initiative for policy dialoguepolicydialogue.org/files/publications/capital_mrkt_lib_s... €...

http://ann.sagepub.com/

ScienceAcademy of Political and Social The ANNALS of the American

http://ann.sagepub.com/content/579/1/219The online version of this article can be found at:

DOI: 10.1177/000271620257900114

2002 579: 219The ANNALS of the American Academy of Political and Social ScienceJoseph E. Stiglitz

RewardCapital Market Liberalization and Exchange Rate Regimes: Risk without

Published by:

http://www.sagepublications.com

On behalf of:

American Academy of Political and Social Science

can be found at:ScienceThe ANNALS of the American Academy of Political and SocialAdditional services and information for

http://ann.sagepub.com/cgi/alertsEmail Alerts:

http://ann.sagepub.com/subscriptionsSubscriptions:

http://www.sagepub.com/journalsReprints.navReprints:

http://www.sagepub.com/journalsPermissions.navPermissions:

http://ann.sagepub.com/content/579/1/219.refs.htmlCitations:

at COLUMBIA UNIV LIBRARY on July 8, 2010ann.sagepub.comDownloaded from


http://ann.sagepub.com/content/579/1/219

http://www.sagepublications.com

http://www.aapss.org/

http://ann.sagepub.com/cgi/alerts

http://ann.sagepub.com/subscriptions

http://www.sagepub.com/journalsReprints.nav

http://www.sagepub.com/journalsPermissions.nav

http://ann.sagepub.com/content/579/1/219.refs.html


219

Capital Market Liberalizationand Exchange Rate Regimes:

Risk without Reward

By JOSEPH E. STIGLITZ

Joseph E. Stiglitz holds joint professorships at Columbia University’s EconomicsDepartment, School of International and Public Affairs, and Business School. From1997 to 2000 he was the World Bank’s Senior Vice President for DevelopmentEconomics and Chief Economist. From 1995 to 1997, he served as Chairman of the U.S.Council of EconomicAdvisers and a member of President Bill Clinton’s cabinet. He waspreviously a professor of economics at Stanford, Princeton, Yale, and All Souls College,Oxford. He was awarded the Nobel Prize in Economics in 2001 for his analysis ofmarkets with asymmetric information.

ABSTRACT: This paper examines the consequences of capital mar-ket liberalization, with special reference to its effects under differentexchange rate regimes. Capital market liberalization has not lead tofaster growth in developing countries, but has led to greater risks. Itdescribes how International Monetary Fund policies have exacer-bated the risks, as a result of the macro-economic response to crises,with bail-out packages that have intensified moral hazard problems.The paper provides a critique of the arguments for capital market lib-eralization. It argues that capital flows give rise to large externali-ties, which affect others than the borrower and lender, and wheneverthere are large externalities, there is potential scope for governmentinterventions, some of which are welfare increasing.



220

F OR almost half a decade, capital-L market liberalization raged asthe prime battleground betweenthose who were pushing for andagainst globalization, and for goodreason: By the mid-1990s, the notionthat free trade or at least freer trade

brought benefits both to the devel-oped and the less developed coun-tries seemed well accepted. Presi-dent Clinton could claim passage ofNAFTA and the Uruguay Round,with the establishment of the WorldTrade Organization, among the ma-jor achievements of his first fouryears. APEC and the Americas hadboth committed themselves to creat-

ing a free-trade area. Not only hadthe intellectual battle been won-

only special interests resisted tradeliberalization-but so seemingly hadthe political battle. On other fronts,the broader liberalization/free-market agenda was winning victoryafter victory: the Uruguay round hadextended the scope of traditionaltrade liberalization to include liber-alization in financial services, theprotection of intellectual propertyrights, and even investment. Al-though the Multilateral InvestmentAgreement was having trouble, in-vestment protections in NAFTAwere cited as a basis on which further

agreements could be reached. Even&dquo;liberal&dquo; governments-the demo-cratic administration in the United

States, the labor government in Brit-ain-embraced privatization and de-regulation, with the United Statesgoing so far as to push through theprivatization of the corporation mak-ing enriched uranium, the core ingre-dient in making nuclear weapons (as

well as fuel for nuclear reactors).Only capital market liberalization-eliminating the restrictions on thefree flow of short-term capital-re-mained as a point of contention. Atits annual meetings in Hong Kong,the International Monetary Fund(IMF) sought to settle this issue too:it asked for a change in its charter, togive it a mandate to push for capitalmarket liberalization, just as it had amandate, in its founding, for theelimination of capital controls thatinterfered with trade.

