of Chile and Thailand. Moreover,the offset between public and private sav-ing was moderated by a history of macro-economic stability and a fall in thedependency ratio (the share of the old andthe young in the total working-age popula-tion) in Thailand, and public sector reforms(of the tax, public enterprise, and pensionsystems) in Chile. Mexico, on the otherhand, has recorded a marked decline innational saving, resulting from a boom inconsumer spending. This boom was fueledby concerns about policy reversals, easedliquidity constraints, and expectations ofhigher real incomes.

Structural reformMacroeconomic adjustment efforts must

be accompanied by structural reformsaimed at improving the efficiency ofresource allocation and productivity inorder to permanently increase a country'sgrowth rate.

Those countries that began withrelatively small structural distortions(Thailand) or made significant progresstoward eliminating major distortions (Chileand Ghana) experienced the greatest pro-ductivity gains. In contrast, Senegal's slowand faltering progress in structural reformappears to have yielded few productivitygains. The modest gains in measured pro-ductivity developments in Mexico andMorocco cannot be explained solely by theextent of their structural distortions; both

countries had only moderate distortions tobegin with and both made major progressin carrying out several aspects of reform.Some observers have noted that growth innewly emerging sectors of the Mexicaneconomy may not have been fullyrecorded. Moreover, important weaknessesin Mexico's financial and agricultural sec-tors either remain or have been addressedtoo recently for the reforms to have hadtheir full impact. The same is true ofMorocco's financial sector and the develop-ment of its human capital.

There is also evidence that certain clus-ters of coordinated reforms can provide acritical mass that will enhance efficiencygains. The linkages among reforms can,however, vary from country to country, andstructural reforms will also be more effec-tive if implemented along with restorationof macroeconomic stability. Three clustersof reforms stand out. First, the economy'ssupply response to trade reforms is influ-enced by the strength of supportingsectoral reforms, such as reducing the dom-inance of the public sector (Bangladesh,Ghana, and India); building a legal infra-structure to accompany reductions inindustrial licensing requirements (India);and increasing labor market flexibility(notably Chile and India). Second, it is criti-cally important that bank supervision andregulation be strengthened at the same timethat the financial sector is liberalized.Otherwise, a serious banking crisis mayresult, as in Chile during the early 1980s,when its banking regulations were tooweak to prevent rapid growth in the vol-ume of nonperforming loans. Third, inade-quate public enterprise reforms can be amajor obstacle to the development of effi-cient financial intermediation. Decisionsto support a weak public enterprise sec-tor—as were made, for example, inBangladesh, Ghana, India, and Senegal—have resulted in large intermediation costsand high lending rates, which, in turn, riskdampening the private investmentresponse.

Key lessonsThe linkages between a country's poli-

cies and its economic growth are often indi-rect and can operate with significant andvariable lags. Moreover, many factors otherthan policies have an impact on growth.Nevertheless, macroeconomic and struc-tural policies clearly matter. How can thedesign and implementation of adjustmentpolicies help foster growth? The differentexperiences of two countries that appear tohave achieved higher sustained growth

rates—Chile and Thailand—suggest thatno single blueprint exists. Still, two centralthemes emerge despite the difficulties ofgeneralizing from a small group. First, mar-ket assessments of the internal consistencyand sustainability of policies greatly influ-ence the size and speed of the response toforward-looking decisions such as privateinvestment and saving. Second, in eachcountry, there appear to be close linksbetween particular aspects of macroeco-nomic and structural reforms that are likelyto be mutually supporting; hence the impor-tance of a critical mass. Several specificlessons for the design of adjustment poli-cies stem from these propositions:

• Delayed adjustment is costly. Sta-bilization policies implemented in a macro-economic crisis will generally have deepercontractionary effects than those imple-mented in a more timely fashion.

• A forward-looking, medium-termframework is essential, in order to ensurethe sustainability and consistency of poli-cies that can prevent an economy frombeing locked into a low-investment, low-growth equilibrium.

• Fiscal adjustment should be strongenough both to minimize any adverseeffects on private investment and to sup-port real exchange rate adjustment to pro-mote resource switching and therebyminimize the initial contraction of output.

• Increasing public saving is likely to bethe most effective means of raising nationalsaving in the short run, although a par-tially offsetting decrease in private savingshould typically be expected.

• Structural reforms need an early start.In each country discussed, there werestrong indications that certain types ofstructural reform tend to reinforce eachother, suggesting that carefully combiningmutually supportive reforms is likely tomaximize their beneficial impact ongrowth. Moreover, insufficient emphasis on,or delays in, implementing sectoral reformmeasures can dampen the economy's sup-ply response to macroeconomic reforms.

This article summarizes the principal findingsof "Reinvigorating Growth in DevelopingCountries: Lessons from Adjustment Policies inEight Economies," by David Goldsbrough, LouisDicks-Mireaux, Sharmini Coorey, MauroMecagni, Balazs Horvath, Kalpana Kochhar,Erik Offerdal, andjianping Zhou, which will bepublished shortly in the IMF's Occasional Paper


serices


The Quality of Fiscal Adjustment and GrowthGEORGE A. MACKENZIE AND DAVID W.H. ORSMOND

"^~~~~~^~^ Growth-friendly policies duces reliable, comprehensible, transparent,Fiscal adjustment is usually The magnitude of the change in the fis- and timely data: and the use of project., f. ., , . cal balance is relatively easy to measure, appraisal techniques where appropriate.tne first Step on me roaa to Thg game camot bg gaid Qf the oyerall These activities must be financed by a

macroeconomic Stability. But quality of fiscal adjustment—it is not easy tax system that does not impede growth.*u«taitu>H0rniutl, veauire* that to tdl exactly what combination of tax and While all—or virtually all—taxes are dis-SUStainea growm requires mat expenditure reforms begt promotes growth torting and entail some loss of welfare, aOS much attention be paid to Nevertheless, as a general rule, public sec- growth-promoting tax system is one thatthe wav the government's tor activities wil1 best achieve that end exercises the least distorting effects on

. . . . when they complement, rather than com- work effort and the quantity and allocationdeficit is cut as to the size of pe^e with, what the private sector does. °f investment and saving, and minimizesthe cut They would include: reliance on scarce administrative resources

• Primary education, basic health care, and demands on taxpayers' compliance.and basic infrastructure, which are activi- Such a system would have the following

___ ties with a high social rate of return that are characteristics:fT^HE PRINCIPAL objective of typically not supplied in adequate quantity • A company income tax with a uniform

I adjustment programs is to foster by the private sector; and moderate rate, and a personal incomeI sustained economic growth. • Basic administrative and regulatory tax with a moderate top rate and few

^L Because countries beginning services to protect private property and exemptions;adjustment usually suffer from a combina- promote a stable, predictable climate for • A general sales tax, like the value-tion of a balance of payments crisis and entrepreneurial activity. This requires a added tax (VAT), preferably with one ratedomestic financial instability, and because well-functioning and adequately remuner- and minimal exemptions;of the key role fiscal policy plays in elimi- ated civil service; • A customs tariff with as low an aver-nating these disequilibria, a substantial • A cost-effective basic social safety net age rate as possible, and limited dispersionreduction in the size of the budget deficit is to foster acceptance of adjustment proof rates;typically at the core of adjustment programs when these are necessary and 'Export taxes only as a proxy forgrams. However, the way the deficit is increase the human capital of the poor; and income taxes in hard-to-tax sectors; andreduced—the quality of fiscal adjust- • Efficient and effective public expendi- • An efficient tax administration thatment—can also affect growth, since some ture management, without which even the encourages "voluntary compliance" (e.g.,components of government expenditure are best-designed expenditure policy will not by designing simple forms), effectivelymore productive than others, some tax be well executed. This entails the use of a monitors tax payments, promptly followsreforms more beneficial in their impact on consistent macroeconomic framework in up on late payment or nonpayment, andresource allocation than others, and some preparing the budget; an effective ceiling discourages evasion and fraud by means oftypes of fiscal adjustment more sustainable and control mechanism on overall expendi- an adequate program of audits and anthan others. ture; a budget reporting system that pro- appropriate penalty system.

George A. (Sandy) Mackenzie, David W.H. Orsmond,a Canadian national, is Chief of the Fiscal Operations Division II of the an Australian national, is an Economist in the Fiscal OperationsIMF's Fiscal Affairs Department. Division II of the IMF's Fiscal Affairs Department.

Finance & Development / June 1996 21


Adjustment in practiceAlthough we know roughly what a

growth-promoting public sector should looklike, knowing the desired destination of eco-nomic policy cannot give us a road mapthat a country can use to get there. Toimprove understanding of this issue, arecent IMF study analyzed the nature ofthe fiscal adjustment and public sectorreform efforts of eight countries—Bangladesh, Chile, Ghana, India, Mexico,Morocco, Senegal, and Thailand. For mostcountries, an observation period of about 15years—usually from 1978 to 1993—wasdivided into three subperiods; that is,a pre-adjustment period and, for ana-lytical purposes, two subsequentadjustment periods. (There was onlyone adjustment period in India.)

Initial adjustment efforts. Allthe sample countries reduced theirprimary deficits (their overall deficitsexcluding interest payments) over thewhole period by an average of 6.5 per-cent of GDP, ranging from 1.1 percentof GDP in Ghana to 10.6 percent inMorocco. But most of the declineoccurred during the first adjustmentperiod, reflecting the urgency of restoringmacroeconomic stability. The adjustment inthe first period was achieved mainly by areducion of non-interest expenditure (3.6percentage points of the 4.6 percent declinein the primary deficit occurred during thefirst adjustment period). In the secondperiod there was a more balanced mix ofexpenditure reductions and revenueincreases (excluding the large adjustmentsin Chile).

Non-interest expenditure bore the bruntof the early adjustment because there wasonly limited scope for raising revenue. Taxrates were already high and probably quitedistorting, and the scope for broadeningthe base of existing taxes was limited by,among other things, the difficulty of taxingagriculture and services, retroactively re-scinding tax incentive schemes, and theincreased cascading that broadening thebase of the existing sales tax would haveentailed.

This left no alternative but fundamentalreform—introducing new and morebroadly based taxes, like the VAT. Mexicoand Senegal had already introduced VATswith multiple rates prior to their firstadjustment periods, though the revenueyields were relatively limited; in contrast,Chile, which had a relatively efficient taxadministration, successfully introduced asingle-rate VAT early in the first adjust-ment period. For reasons that included the

extensive administrative preparations nec-essary for its introduction, none of the fiveremaining countries introduced a VAT dur-ing the first adjustment period. Of thesecountries, only Ghana substantially in-creased revenue, and this was mainlyattributable to the effect of the exchangerate devaluation on the effective tax rateson existing major tax bases.

Not only were the initial adjustments pri-marily focused on non-interest expenditure,but the cutbacks were not spread equally.The scope for quick but substantial andlasting reductions in the civil service wage

".. . it is not easy to tellexactly what combination oftax and expenditure reforms

best promotes growth."

bill and in subsidies and transfers provedlimited. This left two vulnerable forms ofexpenditure that were cut in virtually allcountries to enable them to realize budgetsavings quickly: other current goods andservices, and capital expenditure. Reduc-tions in capital spending that cut sharplyinto productive capital expenditure wouldhave eventually reduced medium-termgrowth prospects if they had not later beenreversed. Similarly, over time, underspend-ing on operation and maintenance wouldsharply reduce the productivity of the pub-lic capital stock, while cuts in expendituresfor other goods and services would con-tribute to a growing imbalance betweenlabor and materials and supplies.

The study was mindful of this initial fis-cal rigidity, since it was experienced tosome extent by all eight countries. Thus,each country's performance was evaluatednot simply from the point of view of howclose it came to the growth-fostering idealbut also on the basis of the direction, depth,and sequencing of its reforms. One basicquestion the study posed was how quicklya country undertook reforms that enabled itto accomplish some reversal of the initialcuts in expenditure on capital and othergoods and services. More generally, coun-tries with the best adjustment profileswould be those that

• reallocated expenditure prioritiestoward, or at least preserved, the most pro-ductive areas;

• minimized reliance on indiscriminateexpenditure cuts, which would typicallyprove to be temporary;

• took early action to reform the tax sys-tem;

• strengthened expenditure and taxadministration, so as to improve the effi-ciency of expenditure, increase the taxyield over time, and allow for the adoptionof a less distortive tax system; and

• reformed public enterprises to increasetheir operational efficiency and eliminateuneconomic subsidies.

Later adjustment efforts. While alleight countries adopted policies alongthese lines, there were distinct differ-ences within the group.

After making disproportionatelylarge cuts in capital expenditure andoperations and maintenance in theinitial adjustment period, almostevery country increased governmentinvestment in the second period, inreaction to growing bottlenecks andless budgetary stringency. Chile,Ghana, Morocco, and Thailandappear to have best preserved invest-

ment in the more productive areas.The shares of total education and health

expenditures in GDP decreased onlymarginally during the adjustment period.Primary education standards, of particularimportance for growth, were protected andimproved in Chile and Thailand, but not inmost of the other countries. Although ingeneral there was some improvement in pri-mary health care, the countries with thepoorest health indicators at the outset ofthe adjustment effort—Bangladesh, Ghana,and Senegal—made less progress than theothers in this respect.

Overall, the budget-cutting ax fell lessheavily on the civil service wage bill. Wagefreezes were tried in almost every coun-try—especially in the early years—butwere usually reversed. Hiring freezesproved ineffective, because the policy madeit difficult to fill more senior and profes-sional positions. A growth-enhancing pol-icy to reduce employment at the unskilledlevel, combined with increased pay andemployment for skilled workers, was fol-lowed by only two countries—Chile,toward the end of the first adjustmentperiod, and Ghana during the second.

Some success was achieved in cuttingsubsidies and transfers, in part through bet-ter targeting of commodity subsidiestoward the poor. Transfers to enterpriseswere decreased mainly by priceadjustments.

All of the countries achieved some ratio-



nalization of income taxes. The top personalincome tax rate was reduced and the mini-mum taxable income, an indicator of theadministrative burden of the tax, wasincreased in most. Company income taxrates were also reduced and unified in a fewcountries. Efforts in most countries towiden the income tax base were confined tothe introduction of a minimum corporatetax and to some rollback of tax holidays.Exemptions and preferences in most incometax codes remained extensive.

Reforms of indirect taxation were moresubstantial. Bangladesh, Chile, Morocco,and Thailand all introduced VATs,which tended to raise domestic salestax receipts as a share of GDP. TheVATs in Bangladesh, Morocco, andSenegal excluded the retail and/orwholesale sector, while in Morocco,Senegal, and, initially, Mexico, theVAT had up to five rates, which madeit more distortive, harder to adminis-ter, and less elastic than it could havebeen. Governments usually switchedfrom specific to ad valorem exciserates, which increased their elasticity,but the excise tax structure oftenremained complex.

The dispersion and maximumrates of import duties declined in all coun-tries—and in consequence their complexityand distortive effects were diminished-—though the extent and pace of thesereforms varied, being most substantial andrapid in Chile, India, Mexico, and Morocco.Export taxes were virtually abolished dur-ing the adjustment efforts, except in Ghana.

In Chile, India, Mexico, Morocco, andThailand, the public expenditure manage-ment systems were relatively strong, incontrast to the corresponding systems inBangladesh, Ghana, and Senegal; of the lat-ter group, only Ghana improved its systemduring the adjustment effort. The mostsubstantive improvements were under-taken by countries that had alreadyachieved reasonably effective basic sys-tems. Countries with strong public expendi-ture management systems seemed to havemore efficient expenditure programs.

More efforts were undertaken to improvetax and customs administration systems.Like public expenditure management, themost substantial reforms were undertakenin countries—for example, Chile, Mexico,and Thailand—that started with a solidbase. Ghana also made substantial reforms.Bangladesh and India did not achieve sub-stantial improvements, and in Senegal,reforms were begun but not sustained. Insome countries, administrative efficiency

and effectiveness were hampered by inflexi-ble public sector pay and employment poli-cies, inadequate equipment and supplies,the complexity of the tax code, and the con-straints imposed by fiscal federalism (forexample, in India).

Although the data are incomplete, theoperating position of the public enterprisesector improved and the cost of budgetarysupport declined during the adjustmentperiod in all countries but Bangladesh.Enterprise reform was typically limited toincreases in tariffs during the initial adjust-ment years; reforms to increase productiv-

" Whatever the pattern ofexpenditure reduction

and/or tax increases ini-tially adopted, structural

reform will, for most coun-tries, be an essential partof growth-oriented fiscal

adjustment."

ity were generally delayed until later in thesecond adjustment period. Reform took avariety of forms including outright privati-zation, establishment of performance con-tracts for enterprises remaining in thepublic domain, and private sector involve-ment in public enterprise operations.

LessonsThe study suggested a number of conclu-

sions for a growth-enhancing fiscal adjust-ment strategy:

• Stabilization need not play havoc witha growth-friendly fiscal policy regime. Asshown in Chile and Thailand—whosegrowth-enhancing policies stood out rela-tive to those of the other countries—it ispossible to protect a core program of themost productive expenditures.

• Whatever the pattern of expenditurereduction and/or tax increases initiallyadopted, structural reform will, for mostcountries, be an essential part of growth-oriented fiscal adjustment. Ad hocapproaches, though they may cut thedeficit, fail to address more basic problemsand are thus prone to reversal. Some mea-sures—like civil service and enterprisereforms—important though they may be,yield little budgetary savings in the shortrun. Reform in these areas needs to startas soon as possible to reap savings

and enhance growth prospects over themedium term.

• The countries that either have basicallygood public expenditure management sys-tems to begin with or that improve a defi-cient system are on the whole moresuccessful in protecting expenditure with ahigh social rate of return. In general, thebetter the system is in permitting the gov-ernment to set priorities and then enforcechanges in the composition of expenditurethat are consistent with them, the more pro-ductive the ultimate outcome will be.

• Major tax reform, such as the introduc-tion of a VAT, requires either areasonably well-functioning tax ad-ministration or a concurrent effort atadministrative reform. At the sametime, tax design must be cognizant ofthe limitations of tax administration.As is true of public expenditure man-agement, early action to reform taxadministration will be necessary formost countries to achieve a growth-enhancing fiscal structure.

• Increases in revenue are possiblewithout higher average tax rates.With the exception of Ghana, thecountries achieving substantial in-creases in the ratio of taxes to GDP

did so by broadening the effective base ofthe tax system, usually with the introduc-tion of a VAT.

• A reorientation of social expendituretoward primary education and publichealth and clinics is doubly desirable, sinceit boosts human capital formation and rein-forces the social safety net.

• Certain reforms—such as an ambitiousreform of the civil service—can entail atrade-off between deficit reduction now anddeficit reduction in the future, which can bean issue when there is some flexibility inthe timing of fiscal adjustment. In suchcases, the budget needs to be viewed in amultiperiod, rather than a single-period,framework. iF&Dj

This article is based on a longer paper by theauthors, "The Quality of Fiscal Adjustment andGrowth," forthcoming in the IMF's OccasionalPaper series.



GUEST ARTICLE

The Tobin Tax and ExchangeRate Stability

PAUL B E R N D SPAHN

Recent turbulence in worldfinancial markets hasrekindled interest in theso-called Tobin tax on interna-tional financial transactionsas a way to discouragespeculative currency tradingand reduce exchange ratevolatility. A two-tier structuremight be more effective than apure transaction tax.

T HE PRIMARY aim of the Tobintax, first proposed in 1972 byProfessor James Tobin during hisJaneway Lectures at Princeton

University, is to reduce exchange ratevolatility. A uniform international taxpayable on all spot transactions involvingthe conversion of one currency into another,in both domestic security markets and for-eign exchange markets, the Tobin taxwould, in theory, discourage speculation bymaking currency trading more costly. Thevolume of destabilizing short-term capitalflows would decrease, leading to greaterexchange rate stability.

The proposed tax is said to possess anumber of advantages. It would reduce"noise" from market trading while allowingtraders to react to changes in economic fun-damentals and policy, and would therefore

be superior to protective measures such ascapital controls. Because it would requirethe international coordination of macroeco-nomic policies and could be used as a pol-icy instrument by such organizations asthe IMF and the World Bank, it wouldenhance not only market efficiency but alsoglobal financial stability.

As a pure transaction tax, however, theTobin tax would not be effective. Because ofproblems in the way it is structured, theproposed tax would impair the operationsof the international financial markets andcreate liquidity problems without deterringspeculation. An alternative solution mightbe a two-tier structure consisting of a lowtax rate for normal transactions and anexchange surcharge on profits from veryshort-term transactions deemed to be spec-ulative attacks on currencies. Under thisscheme, an exchange rate would be allowedto move freely within a band, but over-shooting the band would result in a tax onthe discrepancy between the marketexchange rate and the closest margin of theband. Exchange rates would thus be keptwithin a target range through taxationrather than central bank intervention or thedepletion of international reserves.

Policy dilemmasThe main problems limiting the effective-

ness of the Tobin tax lie in four crucialareas: establishing the tax base, identifyingtaxable transactions, setting the tax rate,and distributing tax revenues.

The tax base. To limit financial marketdistortions, the base for any tax on interna-tional financial transactions would have tobe as broad as possible. No category of

market participants would be excluded,and the nationality of traders would beirrelevant. The tax would apply to transac-tions by financial institutions, governmentsand international organizations, producersof goods and services, commercial enter-prises, and private households.