The timing could not have beenworse: the East Asia crisis was brew-

ing. Thailand had already suc-cumbed, with a crisis that began on2 July. The delegates to the HongKong meeting had hardly unpackedtheir bags on returning home whenthe crisis struck in Indonesia. Withina little more than a year, it had

become a global economic crisis,touching virtually every corner of theglobe, with bailouts billed at morethan 150 billion dollars occurring notonly in Thailand and Indonesia, butKorea, Brazil, and Russia. And it wasclear that hot, speculative money-short-term capital flows-was at theheart of the crisis: if they had notcaused it, they at least played a cen-tral role in its propagation. The onlytwo large emerging markets to bespared the ravages of the globalfinancial crisis were India and

China, both of which had imposedcapital controls. (Even as the globaleconomy faced a major slowdown,China managed to grow by morethan 7 percent, India by more than 5percent). Malaysia had imposed capi-tal controls to help it manage its way



221

through the crisis, and as a result, itsdownturn was shorter, and as it

recovered, it was left with less of alegacy of debt than the other coun-tries because it had imposed capitalcontrols. By the time matters settleddown, the IMF had markedlychanged its tune: its chief economist(Mussa 2000) admitted that finan-cial market liberalization could have

markedly adverse effects on lessdeveloped countries that were notadequately prepared for it (in theview of many economists in the

developing world, this meant virtu-ally all developing countries).’ It

admitted that its predictions (andthose of the U.S. Treasury) thatMalaysia’s imposition of controlswould prove to be a disaster had been

wrong-they had succeeded in spiteof what might have seemed as effortsto undermine the country throughpublic criticism of an almost unprec-edented nature.2 2

But while the intellectual battlewas thus seemingly over, the politicalbattle continued: the managingdirector of the IMF, Michel Cam-dessus, continued pushing for capitalmarket liberalization in his annual

speeches until his departure from theIMF. And countries that proposegoing back on capital market liberal-ization are strongly advised againstit-to the point of implicit or explicitthreats of having programs cut off.3

But these political battles are, forthe most part, going on behind thescenes. The more visible politicaldebate has moved back to issues

thought at one time settled-forinstance, to trade and intellectualproperty rights. Still, revisiting thatearlier debate has much to teach us,

both about economics and politics. Itis precisely because the disjunctionbetween the positions that the IMFtook and the theory and evidenceconcerning capital market liberaliza-tion, between their mandate to pro-mote global stability, and the policywhich seem so patently to lead toglobal instability was so great thatthe debate on capital market liberal-ization throws into such stark reliefbroader aspects of the globalizationcontroversy. (In other arenas, such astrade liberalization, theory and evi-dence are more ambiguous, andwhile the IMF may have pushed poli-cies that could not be defended asfulfilling its mandate, neither couldthey, by and large, be criticizedfor actually going against theirmandate.)

The consequences of capital mar-ket liberalization depend, of course,in part, but only in part, on theexchange rate regime, which is thefocus of many of the articles in thisissue. But before turning to thatissue, it is important to understandthe more general case against and forcapital market liberalization. Ac-cordingly, in this article, I proposefirst to explain the strength of theopposition to capital market liberal-ization : it increases the risks facing acountry while it does not promoteeconomic growth. Given the over-whelming theory and evidenceagainst capital market liberaliza-tion, one wonders: how could themajor international organizationresponsible for promoting growthand stability have promoted a policythat seemed so contrary to its objec-tives ? I first review the argumentsthat were put forward for capital



222

market liberalization. I then turn tothe deeper political economy ques-tions, exploring the role of ideologyand interests. I conclude by arguingthat at the root of the problem is gov-ernance : the governance structure ofthe IMF led it to push for policies thatwere contradictory to its mandate forpromoting global stability and thatreflected the interests and ideology ofthose to whom it was directlyaccountable. There is, in this, animportant lesson for the evolvingglobalization debate, to which I turnbriefly in the concluding section.

THE ECONOMIC CASEAGAINST CAPITAL

MARKET LIBERALIZAITON

The evidence is that capital mar-ket liberalization is not associatedwith faster economic growth orhigher levels of investment but isassociated with higher levels of eco-nomic volatility and risk. And, ingeneral, the poor bear the brunt ofmuch of this risk, especially in devel-oping countries, where safety nets(like unemployment insurance sys-tems) are nonexistent or inadequate.