However, even leaving aside exemptionsfor market interventions by central banksand for transactions between governmentsand international organizations, there arestrong economic and political argumentsfor exempting certain types of trades fromthe tax—for example, those made by mar-ket makers and those that increase marketliquidity. Indeed, a case can be made toexempt all financial intermediaries fromthe tax on the grounds that their trading isusually stabilizing (through liquidity trad-ing) rather than speculative.

The first dilemma is, therefore, that theTobin tax cannot distinguish, on an institu-tional basis, between normal trading thatassures the efficiency and stability of finan-cial markets and destabilizing noise trad-ing, which should be the only target of thetax. The Tobin tax would be applied to allforeign exchange transactions whether ornot they involve financial institutions andmarket makers. A solution to this dilemmais not readily found, however. On the onehand, exempting such institutions from thetax would simply encourage tax-free trans-actions by and through intermediaries; tax-ing them, on the other hand, would entailefficiency costs.

Taxable transactions. Applying theTobin tax only to spot transactions involv-ing foreign currencies is likely to be inade-quate because it would be possible to avoid

Paul Bernd Spahn,a German national, is Professor of Public Finance at the University of Frankfurt/Main. He was a consultant to the IMF's Fiscal Affairs Departmentduring 1993-95.



the tax by trading in financial derivatives.The substitutability of financial instru-ments thus poses a severe problem for thescheme. The markets have developed cashsubstitutes that would escape the tax; newshort-term instruments, similar to banker'sacceptances and commercial paper, couldbe used to evade a cash-based tax, as couldforeign exchange market funds and repur-chase agreements (made against collateraland not settled on central bank accounts).Moreover, financial derivatives (for exam-ple, forward transactions, futures, andfinancial swaps) permit the transformationof "long trading" into "short trading" withimportant repercussions on spot markets.The volume of such transactions hasgrown rapidly over the past few years andnow accounts for a significantshare of all foreign exchangetransactions.

The problem cannot beresolved simply by extendingthe tax to transactions inderivatives because the size ofsuch transactions cannot berelated to the underlying longtransactions in a straightfor-ward manner. A Tobin tax onthe transactions themselveswould grossly understate thevolume of funds that can be channeledthrough foreign exchange markets; how-ever, taxing the notional value of a deriva-tives contract would probably severelydamage the derivatives markets and mighteven destroy them completely. Given theimportant role played by the forward andfutures markets in hedging risks related toexchange rate fluctuations, the eventualdisappearance of these markets wouldthreaten the stability of foreign exchangemarkets. Another option would be to taxthe notional amounts of derivatives con-tracts, but at lower rates. However,although this would be justified by thelower costs of derivatives, it is undesirablebecause it would create a selective tax sys-tem that would be arbitrary, formidablycomplex to administer, and biased.

The tax rate. It can be argued that atax on foreign exchange transactionsshould ideally operate with a zero rate (or,equivalently, a zero base) when theexchange rate for the currencies in a giventransaction is in equilibrium and that thetax rate should increase in accordance withthe deviation from equilibrium. This wouldmean a variable tax, however, and Tobin'sproposal calls for a uniform tax rate.Moreover, a low, fixed-rate Tobin tax at, say,1 percent on round-trip transactions (sale

and repurchase of foreign currencies) isunlikely to deter investors who expect ashort-term devaluation of 3 percent duringperiods of speculation. A tax rate highenough to deter speculation would seriouslyhamper efficient financial intermediation.

The distribution of revenues. Reve-nues generated by the Tobin tax woulddepend on a number of factors, includingthe tax base, the tax rate, and the volume ofexempt trading. The tax is likely to provokea significant behavioral response by marketparticipants that is difficult, if not impossi-ble, to assess, although it is likely that ahigher tax rate would result in a lower tax-able base. Nonetheless, because of the sheersize of foreign exchange markets, revenuesfrom the Tobin tax could be considerable.

"Whenever the tax is activated,transaction costs would rise signifi-cantly, inducing markets to smooth

out large fluctuations..."

Net turnover in the world's foreign exchangemarkets (spot, forward, and derivative con-tracts) is estimated at $1.23 trillion a day. Astatic revenue estimate (that is, an estimatethat does not take into account behavioralreactions to the tax) for a 1 percent tax ontotal net turnover in all of the world's spotand derivative markets would amount to$13 billion a day, or about $3,250 billionannually (assuming 250 business days peryear). This figure is unrealistic, however,because it does not take into account theeffect of the tax on the behavior of marketparticipants. But, even if foreign exchangemarkets were to shrink by 99 percent inresponse to the new tax, it would still raisesizable revenues of $32 billion. Alternatively,a tax rate of 2 basis points (0.02 percent) on$1.23 trillion could raise $64 billion annually.Such a small tax is likely to trigger only aninsignificant behavioral response, and thestatic revenue estimate is more realistic forsuch a low tax rate.

The question as to who should be enti-tled to the proceeds from the Tobin tax pre-sents yet another difficulty. Tax assignmentis a highly controversial political question.Although the Tobin tax would be assessedand collected by national governments,international coordination would be re-quired to set and enforce the rules. Tobin

proposed the World Bank and the IMF ascandidates for this role. However, it doesnot necessarily follow that proceeds fromthe tax would be assigned to the institutionresponsible for orchestrating it.

Proceeds could be returned to nationalgovernments, but this poses several diffi-culties. Redistribution to the countrieswhere the tax revenues originated wouldfavor countries with important financialcenters and would be inequitable. Anotherbasis for determining redistribution of pro-ceeds to national governments could be therelative size of countries' quotas (votingshares) in an international organizationsuch as the IMF, although this, too, wouldresult in certain inequities.

The basic philosophy behind Tobin'sideas on tax assignment is eco-nomically sound. Given the dif-ficulty of determining theregional incidence of proceeds,they could be assigned to asupranational body and used tofund the provision of publicgoods or global causes suchas basic research in healthor protection of the environ-ment. However, the revenue-raising potential of the Tobintax is so large that this alterna-

tive is unlikely to be accepted by all coun-tries. Moreover, assignment of tax revenuesto an international organization would con-fer considerable power on that organizationand is likely to arouse national resentments.

Significant costs could thus be incurredsimply in trying to establish a worldwideconsensus on the issue of tax assignment.Even if all technical and policy issues couldbe resolved, the issue of assignment couldbe a serious obstacle to implementation ofthe tax.

Other policy optionsDespite these problems, there are few

alternatives to the Tobin tax that couldserve as stabilizing devices in the financialmarkets.

Foreign assets. A tax on the domesticstock of foreign assets—as opposed to oneon flows—has been used by some countries(for example, Germany and Switzerland) inthe past. Such a tax would increase theopportunity costs of holding foreign assets,causing investors to shift to domesticassets. It is questionable, however, whethera tax on the stock of foreign assets candeter short-term speculation—it is morelikely to have a longer-term structuralimpact. Moreover, discriminatory taxationof foreign and domestic assets may not be



wholly consistent with the Liberalization ofCapital Markets Code of the Organizationfor Economic Cooperation andDevelopment or with the spirit of the WorldTrade Organization.

Capital flows. Another possibilitywould be to levy taxes on capital outflowsor inflows. The United States, for example,taxed outflows during the 1960s (the inter-est rate equalization tax); more recently,Israel imposed a tax on inflows. But suchtaxes have regularly been abandoned, per-haps because of their structuralimpact and ineffectiveness in fight-ing speculation. They have alsofailed to resolve underlying struc-tural problems on a more perma-nent basis and been difficult toreconcile with the freedom of capi-tal movements.

Capital gains. Another alter-native to the Tobin tax would be asliding-scale capital gains tax thatwould be higher for short-termcapital gains. However, such a taxwould presumably have to beembedded in national income taxlegislation; it is difficult to see howit could be coordinated at the inter-national level. Experience withnational withholding taxes oninterest income demonstrates thatfinancial markets that benefit fromlow or no taxes on the incomes or capitalgains of foreigners have little incentive tocooperate at an international level. A slid-ing-scale capital gains tax would also posesevere administrative problems because thetax rate would have to vary according tothe term structure, as well as the source, ofcapital gains. Furthermore, in a world ofintegrated information and telecommunica-tion networks, it is difficult to pin downwhere capital gains originate.

A two-tier Tobin taxAchieving exchange rate stability

through taxation would require high and/orvarying tax rates, which would seriouslyobstruct the workings of internationalfinancial markets. In contrast, a smallcharge on international financial transac-tions would not create distortions butwould also fail to inhibit speculative behav-ior in foreign exchange markets. A possiblecompromise would be a two-tier structure:a minimal-rate transaction tax and anexchange surcharge that, as an antispecula-tion device, would be triggered only duringperiods of exchange rate turbulence and onthe basis of well-established quantitativecriteria. The minimal-rate transaction tax

would function on a continuing basis andraise substantial, stable revenues withoutnecessarily impairing the normal liquidityfunction of world financial markets. Itwould also serve as a monitoring and con-trolling device for the exchange surcharge,which would be administered jointly withthe transaction tax. The exchange sur-charge, which would be dormant so long asforeign exchange markets were operatingnormally, would not be used to raise rev-enue; it would function as an automatic

A tax on exchange rate movements outsidea band could discourage speculative trading

8 12 16 20 24 28 32 36 40 44 48(number of days)

Source: Author's calculations.

circuit-breaker whenever speculativeattacks against currencies occurred (if theyoccurred at all under this regime). The twotaxes would thus be fully integrated, withthe former constituting the operational andcomputational vehicle for the latter.

The underlying transaction tax. Aminimal nominal charge of, for example, 2basis points on foreign exchange transac-tions would raise the cost of capitalinsignificantly and would probably have noeffect on the volume of transactions involv-ing currency conversions. A transaction taxcould also be imposed on derivative tradesat half the standard rate, or 1 basis point.This would allow the derivatives marketsto continue functioning at low cost whilepreventing the use of derivatives to evadetaxes.

The exchange surcharge. Theexchange surcharge would be administeredin conjunction with the underlying transac-tion tax, but its aim and implementationwould be different. The aim would be to taxnegative externalities associated withexcessive volatility. For normal operations,the fixed-rate surcharge would be zerobecause the tax base is zero, which assuresmarket liquidity and allows efficient trad-

ing. The surcharge would be levied onlyduring periods of speculative trading whenthe tax base becomes positive. It could beconfined to cash transactions or, if neces-sary, could easily be extended to the deriva-tive market. Ideally, if the exchangesurcharge achieves its objective, it wouldgenerate no revenues.

The surcharge would be price-sensitive;it would be switched on whenever the trad-ing price for a currency passed a predeter-mined threshold, which would be

determined by a crawling peg(like a moving average) plus asafety margin (defined as a per-centage of the target rate). Themargins may vary for differentcurrencies but the same ruleswould apply to all markets andto all institutions operating inthe markets. Of course, theshorter the time interval for thecrawl, the greater the scope forshort-term fluctuations. Theinterval should, however, beshort enough to avoid "leaningagainst the wind"—sustainingan effective exchange rateagainst market trends—so thatmarkets can adjust fully tochanges in fundamentals.

Whenever the tax is acti-vated, transaction costs would

rise significantly, inducing markets tosmooth out large fluctuations to avoid suchhigh costs. Traders would be given theright to recontract, however, because trans-action costs could not be known in advance.This would make a significant proportionof contracts contingent, and speculativeattacks would become more difficult tocarry out because traders would automati-cally withdraw from markets during peri-ods of large fluctuations in prices. Ideally,this should induce markets to behave moresmoothly, in their own interests, and the taxwould seldom need to be activated.

The idea is illustrated in the chart. Aneffective exchange rate is simulated over 50days with a forward-looking moving aver-age as the target exchange rate for eachday. A higher and lower tolerable rate aredefined in proportion to the target rate. Aslong as daily fluctuations remain within theband, no tax is levied. Once the effectiverate moves beyond the tolerable range, thedifference between the band and the effec-tive rate (shaded area) is taxed at a constantbut high rate.

The scheme is very similar to theEuropean Monetary System's mechanismfor achieving exchange rate stability



through a target rate and an admissiblespread or band. However, rather than sup-porting a weakening currency throughhigh premiums on overnight moneydeposits or through sales of foreignexchange reserves, with the exchange sur-charge, currencies are defended by taxingtechnically well-defined windfall profits.Unlike the Tobin tax, which would indis-criminately tax all transactions at the"wrong end" and therefore penalize normalliquidity trading, the exchange surchargewould apply only to transactions at the"speculative end" and would not affect nor-mal trading.

Ideally, the two-tier scheme would workon a global scale, as would the tax origi-nally proposed by Tobin, but, initially, itcould be implemented unilaterally by oneor a few countries.

BenefitsThe exchange surcharge proposed in

this article is intended to be a short-termstabilizing tool of monetary policy, of

course; it would not be used to correctstructural problems. Its purpose would beto allow smooth adjustment of exchangerates to economic fundamentals, not torestore an ailing economy to health. Norwould the surcharge be able to preventspeculative trading triggered by suddenfears of payment defaults or political crises.

The exchange surcharge would, how-ever, avoid the negative effects of othermonetary policy measures that sacrificevaluable international reserves or offerexcessively generous interest rates to com-bat speculative attacks. Instead of deplet-ing public assets, it would generaterevenues. It would also eliminate expecta-tions of recurrent bailouts by central banks,reducing moral hazard as well as the inci-dence of financial crises. Moreover, the sur-charge would be likely to stabilizelonger-term exchange rate movements, pri-marily through its impact on investors'expectations.

The goal of the surcharge—exchangerate stability—is, of course, not an end in

itself. Stability reduces the scope for pricebubbles and false signals, improves theallocation of international resources, andreduces risk premiums—in particular thoserelated to inflation. It also restores some ofthe autonomy governments and centralbanks can lose as a result of heavy specula-tion and may help prevent harmful politicalmeasures taken in attempts to correct mis-aligned exchange rates. IF&DI

Suggestions for further reading:James Tobin, 1978, "A Proposal for

International Monetary Reform," EasternEconomic Journal, Vol. 4 (July-October),pp. 153-59.

, 1984, "On the Efficiency of theFinancial System," Lloyds Bank Review,Vol. 153 (July), pp. 1-15.

, 1991, International CurrencyRegimes, Capital Mobility, and MacroeconomicPolicy, Cowles Foundation Discussion PaperNo. 993, New Haven, Yale University.

The World Economic Outlook, May 1996, available now

A Survey by the Staff of the International Monetary Fund

The World Economic Jftlaek, May 1994, presents ttff staff economists'analyses of global tfJQOmk developments during tf>t sear and mediumterms. Particular ewfjfiasis is g Win to growth if» emerging market coun-tries; savings, investment, and debt; budgetary imbalances in industrialcountries; and problems feting those transition economies well advanced intheir move to a market economy. Annexes, boxes, charts, and an extensivestatistical appendix augment the text.

US$35.00 (academic rate: $24.00) (paper).Available in English, French, Spanish, and Arabic.

To order, please write or call:

International Monetary FundPublication ServicesBox FD-296700 19th Street, N.W.Washington, D.C. 20431 U.S.A.

Telephone (202) 623-7430Telefax: (202)623-7201Cable Address: INTERFUND

American Express, MasterCard, and VISA credit cards accepted.



COUNTERPOINT

Why a Two-Tier Tobin Tax Won't WorkJ A N E T G. S T O T S K Y

Financial market variabilityis a perennial problem forinvestors and policymakersalike. While attractive in someways, a two-tier Tobin taxwould not solve this problemand would have a number ofundesirable side effects.

I N HIS article in this issue of Finance &Development, Paul Bernd Spahn offersa creative extension on the Tobin taxidea. His proposed two-tier tax is

designed to deal with a major problem of aTobin tax, namely, if imposed at a high rate,the tax would seriously impair the normaloperations of financial markets; while if thetax is imposed at a low rate, it would notdeter currency traders who expect signifi-cant short-term changes in currency values.The top tier of Spahn's proposed tax—theexchange surcharge—would function as anautomatic circuit breaker whenever specu-lative attacks against currencies occurred.Spahn argues that when the surcharge wastriggered, transactions costs would riseenough to cause some traders to delaytransactions, thus smoothing out changesin the value of the currency. Revenues fromthe tax might be allocated to the countriesof origin, to an international body, or insome other fashion.

Spahn's argument in favor of such a taxrests on the assertion that imposing thehigh rate during speculative attacks wouldcurb volatility, while the low rate thatwould apply during normal market condi-

tions would not impair market efficiency.He acknowledges that there are severalimportant drawbacks to the proposal. Thetax would apply to all transactions, withoutdistinguishing stabilizing from destabiliz-ing ones; it would be difficult to design thetax so that it applied equally to all financialinstruments; and the allocation of revenuesfrom the tax would be contentious.

Spahn faces an uphill battle in gainingacceptance for such an idea. There are theusual arguments against the Tobin taxitself to contend with. The main one is thatthere is little evidence that such taxesreduce market volatility—volatility couldeven increase. At the same time such taxesincrease transaction costs and hinder theoperation of financial markets. Therewould also be enormous administrative dif-ficulties in implementing such a tax.

Spahn's proposal raises several otherimportant issues. One issue is the desirabil-ity of using variable rate taxes. In practice,such taxes are rarely used. The main rea-son for their unpopularity is that they cre-ate uncertainty over prices in markets. Thisis, in fact, the feature of Spahn's proposedvariable surcharge that could make it effec-tive in altering market behavior. But at thesame time, in the absence of volatility, theadditional uncertainty it would create infinancial markets is likely to impair theiroperation and increase spreads. Variable-rate taxes are also rarely used because theycomplicate considerably the burdens ontaxpayers and tax administrations. Thiseffect would be particularly severe in thecase of a two-tier Tobin tax because thenumber of separate transactions to whichthe tax would apply is so huge.

A second issue is the advisability of mix-ing monetary policy with tax policy, giventhe different constraints under which each

operates. Monetary policy can changequickly and is generally somewhat insu-lated from politics. Tax policy changes onlyinfrequently and political considerationsare generally paramount. Spahn does notaddress either of these two issues, thoughthey are crucial to assessing his proposal.

Effect on market volatility andefficiency. Since Tobin taxes do not existin practice, there is little empirical evidenceto suggest that such taxes would be effec-tive in reducing currency fluctuations, asSpahn asserts. Countries have used variousforms of capital controls to reduce currencyfluctuations, including implicit taxes suchas nonremunerated deposits in the centralbank. An extensive literature has examinedhow capital controls affect foreign ex-change markets (see Dooley, 1995), reach-ing the broad conclusion that while capitalcontrols may delay currency adjustmentsin the short run, they are ineffective in thelong run. Similar conclusions have beenreached in studies of the US stock market,where a variety of methods, such as circuitbreakers and margin controls, have beenemployed to reduce volatility.

More generally, empirical observationsdo not provide a basis for asserting a firmlink between transaction costs and volatil-ity. Even in the past, when transactioncosts in financial markets were generallylarger than today, fluctuations in capitalflows and prices were observed. In recentyears, transaction costs have fallen signifi-cantly for participants in major foreignexchange and stock and derivative marketswithout any apparent increase in volatility.

The main argument against financialtransaction taxes is that they reduce mar-ket efficiency, as Spahn acknowledges.Such taxes could impose a cost on financialmarkets by creating a disincentive to trade

Janet G. Stotsky,a US national, is an Economist in the Tax Policy Division of the IMF's Fiscal Affairs Department.



assets by inducing investors to hold a lessdesired portfolio and by potentially reduc-ing stabilizing arbitrage. Moreover, thesetaxes would increase the cost of capital,and thereby lead to lower rates of capitalformation and economic growth.

Since transaction taxes levied equally onall assets would effectively tax more heav-ily short-term assets and those traded morefrequently, the most critical issue is thedesirability of taxing short-term transac-tions more heavily. Proponents of financialtransaction taxes have argued that short-term traders are precisely those whoseactivities are most destabilizing and thusadvocate financial transaction taxes on thegrounds that they would effectively targetthis group. The motivations of differenttraders in financial markets are not, how-ever, well understood (see IMF, 1995,International Capital Markets: Develop-ments, Prospects, and Policy Issues), andthere is no way to target only destabilizingtraders. The case for limiting the activitiesof all short-term traders is not persuasive.

Intermediaries. Financial transac-tions taxes would have a negative effect onthe short-term liquidity trading of financialinstitutions, which rely heavily on trade inshort-term financial assets to hedge cur-rency and other investment risks and toprovide liquidity to enterprises and individ-uals. Acknowledging this problem, Spahnsuggests that the tax could exempt finan-cial institutions and market makers.However, financial institutions and marketmakers are not necessarily stabilizingtraders. In addition, exempting certainintermediaries would encourage their useby taxable market participants, irrespec-tive of any advantages these intermediariesmight provide.

Cascading effects. Spahn dismissesthe efficiency effects of the low-rate Tobintax that would apply under normal marketconditions on the grounds that the taxwould be negligible. Under circumstancesof stable currency values, even whatappear to be small tax rates on turnovermay constitute a significant tax on thereturns to trading financial assets in viewof the narrow spreads that characterize for-eign exchange market transactions. Even a2 basis point (0.02 percent) tax on transac-tions value, as Spahn suggests, wouldresult in a significant tax on the return totrading.

It is often assumed that taxes on turnoverare shifted forward to the consumer andhence do not reduce the return to an activ-ity. However, this is not likely to be the casewith a tax on financial transactions, since

most financial transactions are undertakenby intermediaries. Taxes levied on turnoverin this manner tend to cascade, so that theeffective rate can be many times the nomi-nal rate applied to a single transaction.