Growth

Ascertaining whether trade liber-alization, or capital market liberal-ization, leads to faster economicgrowth is not an easy matter. A stan-dard, though widely discredited,methodology entails looking at thegrowth rates of different countries,attempting to ascertain whetherthose who have liberalized more orfaster have grown faster, controllingfor other factors that might haveaffected growth. The problematic

nature of such studies is highlightedby the contradictory results thathave been obtained in the trade liber-alization literature, with scholarslike Sachs and Warner (1995) argu-ing that trade liberalization is sys-tematically associated with growth,and others, like Rodriguez andRodrik (1999), questioning theresults. The results are highly sensi-tive to issues like how to weight theexperience of different countries andhow to separate causal factors withmere association. For instance,China with its more than 1 billion

people has been the fastest growingdeveloping country in the world;growth in China accounts for a sub-stantial fraction (by some accounts,two-thirds or more) of total growthamong the low-income countries. Butshould China-which did not liberal-ize-be given the same weight in theanalysis as some small country inAfrica with a couple million people?If some of these small countries grewslightly faster, some grew slightlyslower, and on average, those who lib-eralized grew slightly faster, are weto infer that liberalization is an

important ingredient in growth-when the world’s major success story,with growth a multiple of that of anyof the African countries, did not liber-alize ? The matter can be put anotherway: if one treated the separate prov-inces of China as separate datapoints-and they are each large,many times the size of the averageAfrican states, and each followed

slightly different policies-then thetwenty fastest growers in the pasttwo decades are all in China. A studythat treated these provinces as unitsof analysis might conclude strongly



223

that liberalization was not good forgrowth, while a study that treatedChina as a single data point mightconclude that it was.

There are other statistical prob-lems. Assume it were the case thatcountries that did not liberalize, onaverage, were more authoritarianand that authoritarianism is bad for

growth; but in the statistical analysisof growth, no measure of authoritari-anism was included, or a measurethat did not capture the relevantdimensions of a multidimensional

political construct. Then, the statisti-cal analysis might conclude that lib-eralization was good for growth,when the correct conclusion is thatnonauthoritarian political struc-tures, appropriately defined, aregood for growth.What was most remarkable about

the drive for capital market liberal-ization from the IMF was that at thetime they pushed for this change inthe global economic architecture,there was no study even of the cross-country statistical kind that sup-ported trade liberalization, as dis-credited as those studies might be,which provided empirical evidence insupport of capital market liberaliza-tion. The one widely cited study byRodrik (1998)--using the IMF’s ownmeasures of liberalization-showedthat it did not lead to faster economic

growth. One might have thought thatthe IMF would have made a majoreffort to refute Rodrik’s study and topresent countervailing studies show-ing the contrary That they did not,and that they seemingly did not feelthe need to refute the even more com-

pelling evidence showing that capitalmarket liberalization led to greater

risks, may say a great deal about thenature of the organization, a point towhich we return later.

But there is, perhaps, a simplerreason: Rodrik’s (1998) study merelycorroborated what was obvious, bothempirically and theoretically. It wasnot only China that had grown rap-idly without capital market liberal-ization ; India, too, had experiencedrapid growth over the 1990s, and ittoo had not liberalized. Russia had

liberalized, and the liberalizationhad led not to a flow of capital intothe country but to massive capitalflight, and the country’s GDP had,partly as a result, plummeted bymore than 40 percent, a decline thatwas reflected in socioeconomic statis-

tics, such as marked shortening oflife spans, and in data, collectedthrough household surveys, showingan increase in poverty from around2 percent to between 25 percent and40 percent, depending on the mea-sure used.

But these results should not havecome as a surprise. Growth is relatedto investment, to new enterprisesand old enterprises expanding. Suchinvestments cannot be based on spec-ulative money that can come into andout of a country on a moment’s notice.On the contrary, the high volatilityassociated with such flows de-stabilizes the economy, as we shallsee in the next section, and thehigher economic volatility makesinvestment less attractive.

There is another channel throughwhich capital market liberalizationhurts growth. The flow of funds intothe country leads, under flexible ex-change rates, into a higher exchangerate, making it more difficult for a



224

country to export or compete againstimports (a version of the so-calledDutch disease problem). In somecases, such as Thailand, the fundshelped feed a speculative real estateboom, which distorted the economy.To prevent inflation, to sterilize theinflow of funds, which might other-wise have led to an excess demand for

goods, the monetary authorities hadto raise interest rates, which stifledinvestment in other sectors.4 The dis-tinction between foreign directinvestment and these short-term

flows could not be clearer. The foreigndirect investment leads directly to anincrease in GDP and in employment;it brings with it new technology,access to markets, and training-none of which accompany specula-tive portfolio flows.

There is another way of seeing theadverse effects on growth: today,countries are told to keep reservesequal to their foreign denominatedshort-term liabilities. Consider the

consequence of a company within a

poor small country borrowing moneyfrom an American bank $100 millionin dollars short term, paying say18 percent or 20 percent interest. Thecountry then is forced to put a corre-sponding amount in reserves-money that could have gone toward

high-return investments in schools,health clinics, roads, or factories. Thereserves are held in the form of U.S.

treasury bills, yielding 4 percent.In effect, the country is lending$100 million to the United States at 4percent and borrowing it back againat 20 percent-at a net cost of $16million a year to the country, a trans-action that clearly might be good forgrowth in the United States but is

unlikely to have a substantial posi-tive effect on the growth of thedeveloping country 5

Risk

The rapid movement of funds intoand out of a country is clearlydestabilizing, a point brought homeforcefully by the East Asian crisis,where the capital outflows exceededin some cases 10 percent of GDP.Flows of that magnitude (equivalentto close to $1 trillion for the UnitedStates) would be highly disruptive,even in a country with strong finan-cial markets. The effects on develop-ing countries have been devastating.

Empirical studies have shownthat there is a systematic relation-ship between capital market liberal-ization and instability (see Demirguc-Kunt and Detragiache 1997, 1999).The period immediately followingliberalization is one in which risk is

particularly marked, as marketsoften respond to the new opportuni-ties in an overly exuberant manner,as they see previously closed oppor-tunities opened up. The increasedmacro-economic risk would imply anecessity for increased monitoring offinancial institutions, as the fran-chise value, the expected present dis-counted value of future profits, is

likely to be eroded given the higherprobability of an economic downturn;and in the absence of increased moni-

toring, the higher level of risk takingitself would contribute to greaterinstability. Unfortunately, typicallygovernments have not done a betterjob of regulation, for two reasons.Often, the capital market liberaliza-tion is in response to external pres-sure (e.g., from the IMF or the United



225

States), and that same pressure hasbeen accompanied by pressure to lib-eralize financial markets, that is,remove restrictions that, in part,result in less exposure to risk (theelimination of Thailand’s restric-tions on speculative real estateinvestments are a case in point). Sec-ond, the process of liberalization hasbeen accompanied by huge increasesin demand for the relatively fewtrained personnel, many of whomworked for the regulatory authoritiesand the central bank; the public sec-tor simply cannot compete in payingsalaries against the private sector.Hence, just at the time when the needfor improved regulation is greatest,the country’s capacity is reduced (seeHellman, Murdock, and Stiglitz1996).

Capital market flows to which cap-ital market liberalization gave riseare now recognized to be the pivotalfactor in the East Asia crisis.~ Butthere have been more crises thathave been deeper and longer lastingin the past quarter century-morethan one hundred countries havebeen afflicted (see Caprio andKlingebiel 1996; Lindgren, Garcia,and Saal 1996), and it is apparentthat the trend toward increased capi-tal and financial market liberaliza-tion has been a key factor. The IMFand U.S. Treasury suggested that theEast Asian countries were vulnera-ble because of a number of structural

failings; on the contrary, these coun-tries had performed better over thepreceding three decades, not only interms of growth but also in stability:two of the affected countries had had

only one of economic downturn, twohad had none, a better performance

than any of the OECD countries. If

they were vulnerable, it was a newlyacquired vulnerability because of thecapital and financial market liberal-ization that had been foisted on thesecountries.

IMF responses exacerbated therisk. The nature of the response to acrisis affects the consequences,including who bears the burden. Inthe case of the East Asia crisis, theIMF responded with fiscal contrac-tions and monetary tightening,which deepened the economic down-turns and failed in their intendedeffects of sustaining the exchangerate. In addition, the restructuringstrategy, which involved closingfinancial institutions (in the case of

Indonesia, closing sixteen banks,with an announcement that morewere to follow, but that depositorswould not have their deposits guar-anteed, leading to a run on the bank-ing system), led to a further collapsein the supply of credit; and the morehands-off corporate restructuringstrategies meant that corporate dis-tress was addressed at an extremelyslow pace-four years after the crisis,between 25 and 40 percent of Thailoans remained nonperforming.