Derivatives. Derivative instrumentshave grown rapidly and are especiallyimportant in foreign exchange markets.Since investors can construct equivalentpositions with derivatives as they wouldwith cash instruments, transactions inderivatives should be taxed. It is difficult,however, to achieve equivalent taxation ofcash and derivative instruments, as Spahnnotes. His solution—to tax derivative trans-actions at one-half the standard rate—isnot appropriate. Given the complexity ofthe strategies underlying the use of deriva-tives, it would be impossible to establishone rate for derivatives and one for theunderlying instruments that would yieldexact tax equivalences. Markets wouldquickly figure this out.

Ease of avoidance. Another funda-mental problem with Spahn's proposal isthat if a country were to unilaterallyimpose a financial transactions tax, itwould be easily avoided. If transactionstaxes applied to transactions only indomestic markets, investors could easilysubstitute foreign trading as a means toavoid the tax. For instance, a considerableamount of trading in US currency or equi-ties takes place abroad. If financial transac-tions taxes applied to all currencies, traderscould shift into vehicle currencies to avoidmaking currency conversions, increasingcosts for small, not widely traded curren-cies and impairing monetary control inthese countries.

In terms of the effective implementationof the tax, the mobility of financial transac-tions would make the tax easy to avoidunless the tax were internationally agreedupon and administered by each govern-ment. The effectiveness of the tax would begreatly reduced if only a few governmentswith major financial markets enacted it.Tax havens have proliferated in recentyears. Such a tax would only add impetusto this trend. The rules for applying the taxwould have to be established by an interna-tional consensus. But it has proven difficultto get countries to agree upon uniform tax-ation in other areas of taxation, even by rel-atively homogeneous groups of countries,such as the European Union.

Use of revenues. The use of the rev-enues from the tax is likely to be more con-tentious than Spahn suggests. The ideathat the revenues could be used to supportthe United Nations, for instance, has

recently received an icy reception from USpolicymakers.

Interaction between monetaryand fiscal policies. The introduction ofeven the simpler Tobin tax in major finan-cial markets would entail significantadministrative costs in developing methodsfor its collection, monitoring, and enforce-ment. The introduction of Spahn's proposedtwo-tier tax would be even more complex,requiring not only the apparatus of the sim-ple tax, but all of the monetary policyconsiderations relevant to setting andmaintaining exchange rates. The properuse of the tax would require a degree ofcooperation between monetary and fiscalauthorities that does not exist in practice.Compared with monetary policy, fiscal pol-icy is generally determined in a far moreconstrained environment. It is doubtful thatmonetary authorities would have the abilityand independence to administer such a taxwisclylF&Dl

ReferencesMichael P. Dooley, 1995, "A Survey of

Academic Literature on Controls OverInternational Capital Transactions," IMFWorking Paper No. 95/127 (November).

Barry Eichengreen and Charles Wyplosz,1993, "The Unstable EMS," Brookings Paperson Economic Activity: 1, pp. 51-139.

Barry Eichengreen, James Tobin, and CharlesWyplosz, 1995, "Two Cases for Sand in theWheels of International Finance," EconomicJournal, Vol. 105, pp. 162-72.

Peter Garber and Mark P Taylor, 1995,"Sand in the Wheels of Foreign ExchangeMarkets: A Skeptical Note," Economic Journal,Vol. 105, pp. 173-80.

David Greenaway, 1995, "Policy Forum:Sand in the Wheels of International Finance,"Economic Journal, Vol. 105, pp. 160-61.

IMF, 1995, International Capital Markets:Developments, Prospects, and Policy Issues(Washington).

Peter B. Kenen, 1995, "Capital Controls, TheEMS and EMU," Economic Journal,Vol. 105, pp. 181-92.

Parthasarathi Shome and Janet G. Stotsky,1996, "Financial Transactions Taxes,"TaxNotes International (January 1), pp. 47-56.

James Tobin, 1978, "A Proposal forInternational Monetary Reform," EasternEconomic Journal, Vol. 4 (July-October),pp. 153-59.



World Economy In Transition

T

HE WORLD'S poor are concentrated in 65 low-income countries, mainly in sub-Saharan Africa andSouth Asia, with a population of 3.2 billion and a percapita income of $390 a year. Most live in rural areas;they lack adequate access to safe water and other

basic services; nearly 40 percent of their children are malnour-ished; and over 40 percent of primary-school-age children arenot enrolled in school. These statistics are disheartening, yet,over the past 20 years, significant progress has been made inimproving social conditions in these countries, especially forwomen and children, who make up the majority of the world'sdisadvantaged.

• Fertility rates are declining in low-income countries—from 5.6 births per woman in the early 1970s to 3.3 in 1994(Chart 1). This decline has not been evenly spread, how-ever—over this period, fertility declined by nearly 40 percentin South Asia, but only by slightly more than 10 percent insub-Saharan Africa.

• Declining fertility has been accompanied by significantincreases in access to education, especially for girls (Chart 2).In South Asia, primary school enrollment rates increased dra-matically, from 50 percent of all school-age girls in 1970 to87 percent in 1993. In sub-Saharan Africa, enrollment ratesincreased by 50 percent. In low-income countries as a whole,enrollment of girls increased by about 20 percent, while totalprimary enrollment increased by over 40 percent.

• Preventive health care has improved. For example, child-hood immunization against measles in low-income countrieshas risen from 50 percent of children in 1985 to 86 percent in1992, with South Asia showing the most dramatic increase.Access to health care in South Asia rose from 54 percent of thepopulation in 1980 to 96 percent by the end of the decade, andaccess to safe water increased from 50 percent of the popula-tion to 70 percent over the same period.

• As a result of these improvements, children born in low-income countries in 1994 can expect to live, on average, 63years, compared with 54 years for those born 20 years ago(Chart 3).

Improving living standards and reducing poverty neverthe-less continue to pose a challenge to governments and thedevelopment community. Around 1.3 billion people (1993 esti-mate) in the developing world live on less than one dollar aday. Because this figure conceals large regional differences,poverty will remain a persistent problem requiring concertedefforts.

Thirty-nine percent of the world's poor live in South Asia,while only 17 percent live in sub-Saharan Africa. But nearly40 percent of sub-Saharan Africa's population is below thepoverty line, and the region has seen a gradual erosion of liv-ing standards over the past two decades. Real per capitaincome declined, reflecting weak economic growth and arapidly rising population (Chart 4). A few countries have expe-rienced increases in infant mortality, decreases in lifeexpectancy, declines in women's advantage in life expectancy,and slight increases in child malnutrition. Others have experi-enced some deterioration in the provision of social services:fewer people today than in the early 1970s have access to safedrinking water, and primary school enrollment rates havefallen or leveled off.

By contrast, East Asia—with the exception of China—hasreduced the share of its population living in poverty, fromabout 23 percent in 1987 to 13.7 percent in 1993. Annual GNPper capita growth rates accelerated from 4.3 percent in1970-75 to 7.2 percent in 1989-94; over the same period, popu-lation growth rates declined from 2.3 percent to 1.4 percent. Incountries such as Indonesia, the proportion of people livingbelow the official poverty line declined from 45 percent in theearly 1980s to 17 percent a decade later. Per capita income rosefrom $230 in 1975 to $880 in 1994; infant mortality was morethan halved; and access to basic services, such as safe drink-ing water and education, increased. Much of the progress inthis region reflects a policy of fostering broad-based economicgrowth, coupled with the expansion of basic services andmajor investments in human capital.

Sulekha Patel,a national of India, is a demographer in the International Economics Department of the World Bank.

HaBST" *B8W6K -< am. JBTW& maam^ad^^

World Economy In Transition

T

HE WORLD'S poor are concentrated in 65 low-income countries, mainly in sub-Saharan Africa andSouth Asia, with a population of 3.2 billion and a percapita income of $390 a year. Most live in rural areas;they lack adequate access to safe water and other

basic services; nearly 40 percent of their children are malnour-ished; and over 40 percent of primary-school-age children arenot enrolled in school. These statistics are disheartening, yet,over the past 20 years, significant progress has been made inimproving social conditions in these countries, especially forwomen and children, who make up the majority of the world'sdisadvantaged.

• Fertility rates are declining in low-income countries—from 5.6 births per woman in the early 1970s to 3.3 in 1994(Chart 1). This decline has not been evenly spread, how-ever—over this period, fertility declined by nearly 40 percentin South Asia, but only by slightly more than 10 percent insub-Saharan Africa.

• Declining fertility has been accompanied by significantincreases in access to education, especially for girls (Chart 2).In South Asia, primary school enrollment rates increased dra-matically, from 50 percent of all school-age girls in 1970 to87 percent in 1993. In sub-Saharan Africa, enrollment ratesincreased by 50 percent. In low-income countries as a whole,enrollment of girls increased by about 20 percent, while totalprimary enrollment increased by over 40 percent.

• Preventive health care has improved. For example, child-hood immunization against measles in low-income countrieshas risen from 50 percent of children in 1985 to 86 percent in1992, with South Asia showing the most dramatic increase.Access to health care in South Asia rose from 54 percent of thepopulation in 1980 to 96 percent by the end of the decade, andaccess to safe water increased from 50 percent of the popula-tion to 70 percent over the same period.

• As a result of these improvements, children born in low-income countries in 1994 can expect to live, on average, 63years, compared with 54 years for those born 20 years ago(Chart 3).

Improving living standards and reducing poverty neverthe-less continue to pose a challenge to governments and thedevelopment community. Around 1.3 billion people (1993 esti-mate) in the developing world live on less than one dollar aday. Because this figure conceals large regional differences,poverty will remain a persistent problem requiring concertedefforts.

Thirty-nine percent of the world's poor live in South Asia,while only 17 percent live in sub-Saharan Africa. But nearly40 percent of sub-Saharan Africa's population is below thepoverty line, and the region has seen a gradual erosion of liv-ing standards over the past two decades. Real per capitaincome declined, reflecting weak economic growth and arapidly rising population (Chart 4). A few countries have expe-rienced increases in infant mortality, decreases in lifeexpectancy, declines in women's advantage in life expectancy,and slight increases in child malnutrition. Others have experi-enced some deterioration in the provision of social services:fewer people today than in the early 1970s have access to safedrinking water, and primary school enrollment rates havefallen or leveled off.

By contrast, East Asia—with the exception of China—hasreduced the share of its population living in poverty, fromabout 23 percent in 1987 to 13.7 percent in 1993. Annual GNPper capita growth rates accelerated from 4.3 percent in1970-75 to 7.2 percent in 1989-94; over the same period, popu-lation growth rates declined from 2.3 percent to 1.4 percent. Incountries such as Indonesia, the proportion of people livingbelow the official poverty line declined from 45 percent in theearly 1980s to 17 percent a decade later. Per capita income rosefrom $230 in 1975 to $880 in 1994; infant mortality was morethan halved; and access to basic services, such as safe drink-ing water and education, increased. Much of the progress inthis region reflects a policy of fostering broad-based economicgrowth, coupled with the expansion of basic services andmajor investments in human capital.

Sulekha Patel,a national of India, is a demographer in the International Economics Department of the World Bank.

j ! £'• V PHaBST" *B8W6K -< am. JBTW& maam^ad^^

Social indictators of Delvelopment

j ! £'• V P



Chart 4

Income and population growthGNP per capita

(percent change)

This article is based on information available inSocial Indicators of Development 1996, pub-lished for the World Bank by Johns HopkinsUniversity Press, Baltimore, 1996, and PovertyReduction and the World Bank: Progress andChallenges in the 1990s, World Bank, 1996.Social Indicators of Development 1996 is acompilation of the latest data on fertility,mortality, child malnutrition and immuniza-

tion, illiteracy and school enrollment, access tobasic social services, and shares of spending forselected social services and infrastructure. Thereport monitors and evaluates social progressand human welfare in more than 200 coun-tries, and, through its regional and incomegroupings, presents a framework for compar-ing key country indicators with the relevantregional and income group.

li

Chart 2Girls' access to education

is improving 1

Chart 1

Fertility rates are declining 1

Charts

Life expectancy is rising 1

mr""iii v*w -

Univer


The Revenue Decline in the Countries of theFormer Soviet Union

A D R I E N N E C H E A S T Y

In almost all the countries ofthe former Soviet Union,severe revenue losses havecomplicated stabilizationefforts. The situation can bealleviated by measures tobroaden tax bases andimprove tax administration.

I N 1989, general government revenuein the Soviet Union was 41 percent ofGDP—not far from the average formember countries of the Organiza-

tion for Economic Cooperation andDevelopment (OECD). But by 1995, rev-enues in the countries of the former SovietUnion were down to an average of 25 per-cent of GDP. (See chart and Table 1.)

The revenue decline was not the result ofdecisions by reform-minded governmentsto withdraw from intervention in the econ-omy. Rather, it came at a time when theyurgently needed resources to finance eco-nomic restructuring and a social safety netto cushion the impact of falling living stan-dards. Because spending pressures have

been hard to resist, revenue losses havemade stabilization considerably more diffi-cult, contributed to high inflation rates, andcaused governments to delay structuralreforms.

Main causesIn the general upheaval of the transition,

some decline in revenue was to be expected.Since much of the revenue decline could beblamed on decreasing output, the revenue-GDP ratio was expected to remain stable. Infact, the decline has been sharper anddeeper than predicted. Some causes are

• Many system-wide changes created amore difficult environment for revenue col-lection than had existed during the 1980s.

• Traditional tax bases (state industrialproduction, retail turnover, and wages) allshrank faster than GDP, while agricultureand the private sector—which were moreor less untaxed—did relatively well.

• Many market-oriented tax reformswere revenue-losing, largely because theSoviet tax system had been heavily distor-tionary, and establishing market incentiveswould have been impossible without lowertax burdens.

• The legitimacy of the new govern-ments is being tested by the growth of thegray economy, by powerful interest groupsexerting pressures for tax exemptions, by

tax competition among regions, and by var-ious types of civil unrest that limit thereach of the tax authorities.

System-wide influences. Regionalwars and civil strife clearly explain some ofthe more extreme cases of revenue loss inthe countries of the former Soviet Union.Also, some countries lost important inter-governmental transfers that had boosted

Adrienne Cheasty,a national of Ireland, is Deputy Division Chief of the Fiscal Operations Division I of the IMF's Fiscal Affairs Department.

32 Fmnnre. & Development / June 199fi

Falling revenues in the countriesof the former Soviet Union 1

Source: International Monetary Fund.1 General government revenue as a percentage

of GDP (unweighted average).


tax bases; in the Central Asian countries,for example, transfers had amounted to 20percent or more of GDP toward the end ofthe 1980s.

Two reform-related systemic shocks alsohad a notable impact on revenues. (SeeTable 2.) The first was the decentralizationof foreign trade. Trade-related levies hadprovided the Soviet Union with 15 percentof its revenue (more than 6 percent of GDP).Since much of this trade was in energy, thisshould have been by far the most buoyantrevenue base after energy prices were liber-alized. But trade tax revenues fell below 2percent of GDP in all but three countries,and trade taxes on energy brought in only asmall share of total revenues. This outcomereflected both a policy choice—reforminggovernments opted for extremely opentrade regimes—and a pragmatic deci-sion—the countries had few border con-trols and no customs mechanisms.

The second shock was the 1991 overhaulof the former Soviet pension system, whichraised payroll tax rates from 9 percent in

1989 to an eventual 36 percent in mostcountries. This complicates the interpreta-tion of movements in individual taxes,since, by reducing profits and equilibriumwages, the overhaul was partly responsiblefor the sharp decline in enterprise and indi-vidual income tax receipts.

Declining tax bases. The most signif-icant part of the fall in revenue ratios canbe explained by a sharper decline in taxedactivities than in the untaxed part of theeconomy—notably the earnings of agricul-ture and small private businesses. Systemicshocks, price liberalization, the eliminationof most producer subsidies, and tight mone-tary programs hit industrial productionand official retail sales particularly hard.These developments forced real wagesdown, which eventually lowered revenuefrom both production- and consumption-based taxes.

The impact of the loss of traditional taxbases was most pronounced in the first twoyears of transition. In 11 of the 15 countries,the decline in enterprise taxes during

Table 1

Varying severity of the revenue decline in the countries ofthe former Soviet Union, 1991-95

(percent of GDP)

GeorgiaTurkmenistanTajikistanAzerbaijanArmeniaMoldova 'RussiaKyrgyz Republic 1

Kazakhstan 1

LithuaniaLatviaEstonia

BelarusUkraineUzbekistan 1

Average

1991

3438333626324622221

413741

483831

35

1992

154236482720461623

342833

464432

33

1993

21936412417411522

283640

524442

31

1994

41054261617371917

253641

484636

29

1995

59141914203621516

243641

444135

25

Difference1991-95

-29-29-19-17-13-12-10

-7-5

-18-10

-434

-10

Sources: IMF, Economic Reviews, various countries of the former Soviet Union; and WorldEconomic Outlook, various issues; World Bank, 1 992, Statistical Handbook: States of the FormerUSSR, and 1996, Fiscal Management in the Russian Federation.

1 Does not include all payroll taxes.2 Estimates.

1991-93 could be completely accounted forby the fall in industrial production or enter-prise profits. In all countries, more thantwo-thirds of the decline in domestic indi-rect taxes could be explained by the drop inretail turnover from state stores.Plummeting real wages and the (minor)impact of emerging unemployment fullyexplained the erosion of personal incometaxes and payroll taxes in 10 of the 13 coun-tries for which data were available.

The shrinking of traditional tax baseswas accompanied by an expansion of pri-vate activity. Even in Ukraine, a slowreformer, the officially measured privatesector grew to 20 percent of GDP in 1995from 6 percent in 1993. Since small enter-prises are hard to tax and new firms typi-cally take some time to recover start-upcosts, this transformation would haveimplied some revenue loss even if outputhad not fallen.

The decline in tax bases was exacerbatedby interenterprise arrears. Taxes were, inalmost all cases, levied on a cash basis.Enterprise output that was consumed, butnot paid for, was not taxed. Since arrearscould be sizable (for example, theyamounted to 100 percent of GDP in Russiain the first four months of 1992), the erosionof the true tax base (as opposed to recordedproduction) was considerable. Moreover,high inflation rates meant that even ifarrears were eventually paid, the real valueof tax revenue was seriously eroded.

Revenue-losing tax reform. Many ofthe tax reforms undertaken by transitioncountries have been revenue-losing—anoutcome that is not surprising consideringthat, as late as 1990, the top marginalincome tax rate was 90 percent. Revenue-losing reforms included the dismantling ofthe trade revenue regime and thenarrowing of the enterprise tax base toreflect a more economic definition of prof-its, by eliminating the taxation of wagesand other cost elements. The average effec-tive profit tax rate was an estimated 55 per-cent in 1990; now, the central rate in allcountries of the former Soviet Union is 35percent or less.

Many countries, and particularly slowreformers—such as Belarus, Tajikistan,and Ukraine—sustained their high effectiveyields by reincorporating cost elements



(mainly wages or "excess wages")into the profit tax base. However, asthe output decline has spread, enter-prises have clamored for lighter taxburdens. Belarus and Ukraine haveresponded by granting wider exemp-tions, Tajikistan by increasing bankcredit so that enterprises can meettheir obligations. But even in some ofthe faster reformers, costs are taxedin the sense that depreciationallowances remain significantlybelow industrial country norms, andenterprises remain decapitalized.

The introduction of the value-added tax (VAT) also led to a net lossof revenue for half of the countries inthe region, compared with the previ-ous turnover tax regime, which hadlevied (tax-exclusive) rates of severalhundred percent on some goods.Moreover, to cushion businessesagainst the effects of price liberaliza-tion, several countries subsequentlycut their VAT rates.

Part of the revenue loss from theVAT arose from defects in the formerSoviet Union countries' versions ofthe tax, compared with industrialcountries' VATs. The former SovietUnion countries' VATs were leviedpartially on a cash basis and in anasymmetric way that exacerbatedthe revenue losses from arrearsdescribed earlier. Also, in sectorswhere accounting practices were notsufficiently developed to include the use ofinvoices as payment records, the VAT wasoften levied on a fixed markup over costrather than on actual prices. This meantthat value added higher than themarkup—an amount that could be largewhen prices were rising rapidly and unpre-dictably—remained untaxed. Further,because borders were not well establishedwithin the former Soviet Union, VATs onmutual trade were levied on an originbasis—with exports subject to VAT whileimports were not taxed—while the custom-ary destination basis (with imports taxed,exports not) was (with some delay) appliedto VAT levied on trade with the rest of theworld. This dual system created incentivesand opportunities to divert and misrepre-sent trade flows and values, and resulted ina low compliance rate for VATs on trade.

Deficiencies in governance. Someof the teething troubles of the countries'new governments have added to their diffi-culties in collecting taxes. Much of the newprivate sector activity has disappeared intothe shadow economy, and tax administra-

Table 2

Which taxes have been most affected?