The failure was predicted by econ-omists within the World Bank, andthe reason was obvious: given thehigh leverage, the high interest ratesforced many firms into distress andworsened the problems of the finan-cial institutions. The combination ofthe normal Keynesian demand-sideand supply-side contractions proveddevastating. Although at this junc-ture, the IMF admits several of thefailures, on the critical issue of mone-tary policy it remains adamant. It



226

believes that higher interest rateslead to a capital inflow that supportsa country’s currency, evidently evenin circumstances such as those ofEast Asia with high levels of lever-age, in spite of the absence of evi-dence in support and some evidenceagainst and in spite of the over-whelming theoretical argumentsagainst the policy. The statisticalanalyses of whether raising interestrates leads to higher exchange ratesis even more problematic than thecross-country regressions referred toearlier. Here, the critical issue is toidentify the appropriate counter-factual, that is, what would have hap-pened but for the policy. It is clearthat the IMF interventions in East

Asia, which included not only thehigh interest rates but also massivebailouts and contractionary fiscalpolicies, did not prevent a slide in theexchange rates; indeed, looking atexchange rate movements, it is hardto detect evidence of interventions

having any positive impact. It is, ofcourse, possible that the mistakenpart of the IMF packages systemati-cally undermined the positive effectsof the interest rate policies, or thatjust at the moment of the interven-tions, the pace of decline in exchangerate would have accelerated, and thisacceleration was reversed by theinterventions. But neither of these is

plausible, and a more detailed analy-sis of interest rate increases in othercrises does not suggest that they arevery effective instruments (seeFurman and Stiglitz 1998a). The the-oretical arguments put forward bythe advocates of this policy are notcompelling: the higher interest ratesare supposed to attract funds into the

country, bolstering the exchangerate. In fact, the economic disruptionnot only does not attract funds intothe country but also leads to massivecapital flight. Lenders care not justabout the interest rate promised butalso about the probability of beingrepaid; it was concern about defaultthat led banks to refuse to roll overtheir loans. Thus, this was a variableof first-order importance-but leftout from the IMF analyses. The poli-cies increased the probability ofdefault so that the total impact wasto make it less attractive to put fundsinto the country’ 7

The important point is this: hav-ing failed to identify the reasons fortheir admitted failures in their fiscaland financial policies and havingeven refused to acknowledge the mis-takes in monetary policy means thatin the future, the mistakes are likelyto be repeated so that countries cananticipate facing major economicdownturns in the event of a crisis.

There are further aspects of IMFresponses that exacerbated thedownturn in Indonesia. Even in thebest of circumstances, major eco-nomic downturns can give rise topolitical and social turmoil. In thecase of ethnically fractionated societ-ies, such turmoil is even more likely(see, for instance, Collier andHoeffler 1998). A perception that theburden is borne unfairly by the poorincreases the likelihood of turmoileven more. I predicted in earlyDecember 1997 at a meeting offinance ministers and central bankofficials at Kuala Lumpur that if theIMF maintained its macro-economic

policies in Indonesia, there was ahigh probability of such turmoil



227

within six months. I argued that evenif the IMF did not care about social

costs, especially those borne by thepoor, it was simply bad economic pol-icy. All that the head of the IMF, whowas in attendance, could say in replywas that the country had to bear thepain. I was perhaps overly optimistic:riots, every bit as bad as my worsefears, broke out within five months.But I had not anticipated that just asthe economy was plunged intodepression, with unemploymentsoaring and real wages plummeting,food and fuel subsidies for the very

poor would be cut back. Evidently,the IMF could provide billions to bailout Western banks and lenders, butthere were not the measly millionsrequired to finance subsidies for thevery poor. The outrage was under-standable and the consequences longlasting: it will take years before thatcountry recovers to its precrisis level.

There were other aspects of IMFpolicies that exacerbated the down-turns. In East Asia, the debt waslargely private. When private partiescannot meet their obligations, thenormal mechanism by which theproblem is handled is bankruptcy.(Sovereign defaults pose specialproblems, which is why it is impor-tant to note that the debt in East Asiawas private. The governments them-selves had been running surpluses.)