SovietUnion1989

Unweightedaverage19941

Decline1989-942

Total 41.0

Taxes on wages 8.0Personal income tax 4.4Social insurance 3.5

Taxes on enterprises 3 12.3Taxes on consumption 12.0Foreign activity 6.3Nontax revenue 2.5

Taxes on wages 19.0Personal income tax 11.0Social insurance 9.0

Taxes on enterprises 3 30.0Taxes on consumption 29.0Foreign activity 15.0Nontax revenue 6.0

(percent of GDP)28.7

8.22.65.67.19.01.72.7

(percent of total)

28.09.0

19.025.031.0

6.09.0

-12.3

0.2-1.82.0

-5.2-2.9-4.60.2

-1.714.9

-16.642.223.737.5-1.7

Sources: IMF, Economic Reviews, various countries of theformer Soviet Union; and World Economic Outlook, variousissues; World Bank, 1992, Statistical Handbook: States ofthe Former USSR; and IMF, World Bank, Organization forEconomic Cooperation and Development, and EuropeanBank for Reconstruction and Development, 1991, A Studyof the Soviet Economy (Paris).

1 Countries of the former Soviet Union.2 The calculation of the decline is included for illustrative

purposes only; the USSR figures and the average figuresfor the former Soviet Union are not strictly comparablebecause the 1994 average is unweighted.

3 Miscellaneous "other taxes" have been added to "taxeson enterprises." In all countries where they could be identi-fied, they were property taxes or taxes on naturalresources.

tors are finding it difficult to capture it inthe tax net—partly because much of thetrade is very small scale and partly becauseof the activities of criminal elements. A sur-vey conducted in Ukraine (Kaufman (1994))suggested that 75 percent of privatetraders' activity was carried on outside theofficial economy.

The uncertain allegiance of some regionsto their national governments has also lim-ited the scope for those governments toraise taxes. In Russia, for example, the fearthat energy-rich regions might secede hasreportedly made the government reluctantto increase energy taxes. Moreover, in someareas, unrest has made it physically impos-sible to collect taxes at all. Tax collectionhas also been affected by inadequate taxadministration.

Lessons and policy directionsA successful revenue turnaround in the

countries of the former Soviet Union will beintimately linked to success in otheraspects of reform. The recovery of outputshould help recapture at least some of the

lost tax bases—though the shift ofactivity into the private sector meansthat more dynamic collection proce-dures will have to be put in place.Likewise, the imposition of hard bud-get constraints on enterprises andthe consequent dampening of arrearswill help boost the effective taxbase—as will taming inflation. Thisdoes not mean, however, that policyimprovements should not be made tohelp reverse the revenue decline. Themain feature distinguishing indus-trial countries from countries withlow revenue ratios is the breadth ofthe former's domestic tax bases:exemptions and preferences are rela-tively few and compliance is rela-tively high.

Recapturing the tax base.Several policy initiatives need to betaken to broaden existing tax bases.

• Significant exemptions shouldbe eliminated in the core taxes—enterprise profit taxes, VATs, andexcises. In most countries of the for-mer Soviet Union, exemptions bene-fit agriculture and new technology,but these tend to be so broadlydefined that many unrelated activi-ties can qualify for them. Other cen-tral government exemptions aredesigned to favor energy—in pro-ducer countries by encouraging thedevelopment of the sector, and inconsumer countries by cushioning

the impact of the immense energy pricerises seen since 1991. Another importantgroup of exemptions worth considering forelimination are preferences for foreigninvestors, which may have been necessaryto attract venture capital to risky regionsbut should now be scaled back.

• The lost revenue from trade should bereclaimed. This should not have to be doneby raising tariff walls and obstructingstructural reform. Rather, efforts should befocused on ensuring that traders' transac-tions are fully captured in the domestic taxnet. Exporters and importers must payincome tax, and imports must be fullyliable for VAT, as well as for excises onsome groups of goods (notably energy,tobacco, alcohol, and perhaps vehicles). Asan interim measure (where trade arrange-ments permit), a modest but broad-baseduniform tariff would ensure that traderspaid at least some tax and would providethe information basis for ensuring thattraders were eventually incorporated intodomestic tax registers. As the domestic taxadministration becomes more sophisticated



and comprehensive in its coverage, the tar-iff can be phased out. Where depletable nat-ural resources are being sold, the paymentof royalties is an appropriate way forenergy producers to compensate the statefor the loss of national wealth.

• Finally, taxes should be extended to theprivate sector.

Institution building. Successful taxa-tion of the private sector is one of the mainchallenges facing policymakers. Newincome earners need to be identified andtheir incomes measured, and they need tobe persuaded to pay taxes. This requiresthe reform and development of many kindsof institutions, ranging from tax adminis-trations to accounting procedures and legalsystems, as well as the solidification of gov-ernments' legitimacy. Resources need to bedevoted to registering taxpayers, improv-ing audits, educating the public, removing

red tape, and designing and imposing real-istic and effective penalties for noncompli-ance. Governments also need to hireenough competent tax officials and insulatethem from corruption.

Supporting reforms should includedeveloping good accounting standards,eliminating inconsistencies and loopholesin legislation, providing adequate enforce-ment powers, and establishing trust in gov-ernment through the consistent, strict, andfair application of a transparent set of taxrules—which should change as infre-quently as possible once the basic reformshave been put in place. These reforms, ofcourse, will take time. As an interim mea-sure, levying simple presumptive taxessuch as small-business license fees mightbe considered, both for revenue reasonsand—as is true of tariffs—because it is aneasy way to bring new taxpayers into the

net and to compile information that willeventually allow them to be transferred tothe standard tax system. iF&Dl

This article draws on an earlier study byRichard Hemming, Adrienne Cheasty, andAshok K. Lahiri, 1995, "The Revenue Decline,"in Policy Experiences and Issues in the Baltics,Russia, and Other Countries of the FormerSoviet Union, ed. by Daniel A. Citrin and AshokV. Lahiri, IMF Occasional Paper No. 133(Washington). The material in this artickcovering 1994-95 was researched by/itendraModi.

Suggestion for further reading:Daniel Kaufman, 1994, "Diminishing

Returns to Administrative Controls and theEmergence of the Unofficial Economy,"Economic Policy, Supplement to Issue 19(December), pp. 51-70.

J u s t R e l e a s e d — N i n e t e e n t h E d i t i o n

WORLD DEVELOPMENT REPORT 1996:FROM PLAN TO MARKET

World Development Report 1996: From Plan toMarket steps back from the extraordinary array

of recent events and policy changes in 28 former central-ly planned economies—in particular, those in Centraland Eastern Europe and the newly independent states ofthe former Soviet Union, along with China andVietnam—to ask what we have learned about the ingre-dients of any successful transition and how they shouldbe pursued.

The Report addresses the initial challenges of transi-tion, how they have been met in different countries, andhow they may be met in others. It also looks beyond theearly reforms and analyzes the longer-term agenda ofconsolidating them by developing the institutions and

policies which will help new systems to develop andprosper over time.

One of the Report's key messages is that consistentpolicies, combining liberalization of markets and trade,new entry, and reasonable price stability, can achieve agreat deal—even in countries lacking clear propertyrights and strong market institutions.

Available June 28, 1996. Published for the World Bankby Oxford University Press. Stock no. 61107F (ISBN 0-19-521107-3) $22.95. Softcover. Stock no. 61108F(ISBN 0-19-521108-1) $45.95. Hardcover. Translationsforth-coming: Arabic, Chinese, French, German,Japanese, Portuguese, Russian, and Spanish.

World Bank For payment in local currency, contact your World Bank Distributor. In the U.S. contact: Trie World Bank, P.O. Box 7247-8619, Philadelphia, PA 19170-8619. Phone: (202) 473-1155, Fax: (202) 522-2627. Shipping and handling: US$5.00.For airmail delivery outside the U.S., add US$8.00 for one item plus US$6.00 for each additional item. Payment by US$check drawn on a U.S. bank payable to the World Bank or by VISA, MasterCard, or American Express.

1206

Visit the World Bank's Internet Home Page! http://www.worldbank.org Click on Publications



http://www.worldbank.org

The ImplicitPension Debt

C H E I K H K A N E A N D R O B E R T P A L A C I O S

Underfunded pension plansare a serious problem forindustrial and developingcountries alike. Because thisimplicit pension debt hasimportant macroeconomicimplications, governmentsneed to tackle the problem assoon as possible.

IN CONTRAST to conventional publicdebt, both internal and external, pub-lic guarantees that imply large liabili-ties are not assessed and publicly

reported by the authorities in most coun-tries. These include the obligations implicitin guarantees of bank deposits, rural creditprograms, home mortgages, and studentloans. The difficulty of both measuring andmonitoring these hidden liabilities hascaused concern about their macroeconomiceffects.

One of the largest of these liabilities iscreated by a government's promise to pro-vide income support to some of its citizensduring retirement. These pension obliga-tions fall into two broad categories: guaran-tees of private pension schemes and directpromises to pay pensions to individuals.

Private pension guaranteesPrivate pension guarantees can take the

form of insurance against the default orfailure of private company pension funds,such as is provided by the US PensionBenefit Guarantee Corporation. They canalso include partial or complete protectionagainst rising prices, as is provided inJapan and the United Kingdom. The impor-tance of these guarantees varies accordingto the size of the private pension sector.Coverage in company or occupationalschemes ranges from practically zero inmany developing countries to almost uni-versal in such countries as the Netherlandsand Switzerland.

Another kind of guarantee was devel-oped in Chile, in the context of its pioneer-ing system of mandatory private accounts.In Chile, the government commits itself topay the difference between the accumulatedbalance in an individual's private account atthe time of retirement and the amount thatwould provide an annuity equivalent to theminimum pension. The minimum pensionhas hovered around one-fourth of the aver-age wage, and 20 years of contributions arerequired to qualify for the guarantee. Inaddition, the government assures thatannuities purchased through the schemewill be paid even if the life insurance com-panies from which they were purchased gobankrupt. Like Chile's scheme, the reformedArgentine and Colombian schemes nowguarantee that a minimum pension will beprovided to all workers who contribute for a

specified period to their privately managed,defined-contribution accounts.

Liabilities of public plansAlthough private pension guarantees

can lead to significant expenditures, thelargest pension obligations are almostalways those incurred by governments thatmake unfunded promises to pay defined-benefit pensions (see box). The concept ofthe implicit pension debt (EPD) recognizesthat workers and pensioners have claimson current and future governments that arenot unlike those of government bondhold-ers. Public pension promises are oftenaccorded contractual status in practice, ifnot in the legal sense, owing to the "insur-ance" terminology often used to describethe scheme. In fact, constitutional protec-tions have been invoked in many countriesto prevent governments from reducing thevalue of accrued pension wealth. In thisregard, it could be argued that social insur-ance programs impose more severe fiscalconstraints than other government pro-grams do.

Liabilities have grown rapidly with theproliferation of publicly managed, defined-benefit pension plans throughout the twen-tieth century. Today they are found in morethan 150 countries and probably covermore than one-half of the world's laborforce. The extent of their coverage in indi-vidual countries ranges from a tiny fractionof the labor force to almost everyone and isstrongly and positively correlated with the

Cheikh Kane,a national of Senegal, is a Fiscal Economist in the Labor Market and SocialInsurance Division of the World Bank's Poverty and Social DevelopmentDepartment.

Robert Palacios,a US national, is a Pension Economist in the Country Operations Divisionof Country Department II of the World Bank's Europe and Central AsiaRegional Office.



country's per capita income. (See Palacios,1996.) Many countries have multipleschemes that cover specific groups, mostcommonly the military and civil servants.

Although interest in the IPD is growingas countries re-evaluate their pensionschemes, estimates of its magnitude areincluded in only a few recent studies thathave focused principally on several mem-ber countries of the Organization forEconomic Cooperation and Development(OECD). (See van der Noord and Herd,1993.) Studies of developing or former cen-trally planned economies' pension liabili-ties are rare and have generally beenconfined to Latin America, where they havetended to focus on the unique circum-stances of particular pension reforms. (See,for example, Schmidt-Hebbel, 1995.) Theexisting empirical work does not use a con-sistent methodology, and almost no workcomparing different countries has beendone. The available estimates show thatunfunded pension liabilities are large evenin young, poor countries with limited pen-sion coverage and that they can reachalarming proportions in the demographi-cally advanced developing countries.

Timely, comparable estimates of the IPDcan and should be produced. These will beimportant inputs into the public debate onpension reform and will provide importantindicators that policymakers can use tocompare the results of different reform pro-posals. If this liability is expressed as a

stock, it can be readily compared with themore familiar domestic and foreign publicdebts. Finally, calculating the value of theIPD is a useful step when a country'sauthorities are considering the implicationsof ending a pay-as-you-go scheme. (Forinformation on pension schemes, see"Averting the Old-Age Crisis," by EstelleJames, Finance & Development, June 1995.)

Pension debt overhangSome interesting analogies can be drawn

between a country's external debt and itsimplicit pension debt. It is usually acceptedthat when it comes to servicing externaldebt, two aspects should be taken intoaccount. The first concern is the immediatecash-flow requirement as measured by thecountry's debt-service ratio (that is, theratio of scheduled debt-service paymentsfor the current year to annual exports ofgoods and services). The second, andlonger-term, concern is the extent of debtoverhang, as measured by the ratio of thediscounted present value of all future debt-service payments to annual exports ofgoods and services. Applying these con-cepts to pensions, one can measure cash-flow requirements by using the cost rate—that is, the ratio of pension expenditures tothe portion of the wage bill covered by thepension system. Because most pensionexpenditures are financed through payrolltaxes, the wage bill is the relevant tax base,just as exports are the relevant base for

external debt servicing. A cost rate abovethe statutory contribution rate would indi-cate that pension expenses could not be metby contributions alone and therefore wouldrequire transfers from general revenue, ahigher contribution rate, or a lower pensionbenefit. In short, the cost rate is to a coun-try's pension system what the debt-serviceratio is to its external debt.

The concept of external debt overhangcan be applied to pensions. The pensiondebt overhang would be captured by theratio of discounted future pension liabilitiesto the covered wage bill. This ratio isshown in columns (2) and (3) of the table,using two different discount rates to calcu-late net present value. The discount rate of8 percent is added to allow direct compari-son with the external debt overhang shownin column (5). Usually, an external debt-service ratio of 200 percent, calculated on apresent-value basis with an 8 percent dis-count, indicates a severe external debtproblem. For the countries in the table,except Peru and Venezuela, the pensiondebt overhang is well above 200 percent.These figures show in stark terms that pen-sion liabilities deserve far more attentionthan they have received so far. Indeed, acase can be made for making the calcula-tion of a country's implicit pension debt arequirement for any long-term assessmentof its fiscal policy.

When pay-as-you-go pension schemesbecome unsustainable, either on their own

Measuring defined-benefit pension liabilitiesMeasuring defined-benefit pension obligationsis a tricky business, and international account-ing standards for the private sector do notconcur on a single methodology. The complexparameters of the actuarial calculations can beand are manipulated toward various ends, withthe result that "defined benefits" are really notvery well defined. These ambiguities canfrustrate workers covered by the plans, firms'shareholders, and government regulators. Mostformulas involve multiplying a certain"accrual factor" by the number of years of aworker's service, and then multiplying theresult by an average of that worker's wagestoward the end of his or her career. If it isdifficult to calculate the private pension wealthaccumulated by an individual worker, it shouldbe even more difficult to evaluate the liabilitiesof all individuals working at a particular firmand to compare these with the assets of thepension plan. There are, however, strong legaland economic incentives for firms to makethese estimates. Firms need such informationto make informed long-term business deci-

sions, and these liabilities must be reported toregulators and tax authorities. In some coun-tries, regulations even require plan members tobe informed of their expected benefits.

For example, in the United States, thecomplex Employee Retirement InsuranceSecurity Act (ERISA) of 1974 and subsequentamendments set minimum and maximumfunding ratios along with rules about how topay off the liabilities of terminated plans. Theminimum funding requirements are supposedto protect the taxpayer from abuse of PensionBenefit Guarantee Corporation (PBGC) insur-ance while the maximum limit reduces possi-ble tax expenditures that could arise ascompanies "overfund." It also protects workers,because the PBGC insurance covers only aportion of each worker's accrued pensionrights when a pension scheme is terminated. Inaddition, since the mid-1980s, the FinancialAccounting Standards Board has required thatthe unfunded liability, which is to be calculatedin a specified manner, appear on a firm'sbalance sheet. In Germany and Japan, the

measurement of companies' liabilities isnecessary to satisfy criteria for tax-exempt"book reserves." In Japan, up to 40 percent ofthese liabilities are tax deductible.

Pension liabilities are financed using various"actuarial cost methods." This is the umbrellaterm used in the United States to describe "themethod of allocating the cost of a definedbenefit pension plan to each year of the plan'sexistence in an orderly fashion" (Archer, 1993,p. 123). Examples of the methods availableinclude pay-as-you-go, terminal funding (wherelump-sum contributions are made as theemployee retires that are equivalent to theestimated present value of the annuity), andbook reserve (where liabilities are entered onthe balance sheet of the plan sponsor). Thepay-as-you-go method is used by occupationalpension schemes in France, while book-reserveschemes are prevalent in Germany and Japan.The actuarial cost method will influence theway the unfunded liability is financed overtime, but not the value of the liability itself.

Finance & Development / June 1996 37


Selected countries: implicit pension debt during the early 1990s

IPD/GDP

(percent)

IPD/Wage bill External debt indicators(discount rate (discount rate (discount rate= 4 percent) = 4 percent) = 8 percent)

Stock/ PresentGNP value/exports

(1) (2) (3) (4) (5)

Latin AmericaUruguay 214Brazil 187Peru 37

(recognition bonds) (3)Venezuela 30

AfricaCameroon 44Senegal 27

AsiaChina 63

Europe

959591174(15)303

410467

414

713390113(7)

199

318324

295

592863

65

6564

19

287300436

195

253161

76

CroatiaHungaryUkraineTurkey

35021314172

739745779592

486549220388

672

51

1867

170

Sources: Authors' estimates for implicit pension debts (IPDs); World Bank, World Debt Tables(1994-95) for external debt indicators.

Note: The IPD estimates in this table are based on data for different years in the early 1990s. TheIPD calculation is based on a termination liability concept that measures the value, under certainassumptions, of the accrued pension wealth of contributors and pensioners at a given point in time.In contrast, some other studies calculate the present value of all future pension spending assumingthat the scheme continues indefinitely. The present value of debt-service payments is based on adiscount rate of 8 percent. For a detailed explanation, see Cheikh Kane and Robert Palacios, "TheImplicit Pension Debt: Concepts and Measurement," forthcoming in the World Bank's DiscussionPaper series (Washington).

...: Data not available.

or when the IPD is considered with otherpublic debt, governments must seriouslyconsider reducing the IPD. They can do thisby raising the retirement age and reducingthe statutory value of the pension as a pro-portion of the worker's wage (the replace-ment rate). In Ukraine, for instance,equalizing the retirement age at 65 (fromthe current levels of 60 for men and 55 forwomen) would reduce the pension debtoverhang by about 23 percent. By compari-son, under the Toronto terms for repay-ment of bilateral nonconcessional loans,low-income developing countries weregranted an average reduction of their exter-nal debt, in present-value terms, of about 20percent of their nonconcessional debts.Correcting the imbalances of pension sys-tems thus might very well bring more debtrelief than a typical debt-reduction scheme.

Don't ignore the IPDThe IPD is a fiscal burden that cannot be

ignored in low- and middle-income coun-tries, especially those whose economiesused to be centrally planned, or anywhereelse for that matter. In addition to the long-term balance of the pension scheme, the

IPD measures intergenerational transfers ofmassive proportions. Changing the param-eters of a country's pension system canhelp, but a more important objective wouldbe to move away from pension schemesthat entail large burdens that get heavier aspopulations become older. This can beachieved by moving toward fully funded,defined-contribution schemes.

Standardized IPD estimates would addan element of transparency to pay-as-you-go pension financing and provide a moreaccurate indicator of long-term fiscal com-mitments than estimates of implicit govern-ment debt alone. Once an acceptedmethodology for calculating the IPD isdesigned and applied, it can be comparedacross countries and over time, and viewedin relation to national income and the wagebill. Such calculations would also provideinternational lending institutions and gov-ernments with an objective indicator theycould use to measure the impact of suchpolicies as increasing the retirement age orrevising the benefit formula. Perhaps poli-cies expanding pension eligibility orincreasing pension benefits in many coun-tries over the last few decades would have

been reconsidered had the impacts of thesemeasures been shown using a widelyunderstood indicator. An IPD indicatormight also provide useful input into policydebates in India and other countries thatare now beginning to make unfunded pen-sion promises to their citizens. Measuringthe IPD is also an important means ofencouraging the spread of comprehensivepension reforms. Ironically, current pensionaccounting systems in many countriespenalize, to some extent, reformers whomake the IPD explicit by ignoring the trueimprovement in their long-term fiscal posi-tions that can result from pension reform.

Much research has taken the perspectiveof pension plan participants and looked atthe possible impacts of pension wealth, bothpublic and private, on saving rates and indi-vidual saving decisions. Feldstein and oth-ers have also examined the relationshipsbetween unfunded pension liabilities in theprivate sector and stock market prices onthe assumption that share prices shouldreflect these liabilities if the market recog-nizes their existence. Future research maylook at unfunded pension liabilities of gov-ernments in a similar way to see to whatextent pension debt influences their cost ofborrowing. To the extent that such a costexists, this would reinforce the case for pen-sion reform aimed at reducing the size of thepension debt. In the meantime, the implicitpension debt continues to grow as pay-as-you-go schemes expand and the world con-tinues to undergo demographic transition.

This article is based on a longer paper by theauthors, "The Implicit Pension Debt: Conceptsand Measurement," which is to be published in1996 in the World Bank's Discussion Paperseries.

Suggestions for further reading:Michael Archer, 1993, "Minimum Funding

Requirement," in ERISA: A ComprehensiveGuide, ed. by Martin Wald and David Kenty(New York: John Wiley).

Martin Feldstein and S. Seligman, 1981,"Pension Funding, Share Price and NationalSaving," Journal of Finance, Vol. 36(September), pp. 802-24.

Robert Palacios, 1996, International Patternsof Pension Coverage and Expenditures, forth-coming Working Paper (Washington: WorldBank).

Klaus Schmidt-Hebbel, 1995, Colombia'sPension Reform: Fiscal and MacroeconomicEffects, World Bank Discussion Paper No. 314(Washington: World Bank).

Paul van der Noord and Richard Herd, 1993,"Pension Liabilities in the Seven MajorEconomies," OECD Working Paper No. 1142(Paris).



IMF's New Data Standards Online

W HEN corporations want toissue stock on an exchange,they typically offer investorsa regular and timely flow

of corporate financial statements thathave been vetted using generally acceptedaccounting principles—otherwise, a stockexchange listing is not likely. Butwhen countries—particularly developingones—want to tap capital markets,investors are often left to wonder: Howreliable are the data used to analyze thecountry's economic prospects? Is themethodology very different from that usedby other countries? Are there reasons todoubt the integrity of the data? When willnew data come out?

These unanswered questions—reflectinga lack of uniformity in data provision—have long impeded the efficient operation ofcapital markets, raising the costs of doingbusiness. In an era of increasing globaliza-tion of financial markets, the need foranswers has taken on a new urgency—asunderscored by the turbulence in interna-tional markets in late 1994 and early 1995following the Mexican financial crisis.

For this reason, the Interim Committee ofthe Board of Governors—the IMF's advi-sory body—asked the IMF in April 1995 toestablish standards to guide countries inpublishing a regular and timely flow ofcomprehensive economic and financialdata. The goal is to help reduce surprisesfor markets and aid policymakers in imple-menting sound economic policies. One yearlater, a framework is now in place. And,before the year is out, the general publicshould be able to bone up on various coun-tries' data practices by tapping into an IMFbulletin board in cyberspace.

Laura Wallace,a US national, is a Senior Public Affairs Officerin the IMF's External Relations Department.

A two-tier standardThe new data publication standards con-

sist of two tiers: (1) a general standard,which provides guidance to all membercountries for publishing data; and (2) a spe-cial standard, which provides guidance tocountries participating in internationalfinancial markets, or aspiring to do so.

The details of the special standard, orsecond tier, were hammered out after exten-sive consultations with producers and usersof statistics—paying special attention to theneeds of capital markets. In late 1995, IMFstaff visited 23 countries and several inter-national organizations, and correspondedwith government officials in another 42countries. The initial proposal for the spe-cial standard was released to the public forfurther discussion in February 1996—thefirst time the IMF had ever issued a draftpolicy document for public comment.

As of April, countries wishing to sign upfor the special standard could do so,although they will have until the end of atransition period that finishes at the closeof 1998 to bring their statistical practicesfully into line with the standard, seekingtechnical assistance if needed. These coun-tries, expected to eventually include 40-60industrial and emerging market economies,will be required to provide the IMF with rel-evant "metadata"—a statistical term forinformation about data (e.g., measurementcharacteristics and calendars for release ofdata). The requirements of the firsttier—the general standard—are now beingreadied.

The two tiers have the same four dimen-sions, with the second tier setting toughernorms for the first dimension.

• Coverage, periodicity, and time-liness. The standards call for publishingofficial statistics that shed the most light onmacroeconomic policies and performance.These statistics cover four areas: the realsector (GDP, production, employment, andprices), the fiscal sector (governmentdeficits and their determinants), the finan-cial sector (money, credit, and related

variables), and the external sector (interna-tional reserves, exchange rate, debt, trade,and balance of payments). The standardsalso set norms for the frequency of compi-lation and the speed of dissemination (e.g.,second-tier participants must provide dataon international reserves on a monthlybasis within a week of the reference month).

• Access by the public. Given thatdissemination is an essential feature of offi-cial statistics as a public good, the stan-dards call for timely, nondiscriminatory,and ready access for all—including theadvance dissemination of release calendarsthat indicate when data will be made pub-licly available.

• Integrity. Official statistics mustenjoy the confidence of users. The stan-dards include elements such as identifyingthe group of government officials who haveadvance access to the data before officialrelease.

• Quality. Although the quality ofstatistics is difficult to define and judge, thestandards require proxies of data qualitythat can be monitored (e.g., data for cross-checks) to help users make such judgments.

Electronic bulletin boardHow will the public gain access to all this

information? By the end of the summer of1996, the IMF plans to launch a data dis-semination bulletin board, which will bemaintained at a World Wide Web site onthe Internet. The bulletin board will pub-licly identify countries in the second tier,provide the relevant metadata, and let usersknow where they can obtain data for indi-vidual countries.

Market participants have indicated thatthe metadata should be extremely useful, asthey often have trouble finding out releasedates for data, the reasons for delays, andthe methodology used to compile the data.The hope is that users will serve on thefront line in monitoring compliance withthe new standards—perhaps even calling acentral bank directly if data are inexplica-bly late.[



Forget Convergence: Divergence Past, Present,and Future

Looking for evidence ofincome convergence amongthe world's nations hasbecome a fashionable pursuit.Far from narrowing, the gapbetween the incomes of therich and poor countries hasgrown markedly and is likelyto widen further.

C ONVERGENCE—the tendency forpoorer countries to grow fasterthan richer ones and, hence, fortheir levels of income to con-

verge—has recently received a great deal ofattention in the economics literature. Alongwith "globalization" and "competitiveness,"the theme of "convergence" has spilled overinto public discussions of policies andprospects for developing countries. Well,forget convergence—the overwhelming fea-ture of modern economic history is a mas-sive divergence in per capita incomesbetween rich and poor countries, a gapwhich is continuing to grow today.Moreover, unless the future is different inimportant ways from the recent past, wecan expect this gap to grow ever wider.

Divergence pastThe very feature that marks the begin-

ning of modern economic history alsoimplies a major increase in the difference inper capita incomes across nations. Call itthe industrial revolution, the emergence ofmodern capitalism, or the take-off into sus-tained growth, at some point in the late

L A N T P R I T C H E T T

nineteenth century the annual growth ratesof the now-rich industrial countries acceler-ated from historically low levels (0.5 per-cent or less), to 1-2 percent per year. Thefact that this acceleration was not univer-sal, or even widespread, implies that thegap between rich and poor countries'growth rates widened and the gulf betweentheir per capita incomes—which was prob-ably already wide—began to grow.

Given different exchange rates and dif-ferent mixes of tradable and nontradablegoods among countries, how can we com-pare income levels? We can compare themby using purchasing-power-adjusted mea-sures of income. One important feature ofthis adjustment of incomes is to account forthe relative cheapness of nontradables inpoorer countries. Using a purchasing-power-parity measure substantially raisesthe estimate of income of poor countries rel-ative to their income expressed in US dol-lars at official exchange rates—typically bya factor of 3 to 5, depending on particularcountries' prices.

Measured in purchasing-power-parityterms at 1985 prices (P$), the ratio of theper capita income of the richest country(the United States) to the average per capitaincome of the poorest countries grew fromaround 9 (P$2,181 compared with P$250) in1870 to over 50 (P$16,779 compared withP$325) in 1960. In absolute terms, theincome gap between countries grew evenmore, expanding more than eightfold overthis period. The average absolute differencebetween the income of the richest countryand the incomes of all others was aboutP$l,500 in 1870 but, by 1960, this gap hadgrown to P$12,662.

Alert readers may wonder how theincomes of poor countries in 1870 can beestimated. Most of the industrial countries

have roughly comparable estimates of GDPper capita extending back to 1870. In con-trast, GDP estimates for most developingcountries began only in 1950 or 1960.Moreover, most did not even exist as inde-pendent countries with their present bound-aries in 1870. How then can we venture toguess what the evolution of income gapsfrom 1870 to 1960 might have been?

It can be done. Suppose that we onlyneeded to estimate the change in the gapbetween the richest and the poorest coun-try between 1870 and 1960. To do this wewould need the income of today's richestcountry in 1870 and 1960 (P$2,181 andP$9,900, respectively), and the averageincome of today's poorest country—Ethiopia—for those years. The data for theUnited States are available, as is the incomeper capita of Ethiopia in 1960 (P$260).What is missing is Ethiopia's per capitaincome in 1870. But we are not stuck,because if we can make a sufficiently goodguess at how low incomes could possiblyhave been in 1870, we can work backwardby a process of deduction to estimateincome divergence for all countries.

Since we know the growth rate of theUnited States over the entire period, we alsoknow that the ratio of US income per capitain 1960 to its level in 1870 is about 4.5. IfEthiopia grew faster than the United Statesover this period, then the ratio of Ethiopia'sper capita income in 1960 to its level in 1870would have to be larger than 4.5. But, if theratio between Ethiopia's 1960 income andthe lowest it could plausibly have been in1870 is smaller than 4.5, then we know thatEthiopia in fact grew more slowly than theUnited States and, hence, that there hasbeen a divergence in per capita incomesbetween the world's richest country and thepoorest countries. Moreover, applying this

Lant Pritchett,a US national, is a Senior Economist in the Poverty and Human Resources Division of the Policy Research Department of the World Bank.



methodology to other countries, we canmake rough guesses of the average magni-tude of divergence in the cross-national dis-tribution of income.

In "Divergence, Big Time," a back-ground study for the World Bank's WorldDevelopment Report 1995, five differentmethods were used to estimate the lowerbound of incomes: the lowest recordedincomes in the data available for 1960-90;current estimates of poverty lines (the levelof income that defines poverty in a givencountry); incomes required for nutritionaladequacy; the relationship between in-come, mortality, and demographic sustain-ability; and known historical estimatesof income. Using these five distinctapproaches we arrived at a figure of P$250as a reasonable guess at the lowest levelthat income could have reached in 1870.

But using this lower bound of P$250, weextrapolate incomes backward from percapita incomes observed today. For exam-ple, assume that every country grew at thesame rate as the richest country (of course,to generate convergence, poorer countrieswould need to have grown even faster). Butit is simply impossible for the countriesconsidered poor today to have grownthat fast on average since 1870, as theassumption of equal growth rates—orequivalently, of no divergence—implies im-possibly low incomes for those countries in1870. Therefore, for historical growth ratesto be compatible with the current level ofincome in poor countries, growth musthave been considerably slower for the poorcountries than for the leaders. Even with-out historical data, we know that there hasbeen massive divergence in income levelssince 1870 (see chart).

Divergence presentDivergence is not confined to the past

century. For relative income levels to con-verge, poor countries must grow fasterthan rich countries. Between 1960 and1990, income grew, on average, 2.6 percentper year in the Organization for EconomicCooperation and Development (OECD)countries, and 1.8 percent in other coun-tries. Among the poor countries, 43 percenthave grown more slowly than the slowest-growing OECD country, and 70 percenthave grown at a slower rate than themedian for OECD countries. Since poorcountries are growing more slowly on aver-age, the dispersion in incomes among coun-tries (as measured by the standarddeviation—the dispersion of observationsaround an average measure—of the naturallogarithm of per capita income) between

The income gap has grown manyfold since 1870Per capita GDP in 1985 purchasing power parity dollars 1

Source: Lant Pritchett, 1995, "Divergence, Big Time," World Bank Policy Research Working PaperNo. 1522.

1 Expressed in logarithms.

1960 and 1990 increased by 28 percent(from 0.86 to 1.1) and the ratio of theincomes of the richest to the poorest coun-tries rose by 45 percent just since 1960.

Especially given the recent record ofdeveloping countries, it is very difficult tounderstand an upsurge of interest in con-vergence. During the Great Depression ofthe 1930s, income fell by 32 percent in theUnited States and by 19 percent in France,two of the hardest-hit industrial countries.Since 1960, more than 60 percent of thedeveloping countries have experienced atleast one episode during which incomes fellmore than the decline recorded in France,and almost one-third of developing coun-tries have suffered an episode of incomereduction larger than that which occurredin the United States. Moreover, in manydeveloping countries, the decline in incomehas not been reversed. Estimates of incomein 1990 show that 72 percent of developingcountries still fell short of their own peakincome level and two-thirds were not within5 percent of their peak. In discussionsabout developing countries, it is not sur-prising that the 1980s are often referred toas the "lost" decade, but never as the "con-vergence" decade.

Divergence future?What would happen if current growth

rates in developing and industrial countrieswere to persist? How quickly would devel-oping countries overtake the United Statesin per capita income? Using the data for

the 93 developing countries for which theWorld Development Report 1995 reportsincome growth rates for 1980-93, we calcu-lated how long it would take various coun-tries to achieve three levels of income: theirown peak income level; the current incomelevel of high-income countries; and theaverage future income of high-income coun-tries, assuming that high-income countriesalso continue to grow.

First, more than half of the developingcountries had negative growth during1980-93. These countries are not gainingon anything—their incomes are convergingonly on the floor of subsistence. Unlesstheir growth rates accelerate, they willnever reach even their previous peaks. (Thereported data are, if anything, optimisticabout the number of countries with nega-tive growth, as many of the countries thatdo not report data fail to do so because ofinternal and external strife.)

Second, many developing countries hadpositive growth rates during 1980-93, butin more than four-fifths of these countriesgrowth rates were still lower than theaverage (2.2 percent) registered by the high-income countries. Moreover, many devel-oping countries grew slowly after sufferingrecessions during the 1980s. Against thisadmittedly pessimistic background andassuming unchanged growth rates, ifBrazil, for example, were to grow annuallyonly at its 1980-93 pace of 0.3 percent, itwould take 33 years for the country toregain its own previous income peak, and



487 years before it achieved the currentincome level of the high-income countries.

Third, a few developing countries wereactually "converging," that is, they weregrowing faster than the United States.When are these lucky "convergers" going toovertake the United States? India, for exam-ple, registered an annual average growthrate of 3 percent between 1980 and 1993. IfIndia could sustain this pace for another100 years, its income would reach the levelof high-income countries today. And, ifIndia can sustain this growth differentialfor 377 years, my great-great-great-great-great-great-great-great-great-great-greatgrandchildren will be alive to see India'sincome level "converge."

Fourth, since 1980 only 10 developingcountries have had growth rates that weremore than 1 percentage point higher thanthe average for high-income countries.These countries can be said to be converg-ing rapidly to high-income country levels. Ifthey can maintain this pace, these countriescan look forward to attaining today's levelof income in high-income countries within acouple of generations (50 years in the caseof Indonesia), and they would actuallyreach the future income level of the high-income countries in less than a century.

Does convergence ever occur?Of course, what will happen "if current

trends persist" is not really a prediction ofthe future. First, the 1980s were an excep-tionally bad decade and things may getsomewhat better for poor countries becauseof improved global conditions. Second, thefuture will be determined by policy actionstaken today, and there is no iron law thatdictates divergence. Convergence can hap-pen. There are several instances of absoluteincome convergence among deeply inte-grated economies and there are examples ofvery rapid growth among countries thatwere quite poor.

The best-documented examples of abso-lute convergence are those of economiesthat have achieved deep integration. Thisincludes regions within nations (particu-larly in Europe, Japan, and the UnitedStates), the European Union countries—which have experienced absolute conver-gence—and, perhaps, all OECD countries,as the European countries as a group havemade some gains on the United States inthe postwar period.

Even where deep integration has beenachieved, three points can be made. First,by any absolute standard, the rate of con-vergence within Europe, Japan, and theUnited States has been slow. Robert Barro

and Xavier Sala-i-Martin (1995) haveargued that within these countries, regionalconvergence occurs at a near-uniform rateof about 2 percent per year, meaning thatonly 2 percent of the income gap is elimi-nated each year. Second, the integrationneeded to achieve even that slow pace maywell be very deep. In the United States, forinstance, in any given five-year period, 10percent of the population moves acrossstate borders.

Some countries do not show regionalconvergence. For instance, the data pre-sented in "Regional Economic Growth andConvergence in India," by Paul Cashin andRatna Sahay, Finance & Development,March 1996, show substantial absolutedivergence among the states of India, withthe dispersion of the logarithm of incomesincreasing from 0.29 to 0.33. In anotherexample, China, the evidence for incomeconvergence is mixed, but certainly doesnot show any uniform tendency towardabsolute convergence over time.

Third, the mechanisms that lead toregional convergence may not be applicableto countries. In the United States, for exam-ple, from 1930 to 1970, there is evidence ofconvergence because states like Californiahad high initial incomes and low per capitagrowth, while states like Mississippi hadlow initial incomes and high per capitagrowth. However, one should not ignore thefact that population growth in Californiawas 10 times higher than in Mississippiand hence the growth of absolute (not percapita) output in California was substan-tially higher than in Mississippi. No onereally thinks that California's economy wasoutperformed by Mississippi's.

A second type of absolute convergenceobviously occurs when countries that startout behind experience truly rapid growth.That a country has to be behind to gain onthe leader has led economists from Hume toGershenkron to expect that poor countrieswould gain on the leaders. But can doesn'tmean will.

What can we learn from the examples ofJapan and Korea and, most recently, China?If anything, they demonstrate the possibil-ity of "policy-conditional" conditional con-vergence. That is, if a country's initialincome is low and its government pursuesgrowth-oriented policies, then very rapidgrowth rates may be possible. Jeffrey Sachsand Andrew Warner (1995), for instance,have recently suggested that countries thatadopted such policies did in fact exhibitvery strong conditional convergence, whilethose poor countries that did not adoptthem did not display any conditional con-

vergence. However, it is important to notethat only 12 developing countries, usingtheir criteria, did adopt growth-orientedpolicies. This suggests that the likelihoodof having good policies was lower thepoorer a country might be, and that thisstrong "policy-conditional" conditional con-vergence is compatible with absolute diver-gence and very weak "unconditional"conditional convergence.

Why focus on convergence?If the divergence of incomes is obvious in

past and present data, and is a worrisomepossibility for the future, why has conver-gence received so much attention?

The first good reason that convergence isin the news is that even though a largenumber of poor countries may be fallingbehind, the two largest countries—Chinaand India—have been doing well. Thepopulation-weighted average of incomegrowth over 1980-93 for all poor countriesis 3.7 percent, but when China and India areexcluded, that average falls to 0.1 percent.Obviously, the fact that the world's mostpopulous country is also its most rapidlygrowing has significant implications forboth rich and poor countries. But theseimplications are specific to China—they arenot an example of a more generalized expe-rience of convergence.

The second good reason for the attentionto convergence is its importance for theresurgent economic literature on models ofeconomic growth. Many economists arguethat a critical empirical hypothesis for dis-tinguishing a new generation of endoge-nous growth models that predict steadystate differences in growth rates from theolder Solow/Swann neoclassical models(augmented for human capital) is whetheror not there is conditional convergence. Butboth types of growth models are capable ofpredicting absolute divergence in per capitaincomes (as they, of course, must to retaineven surface plausibility). Whatever its the-oretical importance for growth models, thedebate boils down to whether the observedabsolute income divergence is attributedeither to nondiverging fundamentals thatcause permanent differences in growth(endogenous) or to conditional convergenceto divergent levels of income (exogenous).

Unfortunately, convergence has alsoreceived attention because many peoplehave misunderstood the concept of condi-tional convergence. In this context, "condi-tional" has a very specific econometricmeaning. Conditioning in this technicalsense means extracting from the differencesin actual growth rates across countries the



effects of other variables, particularlyinvestment in physical and human capital.Thus, while growth rates are higher amongthe rich countries, growth rates conditionalon variables like the initial level of school-ing and rate of investment are lower. Butsince initial schooling and investment ratesare themselves higher in the rich countries,this "conditional" convergence is perfectlycompatible with continued absolute diver-gence.

An example might be instructive.Suppose we tried to explain people's weightgain with a model in which weight gains orlosses are predicted based on one's weightlast year and one's height. If an individual'sweight fluctuates around a more or less sta-ble level that depends on height, then onewill find convergence of weight, conditionalon height. People who are thinner than theirlong-term average will, on average, gainweight and those heavier will, on average,lose weight. This does not imply that overtime everyone will weigh the same. The dis-tribution of weight across individuals willremain exactly the same, irrespective of thespeed of "conditional" weight convergence.

There are other, not so good, reasons forattention to convergence in the richer coun-

tries. First, since horse race metaphors ofeconomic competition dominate the think-ing of policymakers, they tend to worryonly about the horse just behind them, notabout what is happening at the back of thepack. There is no question that Germanyand Japan have gained on the United Statesin the postwar period, but this has nothingto do with the prospects for poor countries.Second, there is a near-universal tendencyto give more political attention to domesticjobs lost to imports than either to jobs thatare lost to export jobs forgone or even tojobs gained from exports. The import-competing jobs "lost" to Korean imports,for example, are politically more visiblethan the export jobs "lost" to the collapse ofinvestment in Latin America in the 1980s.