But the IMF was dead set againstbankruptcy and facilitating the pro-cess of debt workout, for example,through a debt moratorium. Its firstdeputy managing director referred tobankruptcy as if it were an abroga-tion of the debt contract, failing tonote that bankruptcy was at the coreof modern capitalism and was an

essential part of limited liability cor-porations. (Critics pointed out thatwhile the IMF was willing to put upbillions to preserve the sanctity ofthe debt contract, it was reluctant to

put up the millions needed to pre-serve the social contract, for example,the food and fuel subsidies for the

poor, and the social and economic

consequences of the abrogation of thesocial contract were far more severethan the consequences of bankruptcycould possibly have been.) Bank-ruptcy (or a debt moratorium) wouldhave relieved downward pressure onthe exchange rate (see Miller andStiglitz 1999~--indeed, it was onlywith the essentially forced rollover ofKorea’s debt that its exchange ratewas stabilized. Those who opposedsuch policies said that it would beimpossible to engineer them, butKorea showed that that was wrong.They argued that capital would notflow back into the country, and thatwas partially irrelevant, partiallyjust wrong, both in terms of theoryand evidence. Capital was not goingto be flowing into these countries inthe short run in any case. And thecountries in East Asia, given theirhigh savings rate, had little need forcapital even in the longer run. Butcapital markets are forward looking:there is not a single participant whocan decide to punish someone whodoes not obey his strictures. Rather,there are a multitude of participants,each of whom must decide onwhether the return is sufficient to

justify the risk. A country in deeprecession, with a large overhang ofdebt, public and/or private, is lesslikely to attract funds than a countrythat has put the past behind it. More-



228

over, why should investors hold anew government to blame for mis-takes made by past governments; ifanything, a new government that hasrectified the problems of the past willgain in credibility. All of these pro-vide part of the reason that countriesthat default do regain access to inter-national capital markets, and oftenafter a remarkably short time-per-haps more determined by the time ittakes to restore economic stabilityand growth prospects than anythingelse.

One country in the region took analternative course; after first flirtingwith an IMF program without the

IMF, Malaysia imposed capital con-trols. Its downturn was shorter, andit was left with less of a legacy of debt,as a result. In the next section, weshall explain why, but first, I want todiscuss briefly some of the conse-quences of the increased risk, besidesthe adverse effect on growth that Ihave already discussed.

The consequences of increasedmacro-economic risk. Normally, ascountries go into a downturn, it is thepoor who disproportionately bearthose costs. Their unemploymentrate goes up more,$ perhaps becauseemployers value of the firm-specifichuman capital of the more skilledworkers, and as their demand for la-bor decreases, they would rather re-deploy them to less skilled jobsrather than having them leave thefirm.

Less developed countries typicallyhave limited safety nets. Even indeveloped countries, unemploymentinsurance in the self-employed andagricultural sectors is limited, and in

developing countries, these sectorspredominate. In some less developedcountries, flexible labor marketsimply that reductions in the demandfor labor do not show up in the form of

unemployment but are reflected inchanges in real wages. But the reduc-tions in real wages may be verylarge-in some of the countries inEast Asia, real wages fell by morethan a quarter.

One of the important argumentsto emerge from the 2000 World De-

velopment Report (World Bank 2000)is that the poor are not onlyadversely affected by lower incomesbut also by higher insecurity. De-prived of the instruments with whichto deal with economic volatility, theirlives are particularly affected by theinstability that is associated with

capital market liberalization.9Moreover, extended periods of

high unemployment and low wagescan have a devastating effect inundermining social capital, the socialfabric that enables a society, and amarket economy, to function-wit-ness the increase in urban violence inLatin America following the debt cri-sis. Not only are there severe socialconsequences, but the social instabil-

ity provides adverse conditions forinvestment and thus growth.

The defenses

Given the overwhelming theoryand evidence against capital marketliberalization, one might wonder, onwhat grounds did the IMF argue inits favor? While they refused torefute the arguments put forwardabove, in particular, never address-ing the issue of the impact of liberal-ization on the poor, they argued that



229

capital market liberalization en-hanced growth and actually reducedrisk!

Growth. Underlying the analysisis a simple analogy: free mobility ofcapital is like free mobility of goods;and just as free trade increases in-comes, so too does free mobility ofcapital. But capital markets are dis-tinctly different from goods markets.Risk and information are at thecenter of capital markets: capitalmarkets are concerned with the ac-

quisition, analysis, and dissemina-tion of information; with makingchoices about how to allocate ofscarce capital to investment opportu-nities ; and with spreading, sharing,and pooling risks. Markets for infor-mation are markedly different frommarkets for goods. While with perfectinformation and perfect risk, compet-itive markets are in general (pareto)efficient, with imperfect informationand incomplete risk markets, mar-kets typically do not behave in theway predicted by standard competi-tive models, and market equilibriumis in general not (constrained pareto)efficient.&dquo; Thus, while there maybe some presumption that trade lib-eralization may be welfare improv-in,&dquo; there is little basis for

presuming that liberalization in fi-nancial and capital markets is wel-fare improving

There are two more specific argu-ments that the advocates of capitalmarket liberalization put forward.One is patently wrong: that withoutcapital market liberalization, coun-tries will not be able to attract the

foreign direct investment, which is soimportant to economic growth. China

has not liberalized its capital market,and yet has been able to attract moreforeign investment than any otheremerging market.The second is more subtle. It

argues that capital market liberal-ization provides an important disci-pline device-countries that fail topursue good policies are quickly pun-ished, and thus capital market liber-alization helps keep countries on asolid path of economic reform.Underlying this argument is a highlyantidemocratic bias: the belief thatdemocratic processes provide aninadequate check and the willing-ness to delegate discipline to foreignfinancial interests. But the argumentis more problematic. If one is tochoose an outside disciplinarian, onewants one that punishes one if andonly if one has &dquo;misbehaved.&dquo; But asmany countries-for example, inLatin America-learned with greatpain, with capital market liberaliza-tion, they can be punished even ifthey do everything &dquo;right.&dquo; If emerg-ing markets fall from favor, in theinevitable vicissitudes that charac-terize capital markets, then even thecountries that have been awarded A’sfrom the IMF are punished. Equallybad, capital markets may have cer-tain biases-they may, in the firststance, overreact to certain actions acountry undertakes and fail to reactto other actions.

This point may be highlighted byconsidering the consequences of dele-gating the responsibility of &dquo;discipli-narian&dquo; to labor markets. Labor mar-kets might discipline a country forfollowing bad environmental policieswith a rush of skilled labor out of a

country should it decide, for instance,



230

to allow arsenic in its water supply.The choice of a disciplinarian deter-mines what a country is to be

rewarded, or punished, for-andtherefore affects what a country doesor does not do. It affects the verynature of the evolution of society.Who gets to play the role of disci-

plinarian is affected by mobility.Capital market liberalization doesgive capital markets more power, inthis sense. And in doing so, it affectsthe ability of society to redistribute:any threat to increase the taxes on

capital can result in quick retribu-tion in the form of the withdrawal offunds. Whether such funds contrib-ute to the long-term growth of theeconomy is not the issue: the with-drawal of funds can impose enor-mous costs, especially in the contextof the IMF-style responses

Enhancing the mobility of capitalthus has real consequences: it affects

bargaining positions and the out-come of bargaining processes in waysthat are advantageous to capital anddisadvantageous to labor.

The argument that capital marketliberalization was good for growthwas, to be sure, fairly unpersuasivein the context of the countries in East

Asia, where domestic savings rateswere very high. Although the coun-tries did an impressive job in invest-ing these savings productively, littleargument could be put forward thatadditional funds from more devel-

oped countries would substantiallyincrease growth.

In the face of this, the IMF and theU.S. Treasury used an even morepeculiar argument.

Risk. They argued that capitalmarket liberalization would reduce

risk, enabling countries to have ac-cess to outside funds in the face of athreatened economic slowdown. Di-versification of the source of fundingwould enhance economic stability.What was remarkable about this ar-

gument was the overwhelming em-pirical evidence against it, even atthe time it was put forward: short-term flows of funds are procyclical,exacerbating, not dampening, eco-nomic fluctuations. As the expressiongoes, bankers are most willing to pro-vide credit to those who do not need

it; and as a country faces a downturn,bankers withdraw credit.&dquo;

Externalities and

capital controls

Thus, today, there is widespreadagreement that capital flows imposehuge costs on others-on innocentbystanders, small businesses, andworkers who neither participated innor benefited from these flows. Theyimpose huge externalities. Wheneverthere are externalities, there arestandard remedies-governmentinterventions-that can take a vari-

ety of forms, for example, regulatoryor tax. There is a large literatureaddressing the relative merits of var-ious forms of interventionOne of the standard objections to

these interventions is that they raisethe cost of funds, especially short-term funds. But that objection is likethe steel industry complaining thattaxes on pollution will discourage theproduction of steel. It will-but thatis precisely the point. Efficiency



231

requires that the marginal social costof production equal the marginalsocial benefit; and the steel industry,in its private calculations, does notinclude the social cost of pollution.Once that is included, productionshould be reduced. So too, since thereis a large social cost associated withshort-term capital flows (and ques-tionable social benefits), these capi-tal flows should be discouraged.