ConclusionThere are three good reasons not to

worry about convergence. First, it justhasn't happened, isn't happening, and isn'tgoing to happen without serious changes ineconomic policies in developing countries.Second, casual talk of "convergence" con-veys the wrong impression; there is nothingautomatic or easy about economic develop-ment. Rapid growth is not the result of

being poor—it is the result of creating a setof policies that facilitate rapid growth. Thepolicy environment that developing coun-tries need to establish rapid growth anddevelopment is difficult to achieve, as is evi-denced by the fact that so few have done so,and there is no "advantage to backward-ness" in this endeavor. Third, talk of con-vergence, especially in the industrialcountries, implies that their real concern isto protect themselves from the "converg-ing" poor countries when exactly the oppo-site is the case. Given the facts, more, notless, concern for the promotion of economicdevelopment and acceleration of growth inpoor countries is in order. IF&D

ReferencesRobert Barro and Xavier Sala-i-Martin,

1995, Economic Growth, (New York: McGraw-Hill).

Jeffrey Sachs and Andrew Warner, 1995,"Economic Reform and the Process of GlobalInterpretation," Brookings Papers on EconomicActivity: 1, pp. 1-118.

Lant Pritchett, 1995, "Divergence, BigTime," World Bank Policy Research WorkingPaper No. 1522 (Washington).

Emerging Stock Markets Factbook 1996

ttlFC International finance Corporation

1996. 300 pages. Order no. 13649F. US$100.00

New York, London, and Tokyo are still the world's financial centers,and will remain so for years to come. But when it comes to

dynamic growth, cities like Shanghai, Bombay, and Santiago are thenew centers of capitalism.

The International Finance Corporation's (IFC's) Emerging StockMarkets Factbook pulls together market data on over 60 of these emerg-ing stock markets. The Factbook is an invaluable reference for institu-tional investors, investment bankers, and academics, providing in a sin-gle volume time-series of fundamental market data and market analysis.

"We at Templeton require a great deal of reliable information aboutemerging markets in order to manage the Templeton Emerging MarketsFund, Inc. IFC's Emerging Stock Markets Factbook is an excellent sourcefor these purposes." Dr. Mark Mobius, Templeton International,Singapore.

"The Emerging Stock Markets Factbook is an essential tool for allresearchers who deal with emerging markets. Its breadth and depth ofcoverage are the best in the field." Professor Robert E. Levin, ColumbiaUniversity.

World BankFor payment in local currency, contact your World Bank Distributor. In the U.S. contact: The World Bank, P.O. Box 7247-8619, Philadelphia, PA 19170-8619. Phone: (202) 473-1155, Fax: (202) 522-2627. Shipping and handling: US$5.00.For airmail delivery outside the U.S., add US$8.00 for one item plus US$6.00 for each additional item. Payment by US$check drawn on a U.S. bank payable to the World Bank or by VISA, MasterCard, or American Express. 1207

Sample the Factbook on the World Wide Web! http://www.worldbank.org Click on Publications


EmergingStock

MarketsFactbook

1996


http://www.worldbank.org

Secured Transactions:The Power of Collateral

H E Y W O O D F L E I S I G

In many developing countries,businesses are unable to getlow-cost, long-term loans fromprivate lenders to financeinvestment projects. Reformsthat make it easier for borrow-ers to use movable property ascollateral would give comfortto lenders, stimulate invest-ment, and boost productivityand growth.

T HE FIRST question any privatelender asks is, "How do I get mymoney back?" Given the risksinvolved in lending—borrowers

may be unlucky, unwise, or dishonest—what conditions give comfort to lenders?

Two basic loan transactions haveevolved in private markets: unsecured andsecured. When a borrower offers an unse-cured promise to pay, the lender must relyon the borrower's reputation, or trust thatthe borrower will pay back the loan inorder to have access to future loans. Insecured transactions, promises to repay arebacked by collateral that lenders can seize

and sell in the event loan payments are notmade as agreed. Collateral may be realestate or personal property—tangible per-sonal property such as inventory, equip-ment, livestock, and tractors; or intangiblepersonal property such as unsecuredaccounts owed to merchants (accountsreceivable) and secured agreements (chattelpaper). Secured transactions have advan-tages for borrowers and lenders alike—transaction costs are lower and lenders donot need to gather as much informationabout borrowers.

The issue of collateral is one of great eco-nomic importance. When borrowers cannotuse their assets as collateral for loans andcannot purchase goods on credit using thegoods themselves as collateral, interestrates on loans tend to be higher to reflectthe risk to lenders. In many developingcountries, where legal and regulatory con-straints make it difficult to use movableproperty as loan collateral, the cost of loansmakes capital equipment more expensivefor entrepreneurs relative to their counter-parts in industrial countries; businesseseither postpone buying new equipment orfinance it more slowly out of their own lim-ited savings. Small businesses, in particu-lar, are hit hard by the scarcity of low-costfinancing, but the whole economy suffersbecause the lack of new investment damp-ens productivity and keeps incomes down.Estimates put welfare losses caused by bar-

riers to secured transactions at 5-10percent of GNP in Argentina and Bolivia.

Legal barriersMovable property is widely used as col-

lateral in the industrial countries. Abouthalf the credit offered in the United States issecured by some kind of movable property:about two-thirds of bank loans are securedby either movable property or real estate,and nonbank institutions that lend againstmovable property—such as leasing andfinance companies—do almost as muchlending as banks.

In contrast, private lenders in developingcountries rarely make loans secured bymovable property unless at least one of twoconditions is satisfied: borrowers must ownreal estate that can be attached if they donot pay, or borrowers must place the mov-able property under the physical control ofthe lenders, as in a pawnshop or warehousefinancing. If neither of these restrictive con-ditions can be met, private lenders rarelymake loans secured by movable property.They may still make unsecured loans, butthese are likely to be smaller loans withhigher interest rates and shorter maturities.This phenomenon has been studied in abroad range of countries with different tra-ditions, income levels, macroeconomic out-looks, religions, and levels of urbanizationand industrial activity. The difficulties insecuring loans with movable property have

Heywood Fleisig,a US national, recently retired as Economic Advisor in the World Bank's Private Sector Development Department. He is now Director of Research at theCenter for the Economic Analysis of Law, located in Washington.



been observed in Africa, Asia, EasternEurope, and Latin America—in countrieswith legal and regulatory systems modeledafter the civil codes of continentalEuropean countries, as well as in Africanand Asian countries whose systems aremodeled after British common law.

What prevents the use of movable prop-erty as collateral in developing and transi-tion countries? Three obstacles stand out:

• The creation of security interests is dif-ficult, expensive, and uncertain.

• The perfection of security interests—the public demonstration of their existenceand the establishment of their priority—isnot effectively possible.

• The enforcement of security interests isslow and expensive.

These abstract notions can be under-stood more easily with an example.Compare the financing available to farmersraising cattle in Uruguay with that avail-able to their counterparts in Kansas, in theUnited States. Uruguay and Kansas havesimilar topographies and well-educatedpopulations interested in advanced tech-nologies and able to apply them, and bothare world-class exporters of beef cattle. InKansas, private banks view cattle as one ofthe best forms of loan collateral; this is alsothe view of the bank examiners at theFederal Reserve Bank of Kansas City.Banks with "cattle paper" are seen as solidwhereas banks with "exposure to farm realestate" are seen as risky. By contrast, inUruguay, because of flaws in the legalframework governing secured transactions,private banks and bank examiners preferreal estate as collateral for loans; they con-sider a pledge on cattle worthless as collat-eral. The unacceptability of cattle ascollateral applies to all types of transac-tions, including sales of cattle on credit,sales of cattle financed by third-partylenders like banks or finance companies, orworking capital loans for other purposesthat might be secured by cattle.

Creating security interests. First, itis difficult to create a security interest inUruguay. Suppose a private bank inUruguay were to lend against 100 cattleworth $200,000. Uruguayan law calls for aspecific description of the property that ispledged. A pledge against cattle mightidentify the individual cows pledged byname (Bessie, Elmer, etc.) or by the num-bers tattooed on them. The need to identifythe collateral so specifically undermines thesecured transaction because the bank mustensure that the cattle designated in thepledge are available to be seized in theevent of nonpayment—the lender is not

allowed to repossess a different groupof cows.

As a result, the supervision of such loansis costly. It would not be enough, as it is inCanada and the United States, for example,to verify simply that there are enough cattlein the farmer's field; in Uruguay, the loanofficer would have to verify that the cattlein the field are the ones specifically identi-fied in the pledge. The Uruguayan bankmight try to get around this problem byusing a more general description of the col-lateral in the pledge contract—say, "100calves." But, with a loan of one year's matu-rity, the calves would become cows, bulls,or steers, and the enforceability of the con-tract would be clouded. In Canada and theUnited States, however, a binding agree-ment can be written with a floating securityinterest in "$200,000 worth of cattle." More-over, in Uruguay, the bank would have toworry that the farmer might sell the cattlewithout notifying the bank, whereas aCanadian or US bank would have a contin-uing security interest in the proceeds of thesale and could automatically attach the pro-ceeds—whether they were placed inanother bank or used to buy a tractor.

Perfecting security interests.Second, Uruguayan lenders cannot easilyfind out whether prior and superior claimsexist on their security interest. In Canada,Norway, and the United States, for exam-ple, all security interests against propertyare registered; the registries are public andindexed by borrower, by description ofsecurity interest, and by other relevantinformation. Lenders can easily conduct athorough search to ensure that they haveidentified any outstanding security inter-ests; security interests that are not regis-tered have no legal standing. By contrast,the registry in Uruguay files security inter-ests in chronological order and does notindex them. The only way lenders can findout whether a security interest exists is ifborrowers inform them.

Enforcing security interests.Finally, repossession and sale of collateraltakes longer in Uruguay than in Kansas. InKansas, repossession and sale of cattletakes one to five days and can be con-tracted between private parties. Typically,judicial intervention or the action of gov-ernment officials is unnecessary. InUruguay, the process requires six monthsto two years. In the case of cattle, there is arisk that the collateral will die, disappear,or get sick. Not surprisingly, under theseconditions, lenders demand collateral thatis sure to outlast a lengthy adjudicationprocess—in other words, real estate.

Economic impactIn Kansas, the ease of creating a security

interest, the inexpensiveness and highdegree of confidence that can be attached tothe perfection of a security interest, and thespeed and low cost of enforcement explainwhy farmers can get private loans for alarge fraction of the value of their cattle atinterest rates close to the prime rate. Thedifficulty of using cattle as collateral inUruguay explains why Uruguayan farmerscannot get any financing for cattle.However, although both rich and poorfarmers are affected, rich farmers own landand have access to some credit by virtue oftheir real estate holdings; poor farmers,who are often tenants, have to use their ownsavings to finance the additional invest-ment required to raise their incomes. Butneither rich nor poor farmers have the easyaccess to credit enjoyed by their Canadianand US counterparts (and competitors inthe world beef markets).

With minor adjustments, the same storycan be told for Argentina, Bangladesh,Bolivia, Bulgaria, Mexico, and other devel-oping and transition economies, both urbanand rural, regardless of whether their legalsystems are based on civil or common law.And the same story can be told, with slightvariations, for lending against all movableproperty in developing countries, except forloans and leases secured by automobiles orloans secured by goods stored—under thecontrol of lenders—in warehouses andpawnshops. Barriers to using movableproperty as collateral block access to creditand make it difficult to obtain term financ-ing for investment. They also make it diffi-cult to reform banking systems burdenedby risky, unsecured loans. Lacking usablecollateral, developing countries do notenjoy the major benefit of financial mar-kets—the transfer of funds from saverswith limited investment opportunities toinvestors with insufficient savings tofinance profitable projects.

Access to credit. Even in the best ofcircumstances, poor people do not oftengo to banks for loans to finance smallequipment purchases. But, in industrialcountries, small-scale farmers and entre-preneurs can usually purchase equipment,livestock, or inventory on credit from mer-chants. In developing countries, however,merchants willing to sell on credit to poorcustomers whose reputation is good havetheir own problems gaining access tocredit: legal constraints on collateral pre-vent merchants from getting financingsecured by their inventories or accountsreceivable. If merchants want to extend



credit to microenterprises and small andmedium-sized businesses, they typicallymust do so out of their own capital.

Term finance. Although operators ofsmall industrial operations may find it pos-sible to finance "plant"—real estate—witha mortgage, it may be impossible for themto get medium-term loans for working capi-tal or equipment. The reason is not the loanmaturities—the same lenders happilymake longer-term loans secured by realestate—but the underlying collateral(equipment and inventory) that is unaccept-able to lenders. In the United States, mov-able property accounts for half of thenonresidential capital stock and two-thirdsof corresponding gross investment. Thedifficulty of financing this investment indeveloping countries greatly hampersindustrial and agricultural development.

Bank reform. In developing countries,well-regulated banks typically only makeloans secured by real estate or make unse-cured loans to those who give evidence ofowning real estate. Because the costs ofmortgaging property are usually quitehigh, most bank loans are actually unse-cured but made to borrowers who own realestate on the theory that, in the event ofdefault, banks can file a lien against thereal estate.

Most bank loans in industrial countriesare secured—70 percent in the UnitedStates, for example. The difference betweenindustrial and developing countries is strik-ing—nearly 90 percent of bank loans inArgentina are unsecured, and the figure issimilar for Bulgaria. Unsecured bankingsystems are necessarily more risky andprone to failure. And, because mortgagesare unregistered, a lender has no idea howmany times a borrower has used the sameproperty to indirectly back up a loan. If theborrower defaults, only the bank that files alien the fastest will be able to back up theloan. Inability to use anything else as collat-eral means that about half of the nonresi-dential private capital stock is inaccessibleto banks as security for loans.

Capital markets. The World Bankhas supported capital market reforms thatencourage the securitization of equipmentloans, real estate mortgages, and creditcard accounts receivable. Such mechanismscan change the conduits through whichfinancing takes place, improve the alloca-tion of capital, and promote competition inthe financial sector. However, their effec-tiveness is limited by the difficulty ofcreating, perfecting, and enforcing theunderlying secured transactions. At themost rudimentary level, the easiest securiti-

zation is one where an enforceable securityinterest is given in the underlying paper.Successful securitizations in Canada andthe United States are based on predictablecollection rates for underlying mortgages,chattel paper, leases, and accounts receiv-able. In countries where the underlyingloans cannot be collected or leases enforced,securitization offers only limited actuarialbenefits. And, if a government guaranteessuch securitizations, it risks accumulatinglarge uncollectible debts.

Leasing, a type of secured transaction,may facilitate repossession by not requir-ing proof of ownership. Even so, reposses-sion can be a lengthy process. Leasing isalso subject to problems of creation andperfection. When these problems are notaddressed, leasing primarily benefits thosewho can already borrow.

Seeking solutionsSome analysts have concluded that, if the

private sector is unwilling to provide credit,market imperfections justify the creation ofstate agencies that make loans. However,this strategy has severe limitations. First,state lenders are no more able to collectthese loans than are private lenders. Themain difference is that state lenders arewilling to make loans despite the risk of los-ing money. Not surprisingly, they have losta great deal. Second, because borrowersknow that these loans are hard to collect,such lenders tend to attract a different typeof client—one who specializes in gettingloans from the government, not one whospecializes in investments with highreturns. What typically takes place withstate agency lending programs is a greatdeal of lending, little repayment, and notmuch positive impact on economic growth.

To get around the difficulties of reposses-sion and sale of collateral, some lenderssimply seize and sell the collateral withoutthe sanction of law. Some leasing opera-tions disguise their underlying financialnature and pretend that their seizures arenot repossessions, hoping that the judicialsystem will not uncover the equivalence ofthe transactions. For large and valuablepieces of equipment, dealers in every devel-oping country tell tales of dispatchingarmed men and bribing police to recovermachinery at gunpoint.

In some countries, lenders use a post-dated check to convert the civil offense ofnonpayment into a criminal act. They maydemand a postdated check in the amount ofboth the loan and the interest. On the datethe loan is due, the lender requests pay-ment. If the borrower cannot pay, the lender

deposits the check in a bank. If the check isreturned unpaid, the lender can take thecheck, stamped "check without funds," tothe police station. In Bolivia, for example,writing a check without funds is prima facieevidence of a criminal act of fraud. The bor-rower may be arrested. If he or she fails toraise the funds, conviction is certain. (Thesentence in Bolivia for writing bad checks isabout four years.)

Lenders in the formal sector, obviously,cannot use such collection techniques; therisk of civil and criminal damages is toogreat. Therefore, the resources of the formalsector are not tapped for credit, and mov-able property remains the province of lend-ing techniques in the informal sector. Thisis costly for developing economies—facedwith the prospect of jail if they are unable torepay, business people tend to borrow lessand borrow only for operations with veryhigh returns—and costly for society,because incarcerating risk-taking entre-preneurs stifles development.

Governments can implement a numberof legal reforms to address the fundamentalproblems in credit markets that make itso difficult to secure loans with movableproperty:

• Changing the law to permit a greatervariety of security interests in a widerrange of transactions by a broader group ofpeople.

• Making registry records public,reforming state-operated registries, restruc-turing public registries to permit competi-tion, and privatizing registry services orallowing private registry services to com-pete with public ones.

• Speeding up enforcement and makingit cheaper, changing the law to permit pri-vate parties to contract for nonjudicialrepossession and sale, and, when possible,allowing private parties to contract forrepossession and sale without governmentintervention. iF&Dl

This article draws on the work of several lawyersand economists who have advised the WorldBank on a variety of issues related to the Bank'slending activity and economic and sector work inthe area of secured transactions.

46 Finance & Development I'June 1996


Stock Market Developmentand Corporate Finance Decisions

ASLI D E M I R G U g - K U N T A N D VOJISLAV M A K S I M O V I C

In developing countries, howis the growth of stock marketsaffecting corporate financingdecisions? Initially, stock mar-ket development tends to beaccompanied by higher corpo-rate debt-equity ratios andmore business for banks.

O VER THE past ten years, totalcapitalization of stock marketsworldwide has grown from $4.7trillion to $15.2 trillion; develop-

ing countries' share of this total has jumpedfrom less than 4 percent to 13 percent.These increases have been accompanied bythe liberalization of stock markets, espe-cially in the developing world. In a success-ful attempt to attract foreign portfolioflows, many developing countries haveremoved restrictions on foreign ownership,liberalized capital account transactions,and improved accounting and informationstandards. Portfolio flows of equity in-vestment to emerging markets haveincreased sharply in recent years, reaching

$35 billion in 1994, compared with $0.1 bil-lion in 1985.

The growth of equity markets in devel-oping countries has won the enthusiasticsupport of policymakers and expanded thefinancing options available to firms. But itraises a number of questions. How do firmsdecide whether to finance investment bydebt or equity? How does stock marketdevelopment affect the financing choices offirms? And how does it affect banks indeveloping countries?

Debt or equity?Finance theory tells us that, in the

absence of bankruptcy costs, corporateincome taxation, or other market imperfec-tions, the value of a firm is independent ofits financial structure. The theory is intu-itive—because a firm's value is determinedby real assets, it cannot be changed bypurely financial transactions. In otherwords, financial assets on the right side ofthe balance sheet have value only becauseof the real assets, including intangibles andgrowth opportunities, on the left side.Therefore, if markets are doing their job, itshould not be possible to create value byshuffling the paper claims on the firm's realassets. However, if there are imperfec-tions—such as taxes, underdevelopedfinancial markets, and inefficient legal sys-

tems—financial structure becomes rele-vant Firms must decide whether to issuedebt or equity securities to minimize thecosts entailed by these imperfections.

Existing theories have focused on twodifferent determinants of financing choicesmade by firms: "agency" theories stressconflicts of interest between owners, credi-tors, and managers; other theories stresstax consequences. Empirical evidenceshows that differences in the capital struc-tures of firms in industrial and developingcountries can be attributed to the potentialfor a firm's owners or managers to engagein opportunistic behavior (captured by fac-tors such as asset composition, liquidityconstraints, industry classification, andgrowth opportunities) as well as to the taxadvantages of debt financing in many coun-tries. However, these differences explainonly part of the cross-country variation incorporate debt-equity ratios. As shown inChart 1, this variation is considerable.

Stock marketsOne possible determinant of corporate

financing choices that theory has over-looked is the level of development of finan-cial markets, especially equity markets.Most of the finance literature assumes theexistence of liquid, well-functioning stockmarkets. However, economies without a

Asli Demirgii£-Kunt,a Turkish national, is a Senior Economist in the Finance and PrivateSector Development Division of the World Bank's Policy and ResearchDepartment.

Vojislav Maksimovic,a Canadian national, is an Associate Professor at the Business School ofthe University of Maryland.



ChartlCorporate debt-equity ratios

(average, 1980-91)

Sources: IFC's corporate finance data base and Global Vantage data.

well-functioning stock market may sufferfrom three types of imperfections.

First, if there is no stock market, or thestock market is not liquid, opportunities forrisk diversification are limited for investorsand entrepreneurs. Outside investors mayrequire a premium to acquire companystock that is traded on an illiquid market.The high costs of diversification mayinduce firms to avoid the use of financialmarkets and may influence the firms'investment decisions. Thus, firms maychoose less capital-intensive productiontechnologies that are subject to lower long-term risk, or they may invest less andremain smaller than if their shares werewidely held.

Second, in the absence of a well-function-ing stock market, firms are unable to opti-mally structure their financing packages.Usually, there are conflicts of interestbetween a firm's managers and its cus-tomers and suppliers as well as betweendifferent classes of investors in the firm.For example, firms with high levels of debtmay have increased their probability ofbankruptcy sufficiently that they may enterinto overly risky projects, thus harmingtheir creditors. Because debt financing cre-ates incentives to take greater risks, ahighly leveraged firm may not be able toobtain additional credit. In these cases, ifthere were a well-functioning stock market,

issuance of equity would mitigate theincentive problems, allowing the firm toborrow more.

Third, besides their role in supplyingcapital, stock markets play an importantinformational role. Well-functioning stockmarkets collect information about theprospects of firms whose shares are tradedand make it available to creditors andinvestors. By improving the flow of infor-mation about firms and simplifyingtakeovers, well-functioning stock marketsmay contribute to corporate control andthus lead to greater managerial compe-tency. Better corporate control and firmmanagement will, in turn, promote invest-ment and efficiency.

However, the effect of stock marketdevelopment on corporate financing deci-sions is ambiguous. Sudden access to awell-functioning stock market could have avariety of possible effects on corporatedebt-equity ratios. One possible outcome isthe substitution of outside equity, throughpublic offerings, for debt; in this case, thedebt-equity ratios of firms previously ableto issue only debt would decrease. Or aclosely held firm might open itself to publicownership by issuing shares and substitut-ing outside equity for inside equity, whichwould not affect the debt-equity ratio. Athird possibility is that the firm's owners'new ability to diversify risks would make

expansion more attractive; such an expan-sion could be financed either through addi-tional debt or equity. A fourth possibility isthat, by facilitating the flow of informationand improving corporate governance, well-functioning stock markets may lower thecost of raising capital. In this case, externalfinance—both debt and equity—wouldbecome less costly, although it is not clearwhich would increase more.

The effect of stock market developmenton corporate debt-equity ratios depends onthe initial level of stock market development.We examined corporate debt-equity ratiosin 30 industrial and developing countries(Chart 1). After ranking countries based onthe level of stock market development (mea-sured by the size and liquidity of stockmarkets), we divided the sample into threegroups of equal numbers of countries. Thefirst group has the least developed stockmarkets; the stock markets in the secondgroup are twice as developed; and the thirdgroup has the most developed stock mar-kets—almost four times as developed asthe first group. When less-developed stockmarkets double, in terms of size and liquid-ity, corporate debt-equity ratios increase by10 percent. (See Chart 2.) When a stockmarket quadruples in size and liquidity,however, corporate debt-equity ratiosdecline by 25 percent.

This finding suggests that, as a rela-tively undeveloped stock market begins todevelop in a given country, firms in thatcountry initially increase their debt-equityratios. Not only do they issue new equitybut they also borrow more. So, at earlystages of market development, improve-ments in information quality, monitoring,and corporate control may be large enoughto induce creditors to lend more. For thesefirms, debt and equity finance are comple-mentary. However, as stock markets con-tinue to develop, the ratio changes. Incountries with relatively developed stockmarkets, as the latter continue to develop,firms begin to substitute equity for debt.

Could this reflect other factors that deter-mine corporate financing decisions? Whilethe simple correlations between debt-equityratios and the level of stock market devel-opment are telling, they do not take intoaccount other possible determinants of cor-porate financial structure. In addition to thedifferences, identified in the corporatefinance literature, between firms, capitalstructures may be different across countriesbecause of differences in economic develop-ment, supporting institutions, tax treat-ment of debt versus equity, and level ofdevelopment of financial institutions.



Differences in growth rates capturedifferences in growth opportunitiesavailable to firms in industrial anddeveloping countries. All countries inChart 1, except Jordan and South Africa,experienced GDP growth in the 1980s.Some countries—especially Brazil,Mexico, and Turkey—experienced highrates of inflation in the 1980s.Differences in inflation rates can alsoexplain some of the cross-country varia-tion in debt-equity ratios. Because debtcontracts are typically written in nomi-nal dollars, the rate of inflation mayaffect the riskiness, in real terms, of debtfinancing.

The countries in the study also presenta wide range of economic development—GDP per capita for 1991 ranges from$359 (Pakistan) to $27,492 (Switzerland).Income is a good indicator of institu-tional development. Although the generaltrend has been toward liberalization,markets are much more heavily regu-lated in most developing countries; thereare often fewer protections for investors;accounting standards are inadequate;and governments are, in general, moreactive in business affairs. The efficiencyof the legal system and the ability toenforce contracts are also highly corre-lated with the level of income.

Countries also vary with respect totheir tax treatment of interest income,dividends, and capital gains. In mostindustrial countries in the sample—forexample, Japan, the United Kingdom, andthe United States—interest payments havetax advantages. However this is true foronly two developing countries in the sam-ple: India and Korea. In the others, the nettax burden is generally lower on equityincome.

Finally, the level of development of finan-cial institutions, especially banks, is alsoimportant in provision of credit and corpo-rate financing decisions. Differences inbank development also reflect differences inlegal structure. For example, different com-binations of financial intermediaries havebeen developed by European countrieswith universal banking than by industrialcountries with regulatory restrictions thatsegregate banking and commerce (theUnited States, for example). Even thoughthe overall size of the financial system maybe similar across these countries, thefinancing decisions of firms may reflectthese structural differences.

Multiple regression procedures suggestthat our findings on stock market develop-ment and leverage hold even after control-

Stock market development in terms of sizeand liquidity

Sources: IFC's corporate finance data base andGlobal Vantage data.

Note: The 30 countries in the study are groupedaccording to the level of stock market development.The index of stock market development is based onsize and liquidity and is a means-removed averageof the ratios of stock market capitalization to GDP,value traded to GDP, and value traded to stockmarket capitalization.

1 Least developed stock markets: Belgium, Brazil,Finland, France, Italy, New Zealand, Pakistan, Spain,Turkey, and Zimbabwe.

2 Developed stock markets: Australia, Austria,Canada, India, Jordan, Mexico, the Netherlands,Norway, Sweden, and Thailand.

3 Most developed stock markets: Germany, HongKong, Japan, Korea, Malaysia, Singapore, SouthAfrica, Switzerland, the United Kingdom, and theUnited States.

ling for these other determinants of corpo-rate financing decisions.

Size matters. It is likely that the roleplayed by the market in gathering and dis-closing information may be more impor-tant for large firms because their stocks aretraded more often and are followed bymany analysts. Small firms may not benefitas much from stock market development, atleast initially, because their access may belimited by high fixed issuance costs. Eventhe stock of small firms that are listed on anexchange may not be traded as often as thestock of larger firms, since it may be morecostly for traders to acquire informationabout the prospects of small firms.

An examination of the quartiles com-posed of the smallest and the largest firms(measured by asset size) in each countrydemonstrates that debt-equity complemen-tarity in developing markets is indeeddriven by large firms. The findings suggestthat the development of a stock market ini-tially affects the financial policies of onlythe largest firms. In countries where stockmarkets do not play a significant economicrole, stock market development permits

large firms to increase their leverage.However, for large firms in countrieswith more developed markets, furtherstock market development is associatedwith lower debt-equity ratios.

Complementarity of banksStock markets serve important func-

tions even in economies with well-devel-oped banking sectors. Because stockmarkets provide a means of diversifyingrisk, mitigate conflicts of interest amongdifferent creditors, and improve infor-mation flow and corporate governance,equity and debt financing are, in gen-eral, not perfect substitutes for eachother. This is especially true in coun-tries with developing stock markets. Inthese countries, although developmentof stock markets makes more invest-ment feasible, new equity sales are notthe only source of finance for invest-ment. By providing better informationand decreasing monitoring costs forinvestors and financial intermediaries,stock markets lower the costs of bothexternal debt and external equity. Someof the new investment stimulated bystock market development is financedby new bank loans and bond sales.Thus, in the early stages of stock mar-ket development, equity issues tend tocomplement rather than replace banklending and bond issues.

In many developing countries, banksfear that the volume of their business willdecrease as stock markets grow. However,analysis suggests that initial improvementsin the functioning of a developing stockmarket produce higher debt-equity ratiosfor firms, and thus more business forbanks. In countries with developing finan-cial systems, stock markets and banks playdifferent, but complementary roles. Policiesundertaken to develop stock markets neednot affect existing banking systemsadversely. Stock markets and banks can bedeveloped simultaneously. iF&Dl

This article is based on two papers by theauthors: "Capital Structures in DevelopingCountries: Evidence from Ten Countries,"World Bank Policy Research Working PaperNo. 1320 (Washington, 1994); and "StockMarket Development and Firm FinancingChoices," World Bank Policy Research WorkingPaper No. 1461 (Washington, 1995). Thesepapers were part of a World Bank researchproject, "Stock Markets, Corporate Finance, andEconomic Growth," organized by Asli Demirguc-Kunt and Ross Levine.


Chart 2

Leverage and stock marketdevelooment M 980-91)


Urbanization: The Challengefor the Next Century

Chart 1

Urban population growth(billions)

Chart 2

Population subject towater srarcitv

them less capable of supportingrural populations and thus addingto the pressures for urban migration(Chart 2). Air pollution alreadyexceeds health standards in manymegacities in developing countries.Sewage and industrial effluents arereleased into waterways with mini-mal or no treatment, threateninghuman health and aquatic life.Thus, in the absence of policyreform, stronger institutions, andenlightened political leadership, eco-nomic and population growth indeveloping countries in the nearterm may lead to a deterioration ofthe urban environment, both physi-cal and social.

Three issues emerge as parti-cularly critical: water supply,sanitation, and water resource man-agement; solid waste management;and air pollution. In each area, thereare compelling economic, social,and environmental rationales forchange. Successful efforts, however,are likely to require significantchanges in urban practices andstrategies.

Beyond the immediate prioritiesfor improving the urban environ-ment lies the need to strengthenlocal governments, to implementnew approaches to alleviatingpoverty and supporting communi-ties, and to develop more environ-mentally friendly cities. Virtually allof the policies needed to improve

the urban environment require more effective urban governance.That will require not only strengthened governments but alsothe involvement of many other actors in the urbanenvironment—including the poor and the private sector.Community-based approaches are essential if urban services are toreach those who need them and if there is to be broad-based sup-port for needed changes in strategies and practices. The sheer sizeof urban populations and economies means that cities must leadthe way toward more environmentally sustainable practices for theworld as a whole. IF&D

This article is drawn from World Resources 1996-97, a joint publicationof the World Resources Institute, the United Nations EnvironmentProgram, the United Nations Development Program, and the World Bank,Oxford University Press, Oxford, 1996.


Source: World Resources Institute.

M ORE PEOPLE will live incities by the year 2025than occupied the wholeplanet ten years ago.

A new report, World Resources1996-97, from the World ResourcesInstitute, the United NationsEnvironment Program, the UnitedNations Development Program, andthe World Bank finds that urbaniza-tion is reshaping the physical andsocial environment as it fuels eco-nomic growth and spurs environ-mental degradation.

Urban environmental conditionsare important to the health andquality of life of a city's inhabitantsand can impose significant costs oneconomic and social development.The impact of urban areas on thesurrounding environment is an issueof growing concern. More than halfof humankind will live in urbanareas by the end of the century, and60 percent by 2020 (Chart 1). In mostnations, cities generate a majority ofthe economic activity, ultimatelyconsume most of the naturalresources, and produce most of thepollution and waste. Thus, urbanenvironmental issues, althoughoften overlooked, are important bothlocally and on national and globalscales. Neglect of these issues couldcompromise larger economic, social,and environmental goals in bothdeveloped and developing countries.

M

The developed world is alreadylargely urbanized. In the developing world, the rapid urbanizationnow under way will increasingly concentrate both population andeconomic growth in cities—as much as 90 percent of future popu-lation growth and a major share of future economicgrowth—intensifying the problems of the urban environment. Inrecent decades, urban areas in developed countries have mademajor progress in cleaning up local environmental problems, butthey remain significant contributors to regional and global envi-ronmental burdens. In developing countries, urban areas oftenhave huge populations living in poverty. Their problems are simi-lar to those of the rural poor—lack of access to clean water, sani-tation, and adequate housing—compounded by overcrowding andexposure to industrial wastes and urban air pollution.

Burgeoning cities are expanding into fragile ecosystems. Citiessometimes deplete nearby areas of water and firewood, rendering


BOOKS

Tony Killick

IMF Programmes inDevelopingCountries (Designand Impact)Routledge, London, 1995,

x + 212 pp., £13 (paper).

Kullick has written a responsible andwell-researched critique of Fund-supportedprograms. After surveying the literatureon the effects of Fund-supported programsand the methodological problems involved,Killick presents the results of his ownresearch. As in all such studies, the exis-tence of a program seems to have littlestatistically identifiable impact on out-comes. In fact, Killick finds a more signifi-cant positive effect on a number ofvariables than has most previous research,including that done within the Fund.

At first sight, the weakness of thisrelationship is surprising, since an agree-ment with the Fund on a program is con-sidered a major event by the partiesinvolved and by third parties, both publicand private. The reason for the weaknessof the relationship is probably that modelsdo not capture the rather complex inter-actions between the Fund and the countryduring the implementation of the program.In this process, the decision to approve anarrangement with the country is only one,albeit an important, element. Fund-supported programs are not missiles that,once launched, have a pre-ordained trajec-tory. The more interesting questions arewhether the Fund is helping a membercountry to adopt the right policies andwhether the institution is using itsundoubted influence wisely.

Among Killick's serious concerns withthe structure of Fund-supported programsis that the bulk of the cut in absorption istypically borne by investment. In defenseof Fund-supported programs, it could benoted that, if a shock is temporary andexternal financing constrained, it may bemore rational and welfare-enhancing to cutinvestment rather than consumption.There is also a tendency for private invest-ment to slump after a period of macroeco-nomic instability as investors wait to see

whether stability has been restored beforemaking irreversible investment decisions.Thus the fall in investment directly associ-ated with Fund-supported programs maynot reflect excessive tightness.

Killick's concern with the cuts in invest-ment feeds, however, into his critique thatthe Fund's integration of growth into thedesign of programs is more rhetoric thanreality. He suggests that programs shouldnormally incorporate a growth target atleast 1 percent above the rate of populationgrowth, and that sufficient financing forthis should be mobilized. In my view, thereal problem is to ensure that the basis islaid for high-quality and sustainablegrowth. The recovery of growth may oftenoccur only after the expiration of the pro-gram, and often it will be impossible toavoid a period of retrenchment as thenegative consequences of past policies areovercome.

Killick recognizes the increased incorpo-ration of safety nets in Fund-supportedprograms but argues that more can andshould be done. He notes that it is veryhard to come by data on the effects ofFund-supported programs on incomedistribution, but the urban poor as a groupare probably often adversely affected.Killick does note that both income-distribution outcomes and cuts in publicinvestment may reflect the priorities of thegovernments involved, rather than those ofthe Fund.

Killick finds no evidence for Fundfinancing as a catalyst for other financing,particularly as far as private sector flowsare concerned, and he describes this"catalytic role" as a myth. By contrast, theFund staff has found a strong link betweenaid flows and loans under the enhancedstructural adjustment facility (ESAF), andthe experience of mobilizing resourcesthrough consultative groups would alsoindicate such a link with regard to officialfinancing. Fund-supported programs arecertainly intended to lead ultimately tofiner borrowing terms and better marketaccess. Killick's failure to find this relation-ship may lie in the estimating techniquesused: simple before-and-after analysis willoften show reduced inflows. Programsusually aim for an improvement in thecurrent account and thus less reliance on

foreign borrowing than in the immediatepast. But Killick may be identifying a realproblem reflecting the lags associated withthe restoration of cover by export creditagencies.

Killick strongly criticizes the intellectualunderpinnings of performance criteria andcalls for radical change. While there ismore variation in performance criteria thanKillick gives the Fund credit for, it is truethat the core constellation of financialperformance criteria has not changedmuch over the years. A number of validcriticisms of credit ceilings and other crite-ria are presented, but he does not suggestany alternatives. From the Fund's point ofview, it is crucially important to link theauthorities' policy commitments to opera-tionally important magnitudes controlledby the central bank and the ministry offinance.

Killick's book is a balanced assessmentof Fund-supported programs and a valu-able and independent contribution to theliterature on the Fund. The Fund can onlybenefit from the scrutiny that he gives itsrecord and would be well advised to con-tinue the debate. Finally, the book containshelpful suggestions for improvements inthe Fund's operations that should be takenseriously within the institution.

Mark Allen

Yegor Gaidar and Karl Otto Pohl

RussianReform/InternationalMoneyMIT Press, Cambridge,Mass.,1995, 146 pp., $25.00(cloth).

.his volume reproduces the 1993 LionelRobbins Memorial Lectures given by twofigures who played crucial roles in shapingthe major economic events of recent times.Yegor Gaidar launched economic reformsin Russia in 1992, while Karl Otto Pohlwas at the center of German efforts tocontrol inflation in Europe in the 1970s



and 1980s and to lay the framework for enhanced monetarycooperation among industrial countries.

Gaidar's essay provides a fascinating and candid insight intothe unprecedented policy challenges facing a small group ofreformers as the Soviet Union collapsed in late 1991. He arguesvery persuasively that given the previous failed attempts atreform, traditional theoretical arguments about shock treatmentversus gradualism were largely irrelevant. The only option was toliberalize prices immediately and introduce market mechanismsas quickly as possible. While most observers—or, at any rate,those outside Russia—supported this approach, there was muchdebate, both at the time and afterward, about whether an addi-tional element, tight monetary and fiscal policies, was essentialfor it to work.

Gaidar makes a strong case that, in addition to liberalization,steps (drastically slashing military spending and subsidies) weretaken to stop the very real threat of hyperinflation—to him, thisachieved the "maximum" program of the reformers. To be sure,keeping monthly inflation to no more than 10 percent in the firsthalf of 1992 was considered a failure by some, but it should not beforgotten that, at the time, many observers had predicted muchworse outcomes. Actually, inflation later accelerated somewhat, asGaidar's political influence began to wane. But by then thereformers had begun to push radical privatization, which Gaidarrightly regarded as an essential means of creating a class whosemembers would not tolerate hyperinflation.

Gaidar's personal account is essential reading for anyone wish-ing to understand the political and economic pressures of thosetumultuous times. Notwithstanding the current uncertaintiesabout the future course of Russian reform, his legacy will not beeasily overturned.

While the monetary policy crises faced by industrial countrieswere much less dramatic than those in Russia, Pohl's insider'sview of the German role nevertheless makes for absorbing read-ing. In discussing the events leading up to the collapse of theBretton Woods system, he leaves no doubt that the Germanauthorities were more strongly committed to maintaining domes-tic monetary stability than to maintaining fixed exchange rates.Pohl stresses again and again the (by now fairly obvious) pointthat "exchange rate stabilization cannot work if underlying eco-nomic fundamentals do not converge."

This philosophy also pervades the second part of Pohl's essay,which deals with Europe's experience on the road to economic andmonetary union. For a long time, Pohl and the Bundesbank wereskeptical about the possibility of a convergence of low inflationpolicies within Europe. But gradually this skepticism was over-come, and by the time the Maastricht Treaty was signed, Pohlappeared willing to give other countries the benefit of the doubt.He felt that, first, most politicians (and their electorates) had cometo accept the basic arguments for price stability and againstdebasement of the currency, and, second, the increased integrationof world financial markets had already begun to impose severeconstraints on countries' ability to pursue truly independentmonetary policies. But Pohl remained cautious, stressing repeat-edly that the institutional structure of a European monetary unionmust protect monetary policy against political pressures. Anappendix provides an interesting critique of the alternativeapproaches that seemed possible at the time.

Pohl's account is a well-argued statement of the classic Germanapproach to monetary policy and of the importance he attaches tothe independent role of the Bundesbank. However, one omission issomewhat striking: there is virtually no discussion of the

upheavals in the European Monetary System resulting from theeffects of German unification—an episode suggesting that mat-ters may not be quite as clear cut as Pohl would wish.

Donal Donovan

Making Peace Work:The Role of the International

Development CommunityNicole Ball w i t h Tammy I lalevy

Policy Essay »18 ISBN: 1-56517-022-2 $13.95

Drawing on recent peacebuildingexperiences of Cambodia, ElSalvador, Mozambique, andNicaragua, this essay extractslessons about the scope anddelivery of assistance followingconflicts that are applicable toother post-conflict situationsaround the world.

Investing in Women: Progress andProspects for the World Bank

Mayra R u v i n i c , Catherine Cnvin ,and Lisa M. Rates

Policy Essay »19 ISBN: 1-56517-018-0 $13.95

Investments in women matter not only to theirown well-being, but also to that of their families,communities, and countries. This Policy Essay,a joint publication of ODC andthe International Center forResearch on Women, examinesspecific World Bank initiativesand projects aimed to benefitwomen, draws lessons fromsuccessful efforts, and consid-ers the potential of the Bankto do even more.

Political Essay SeriesPublished by the Overseas Development Council

Distributed by The Jo/ms Hop/tins University Press

To Order Call:1-800-537-5487 from within the U.S. and Canada1-410-516-6957 from outside the U.S. and Canada

Or Fax:1-410-516-6998

Credits: Photo composition for cover: Padraic Hughes and Luisa Watson;photograph of oil well on cover: © Uniphoto South / Photopic / PNI; chartart on pages 3 and 4: Luisa Watson; art on pages 30 and 31: Una Zenni;art on page 36: Massoud Etemadi; photograph on page 47: RobertRathe.



Adrian Wood

North-South Trade,Employment andInequality

Changing Fortunes in aSkill-Driven WorldOxford University Press,

New York, 1994, xii + 505 pp., $24.95 (paper).

Lhe rise in wage inequality betweenskilled and unskilled workers in industrialnations in the 1980s has brought forth awide literature whose authors are seekingexplanations for this phenomenon. Thegenerally accepted view is that a relativedecline in the demand for unskilled work-ers overwhelmed the long-term decline inthe relative supply of them. The principaldispute in the literature is about the causesof this relative decline. Two explanationsare usually offered: one involves the greaterpenetration by developing countries of theglobal market for manufactured goods andthe other involves skill-biased technicalchange.

In his book, North-South Trade,Employment and Inequality, Adrian Woodargues strongly in favor of the first expla-nation. Using factor-content analysis, heestimates that changes in trade with devel-oping countries reduced the demand forunskilled labor relative to skilled labor inthe industrial nations by 20 percent.Moreover, he finds that the differentialbetween the wages of skilled and unskilledworkers, using various measures, liesbetween 10 and 21 percent. Therefore,assuming a relative labor demand elastic-

ity equal to 1, he concludes that the growthof trade with developing countries is suffi-cient to explain all of the increase in thewage differential.

In the final section of the book, Woodoffers various policy options to address therise in wage inequality. He argues that anappropriate policy response would be tocombine improvements in education andtraining for new labor force entrants, witha system of income supplements for low-wage workers in the United States and witha progressive payroll tax in Europe. Thedifferent prescriptions for the United Statesand Europe reflect the differing degree ofwage flexibility in both regions. The flexi-bility of wages in the United States has ledto much greater income inequality than inEurope, where policies have buttressed thebottom of the wage distribution, at the costof higher unemployment. Therefore, thepolicy prescriptions attempt to reduceincome inequality in the United States andto stimulate the employment of unskilledworkers in Europe.

Wood's explanation of the rise in wageinequality relies on the validity of hisestimate of the factor content of trade. Toderive his estimate, he makes variousadjustments to the standard way of calcu-lating the cumulative reduction in thedemand for unskilled labor. First, he useslabor-input coefficients for developingcountries rather than those for developedcountries (the latter coefficients are used inthe standard methodology) because devel-oping countries export different (and non-competing) goods than developedcountries. Second, he corrects for defensiveunskilled-labor-saving innovation andspillover effects on the nontraded sector.

Both these modifications raise the standardcalculation of the factor content of trade bya multiple of 10. While the corrections thatare made by Wood go in the right direction,their magnitude is subject to debate. Itseems more reasonable to attribute part ofthe rise in wage inequality to greater expo-sure of developed country markets tomanufactured products from developingcountries and the remainder to the compet-ing explanation of skill-biased technicalchange.

Wood's general policy prescriptions arenot dependent on whether the fallingdemand for unskilled labor is caused bytrade or by new technology, and they cantherefore be assessed independently of theposition taken on the relative importanceof both explanations. The argument thatproviding education and training to newlabor force entrants will raise the supply ofskilled workers over time and hence reducethe skilled wage differential is debatable.The recent US experience of training pro-grams targeted at youths has been unfa-vorable, and a number of commentatorshave suggested reassessing educationpolicy earlier in the life cycle of labor forceentrants.

In general, the presentation of the factsof wage inequality and rising penetrationof the global market for manufacturedgoods by developing countries is clear andcogent. Moreover, the exposition of thefactor content of trade is comprehensiveand original. For those interested in under-standing the potential causes of, and reme-dies for, the rise in wage inequality in the1980s, this book should certainly be highon their list of publications to read.

Alun Thomas

^he authors of this volume explore thefollowing interlinked and fundamentalissues: What has been the lot of LatinAmerica's poor during the "lost decade" ofthe 1980s? What has been the impact ofstructural adjustment programs on theirwell-being? And, what kinds of programsand results have governments imple-mented to mitigate the impact of structuraladjustment?

The first four chapters explore variousthematic issues—they describe whatactually happened to poverty in a group ofLatin American countries and explore theunderlying determinants of poverty.Although there can be little doubt that the

Nora Lustig (editor)

Coping withAusterityPoverty andInequality in LatinAmericaBrookings Institution,Washington, 1995,

xviii + 460 pp., $39.95 (cloth), $18.95 (paper).

"head-count ratio" (the ratio of personswhose incomes fall below the poverty line)rose in most Latin American countriesduring the 1980s, it is also clear that infor-mation on poverty in the region is still

incomplete. The authors tackle theintractable issue of the impact of structuraladjustment per se (i.e., as opposed to theimpact of the crisis environment thatmakes it necessary) on the poor. In the finalthematic chapter, they provide an excellentdiscussion of the ways in which povertyprograms, such as food distribution, can beevaluated in a reasonably systematic wayeven without the benefit of a rigorousanalytical model. This reader felt, however,that the analyses of Bolivia's experiencewith the Social Emergency Fund andJamaica's with food aid programs couldhave been a bit more pointed.

These chapters are followed by casestudies on the relationship between the


TAh


Chadwyck-Healey Business Publications on CD-ROM

Business research tools on CD-ROMprovide speed, accuracy and

unparalleled search capabilities• The EconomistKeep up with internationalbusiness and politics, finance,science and technology andthe arts across hundreds ofissues of The Economist.

m FT McCarthyFind company, industry andmarketing informationdevelopments from over fiftyinternational newspapers

and business magazinescompiled into onedatabase.

development, and monetarypolicy, with this databasecreated by the staff of theJoint Bank-Fund Library inWashington, DC.

• Leadership Directorieson CD-ROMAccess over 350,000 topdecision makers in business,government, and professionalorganizations using theelectronic version of therenowned "Yellow Books."

• Financial TimesLearn about changes inthe financial, economicand political arena.Financial Times is a leadingsource of accurate andunbiased news for theanalysis of business eventsand trends.


• IntlEc —The Indexto InternationalEconomics,Development andFinanceFocus on globaleconomic trends,international trade and

^j^ CHADWYCK-HEALEY

Chadwyck-Healey, Inc. Chadwyck-Healey, Ltd.1101 King Street, Alexandria, VA 22314 The Quorum, Barnwell Road(703) 683-4890 Cambridge, U.K. CB5 8SWToll free: (800)752-0515 Phone: 01223 215512Fax:(703)683-7589 Fax: 01223 215514E-mail: [email protected] E-mail: [email protected]://www.chadwyck.com http://www.chadwyck.co.uk

For more information contact your Chadwyck-Healey representative at:


http://www.chadwyck.co.uk

http://www.chadwyck.co.uk

economic crisis of the 1980s and the socialsafety net programs in Argentina, Brazil,Chile, Mexico, Peru, and Venezuela. Thesecountry case studies deal, with varyingdegrees of thoroughness,with issues suchas the origins of the country's macroeco-nomic problems, the impact of the crisis onthe prevalence and depth of poverty, andgovernments' efforts to safeguard existingsocial programs or introduce new ones toalleviate poverty. The most comprehensivetreatments are those of Chile andVenezuela; the chapter on Peru has seen itsmajor thesis overtaken by events. TheVenezuela study is notable for its concisediscussion of the macroeconomic back-ground to the late 1980s crisis and for itsanalysis of the opposition to fundamentalreform in various areas by diverse vestedinterests. In the case of social expenditure,the authors argue, it is the public sectorunions that resist the reorientation of, forexample, health and education expenditurefrom the tertiary to the primary sectors,because of its impact on jobs. The chapteralso clarifies how the targeted nature ofprograms introduced in early 1989, after

the riots, distinguishes them from thosealready in place.

The Chilean study provides an exhaus-tive account of the country's social pro-grams. Chile sought to contain educationand health expenditures mainly by squeez-ing teachers' and health workers' salaries,and by postponing capital expenditure.The government did, however, try toincrease the share of primary expenditureand of targeted special programs for moth-ers and infants. The authors argue that thisresulted in undue neglect of others in need,and that, given Chile's demographic andepidemiological profile—"older" than thatof other Latin American countries, andwith fewer diseases of malnutrition andless bad sanitation—it was misguided.This seems debatable given the importanceof early care and attention for future healthand productivity. On the discussion of thedecline in total public expenditure, theauthors could have explained more care-fully the role of devolution of such expendi-ture to the municipalities—it is not clearwhether total expenditure includes expen-diture at this level or not. Chile's experi-

ence contrasts markedly with those ofArgentina and Venezuela, whose politicallydifferent governments were unsuccessful(at least in the 1980s) in achieving funda-mental change in the structure of publicexpenditure.

The vexing issue of the counterfactualcould be more clearly handled by the coun-try studies. Thus, one feels that structuraladjustment is "blamed" for the increase inpoverty when, in reality, one should askwhat would have happened in the absenceof adjustment. The Chilean study arguesthat the impact of the boom-and-bust cycleis asymmetric: the increase in poverty thatoccurs during the recession is not elimi-nated fully during the recovery. This is animportant argument, but one that needs tobe developed.

These criticisms notwithstanding, thisvolume is a worthwhile read—not just forregional specialists, but for anyone inter-ested in the issues it raises. Would thatthere were similar studies for other parts ofthe world.

George A. Mackenzie

Ricardo Lopez Murphy (editor)

FiscalDecentralization inLatin AmericaInter-American DevelopmentBank, Washington, 1995,297 pp., $18.50 (paper).

Jie concept of the nation-state is notdead, but it has come under increasingpressure from two sides. First, the signifi-cant advances in regional and global inte-gration and coordination over the last fewdecades—like the European Union orMercosur—have demonstrated conclu-sively that regional arrangements can bean effective mechanism for carrying outpolicies traditionally in the realm of thenation-state. Second, increased decentral-ization of national decision making tosubnational levels of government, a pro-cess that has paralleled the advance ofdemocracy around the world, has revealeda significant potential for efficiency gains.Further decentralization seems inevitable,but major questions remain as to whatgovernment functions should be decentral-ized and in which sectors and regions; to

what extent and how decentralizationshould be implemented; and how macro-economic stabilization objectives can besafeguarded in this process.

This book examines fiscal aspects ofdecentralization in Latin America, a timelytopic that has gained much prominence inthe context of the sweeping democratiza-tion processes in the region. It containsfour case studies—Argentina, Chile,Colombia, and Peru—and an overviewchapter that synthesizes the lessonslearned from the case studies. All of thestudies are based on the hypothesis that ahigh degree of centralization may bewasteful, as it prevents local or regionalfeatures and advantages from being takeninto account.

Decentralization, when badly done,however, may increase regional disparities,jeopardize macroeconomic stability, andundermine economic efficiency. Theauthors recognize this by emphasizing thatfiscal decentralization without adequateinstitutions (for example, a strong taxadministration and expenditure manage-ment system) at all levels of government,clear rules governing revenue and expendi-ture assignments and responsibilities, anda well-designed system of transfersbetween different levels of governmentactually reduces fiscal accountability and

creates macroeconomic and allocationproblems that may lead to severe economiccrises.

A major strength of this book is that itprovides concrete, detailed policy advice.Indeed, it could even serve as a blueprintfor reforming intergovernmental fiscalrelations in these four countries, given itsdazzling range of policy recommendations.These recommendations cover enhancingthe efficiency of institutional, administra-tive, and regulatory structures; restructur-ing revenues, expenditures, andintergovernmental transfers; addressingincome distribution problems; and safe-guarding macroeconomic stabilizationobjectives within a decentralized fiscalsystem. The book convincingly shows howa well-designed decentralized fiscal systemwould improve the effectiveness of govern-ment expenditure in each of the four coun-tries and, most important, how this couldbe achieved without compromising hard-won macroeconomic stability.

Of course, reform needs differ fromcountry to country. Take, for example,macroeconomic stabilization. Chile is prob-ably the most centralized of the foureconomies, and it already has a strongarsenal of stabilization tools, ranging fromfunds that keep fluctuations in copperrevenues from destabilizing the budget, to



a constitutional framework that preventsthe legislature from making upwardadjustments in the calculation of availableresources, to strict controls over the powerof subnational levels of government toincur debt. In contrast, Argentina, the mostdecentralized of the four countries, has anelaborate revenue-sharing system with astrong procyclical bias and traditionallyhas had insufficient controls on provincialfinances. The book argues that theseremain weak points in the otherwise suc-cessful stabilization program now under

way. The same features, however—a rela-tively powerful central government inChile, and fairly powerful but inefficientprovincial administrations inArgentina—make for important differ-ences in the reforms needed to obtain theefficiency gains that a decentralized fiscalsystem potentially promises.

Unfortunately, the book's coverage ofcountries is not as broad as its title sug-gests, and, in particular, one wishes thediverse experience of large federal stateslike Brazil and Mexico had been included.

Also, the book could have benefited frommore meticulous editing to highlight themain policy lessons. While it requires arather patient reader who does not mindgetting immersed in overwhelming detail,its careful analyses and detailed policyrecommendations make it an invaluablesource for policymakers and researchersinterested in intergovernmental fiscalrelations.

Gerd Schwartz

Dennis T. Yasutomo

The NewMultilateralism inJapan's ForeignPolicySt. Martin's Press, NewYork, 1995, x + 130pp.,$39.95 (cloth).

•Japan's emergence as the world's secondlargest economy has been one of the mostsignificant events of the postwar period.While the bulk of its economic develop-ment was financed by domestic savings, itshould not be forgotten that in 1960, Japanshared with India the status of top recipi-ent of World Bank loans. It received its lastloan from the Bank in 1966 and did notfinish repaying all of its loans from theBank until 1990.

While Japan's economic developmenthas been extraordinary, its internationalrole in influencing foreign affairs and inproviding global leadership has been, untilrecently, quite muted. Postwar regionalanimosity toward Japan, as well as consti-tutional limits on its military capabilities,resulted in a risk-avoiding diplomacy thatseparated economic development frominternational politics. This was reinforcedby Japan's reliance on military protectionfrom the United States, which still causesfriction today. As its economic strength hasincreased in both absolute and relativeterms, however, and as global militarytensions have abated, Japan has beenbelatedly but increasingly speaking with alouder international voice—one that is nolonger automatically aligned with that ofthe United States.

In this well-organized and clearly arguedbook, Professor Yasutomo examinesJapan's evolving foreign policy as

expressed through its membership inmultilateral development banks. He arguesthat, given constitutional and geopoliticalconstraints, it is official development assis-tance (ODA) and not military activities thathas been Japan's contribution to the worldcommunity. Tracing Japan's contributionsfrom its membership in the Colombo Planin 1954, followed by its first bilateral loan(to India) in 1958, Yasutomo characterizesJapan's aid policy in the 1950s and 1960sas largely commercially based, with anexplicit separation of politics from eco-nomics, and almost exclusively focused onits Asian neighbors.

Yasutomo states that in the 1970s, fol-lowing the normalization of US-Chineserelations and the end of the Vietnam War,Japanese aid policy became increasinglydiversified and politicized. ODA terms andconditions were softened, amounts wereincreased, and new aid forms and newrecipients, including Mongolia andVietnam, emerged. Thus, while economicinterests remained paramount, the politicaldimension was increasingly apparent(though yet to be acknowledged). Thesechanges in policy were also visible inJapan's introduction of aid to the MiddleEast and Latin America, which was part ofan attempt to secure and diversify itsenergy sources.

In the 1980s, the emergence of Japan'ssizable and persistent external surplusessimultaneously with the global debt crisisresulted in renewed appeals to Japan tofurther increase ODA. Japan respondedwith three debt-relief plans totaling$65 billion over 1986-89, with increasinglyheavy emphasis on intermediation throughmultilateral institutions. Yasutomo arguesthat four major developments have charac-terized Japanese ODA policy in the1990s—the political and strategic use ofODA has become explicit; the country's

Asian ODA policy has changed, as somerecipients have graduated and others(especially from the former Soviet Union)have emerged; a new ODA philosophy,reflected in Japan's recently adopted ODAcharter, has arisen; and multilateral aid hasexpanded into new areas and received newemphasis.

Yasutomo discusses Japan's evolvingforeign aid policy through detailed exami-nations of its activities in the World Bank,the Asian Development Bank, and theEuropean Bank for Reconstruction andDevelopment. In addition to examiningJapan's increased financial contributions tothe multilateral development banks,Yasutomo discusses Japan's heighteneddesire to increase the share of Japanesepersonnel in these institutions. He alsohighlights Japan's ODA policies in the1990s with a chapter examining the multi-laterialization of Japan's dealings with theformer Soviet Union. In addition, he arguesthat Japan has become increasingly vocal inits view—which is shared by other Asiangovernments and academics—that theprevailing development strategy is exces-sively narrow and ignores the government'srole in recent "miracles" in East Asianeconomic growth. This perception resultedin the Japanese-financed study, The EastAsian Miracle: Economic Growth andPublic Policy, by the World Bank in 1993.

While the debate about whether or notthere have been miracles (with strongdoubts cast by Paul Krugman and AlwynYoung) is far from over—and, if so, whattheir causes have been—there is no doubtthat Japan will continue to influence themultilateral development banks in waysthat are more in line with its economicstrength.

Mark S. Lutz



JOURNAL OF AFRICAN ECONOMIES

Managing Editors: PAUL COLLIER, University of Oxford;JAN WILLEM GUNNING, Free University, Amsterdam;BENNO NDULU, African Economic Research Consortium;T. A. OYEJIDE, University of Ibadan, Nigeria

"This new journal, with its high-level international board ofeditors, promises to be just what is needed: an outlet and astimulant for a growing volume of applied economicresearch on Africa."—Prof. Stanley Fischer, MIT

"A journal that none of us with astake in the future of Africa canafford to neglect."—Prof. DavidStout, Chief Economist, Unilever

Volume 5, 1996 (3 issues)ISSN 0963-8024Individuals: $73/£38Institutions: $118/^66Special rates for subscribers in Africa:Individuals: $40 Institutions: $58

THE CHINA QUARTERLY

Editor: DAVID SHAMBAUGH, SOAS, London

"The leading interdisciplinary journal on modernChina."—Nicholas Lardy, The Brookings Institution

Issues 145-148, 1996 ISSN 0009-4439Individuals: $61/£33 Students*: $29/^16Institutions: $80A£42

INDUSTRIAL AND CORPORATE CHANGE

Editors: NATHAN ROSENBERG and DAVID TEECE,University of California; NICK VON TUNZELMANN,University of Sussex; GIOVANNI DOSI and FRANCOMALERBA, Milan

"Finally, a truly multidisciplinary journal focused on someof the most exciting and important areas of social science:organizational theory, economic history, strategic manage-ment, and institutional theory. Years from now ICC will berecognized for spurring on and integrating innovative,path-breaking -work."—Prof. Glenn R. Carroll, University ofCalifornia, Berkeley

THE ECONOMICS OF TRANSITION

Editors: PHILIPPE AGHION,EBRD and Nuffield College, Oxford,and PAUL HARE, Center forEconomic Reform and Transformation,Heriot-Watt University, Edinburgh

"The leading academic forum forthe analysis of post-Communistreform."—Richard Layard, Centre

for Economic Performance, LondonSchool of Economics

Volume 4, 1996 (2 issues) ISSN 0967-0750Individuals: $82/^49 Academic Libraries: $175A£99Institutions: $280A£160

THE REVIEW OF FINANCIAL STUDIES

Editor: FRANKLIN ALLEN, Wharton School, University ofPennsylvania

The Review of Financial Studies is internationally recognized as amajor forum for the promotion andwide dissemination of significantnew research in financial econom-ics. As reflected by its broadly basededitorial board, the Review balancestheoretical and empirical contribu-tions. The journal is sponsored bythe Society for Financial Studies.

Volume 9, 1996 (4 issues) ISSN 0893-9454Individuals: $72/^50 Students*: $29/£19Institutions: $195/^135

OXFORD REVIEW OF ECONOMIC POLICY

Editor: CHRISTOPHER ALLSOPP, New College, Oxford

"The Oxford Review of Economic Policy has quickly estab-lished itself in the business of conveying the latest academiceconomic thinking to politicians, bureaucrats, and the restof us."—The Times

Volume 12, 1996 (4 issues) ISSN 0266-903XIndividuals: $88A£48 Institutions: $195/£108

OXFORDJOURNALS

Volume 5, 1996 (4 issues) ISSN 0960-6491Individuals: $78A£43 Students*: $50/£3()Institutions: $195/£110

OXFORD UNIVERSITY PRESSTo place a subscription order or request a free sample issue,

call 1-800-852-7323 (USA & Canada).Or call 919-677-0977. Fax 919-677-1714. In Europe, ring +44 (0)1865 267907 or fax +44 (0)1865 267485

*Please provide proof of status. £ rates apply in Europe; US$ rates apply in the USA and elsewhere.E-mail: [email protected] World Wide Web http://www.oup.co.uk/jnls/

ALSO PUBLISHED BY OUP: OXFORD ECONOMIC PAPERS; JOURNAL OF LAW. ECONOMICS & ORGANIZATION; JOURNAL OF ENVIRONMENTAL LAW

STR.vuxiv WITH OXFORD JOURNALS


http://www.oup.co.uk/jnls/

pDWERli BJH!

Amoco Power brings total energy solutions to a power generation project.^SSt~Our century-long heritage of meeting global energy needs is echoed from ^^H

the wellhead, to the pipeline to the power plants that will light tomorrow's

cities and industrial hubs. We have the expertise and financial strength to develop,

supply and manage even the most challenging of projects. Above all, every project

we undertake is rooted in a commitment to the people we serve. So when you look

PMn^n^^f^niAlf^fP for energy, look to the company that

Power is only part of the equation. has no equal. Look to Amoco Power.


letter from the editor - imf elibrary · 2019-12-11 · to attract private investment to its oil...

Documents