The battle of the metaphors. Whilethe economic case for some form of in-tervention is compelling, opponents(and proponents) of interventionhave often resorted to metaphors andfalse analogies to help &dquo;prove&dquo; theirpoint. We have already disposed ofone such: the argument that the freeflow of capital is just like the free flowof goods; and just as free trade is wel-fare enhancing, so are free capitalmovements.

One popular metaphor hasinvolved the automobile. Critics of

capital market liberalization haveargued that capital market liberal-ization, at least for most developingcountries, is like giving a teenage kida high-powered car before makingsure that the tires were in good con-dition and before installing seatbelts,let alone airbags. They noted thatwhen there was an isolated accidenton a highway, one might infer thatthe problem was with the driver, butwhen there were repeated pile-ups atthe same bend in the highway, theproblem was more likely with thedesign of the road. Supporters of cap-ital market liberalization respondedthat the appropriate response was towiden the highway, not to return tothe days of the horse and buggy, and

in the meanwhile, the drivers neededto be better trained. Critics respond-ed that roads and cars have to be

designed for ordinary mortals; if onlythose with years of experience asracetrack drivers can survive, thensomething is fundamentally wrong.Moreover, they suggested that theonly repair work on the road systemthat the international communityhad proposed was better road signs(improved information)-and eventhat initiative was halfhearted and

incomplete, as the United Statesrefused to allow the posting of signsat the most dangerous turns (disclos-ing information concerning the activ-ities of hedge funds and offshorebanking centers).

Another popular metaphor lik-ened small developing countries tosmall boats on a rough and wily sea.Even if well designed and well cap-tained, they are likely eventually tobe hit broadside by a big wave andturned over. But the IMF program of

capital market liberalization had setthem forth into the most tempestu-ous parts of the sea, in boats thatwere leaky, without life vests orsafety nets, and without training.

Still a third metaphor involvedaviation: the undersecretary of Trea-sury (who at one time, before he hadtaken up the job of representing WallStreet’s interests, had argued thatfailing to regulate capital marketswas like failing to regulate nuclearpower plants-doing either was aninvitation to disaster) used to arguethat simply because planes occasion-ally crash was no reason to give upflying. But critics responded: but if aparticular model of a plane consis-tently crashed, one would want to



232

ground it; and all governments takestrong policies to ensure that thosewho fly planes are well trained, andthose who fly more powerful planeshave better training. And one cer-tainly wanted to be particularly care-ful in flying over territory where theterrain was particularly rough sincethe dangers of a crash landingwere then particularly severe, andeven more so if the inhabitants inthe territory had a penchant forcannibalism.

The metaphors, of course, werehardly a substitute for deeper eco-nomic analysis. But, especially whenaccompanied by such an analysis,they helped bring home the concernsand the depth of passions, on bothsides. For instance, the fact that, asPaul Volcker, former governor of theFederal Reserve Bank, emphasized,the total stock market of a countrylike Thailand was smaller in sizethan a medium-sized American com-

pany like Home Depot, and that evenwhen well managed, such companiesexperience huge volatility in theirmarket value, brought home thepoint that the developing countrieswere like small boats on a rough sea.Analytic studies showing that mostof the shocks that developing coun-tries experience were external-caused, for instance, by suddenchanges in investor sentiment in themore developed countries, havingnothing to do with the particular pol-icies and events in their particularcountry-provided the answer to thecharge that the problems were of thecountry’s own making; and by thetime the East Asian crisis of 1997had become the global financial crisisof 1998, touching even the best

managed of developing countries, ithad become clear that the rhetoric

blaming the countries of East Asiafor the crisis had been largely self-serving. By the same token, the pri-vately financed but governmentengineered bailout of the world’slargest hedge fund, Long Term Man-agement Corporation, on thegrounds that the failure of this onefirm would exacerbate markedly theglobal financial crisis, provided theanswer to the IMF study arguingthat speculative hedge funds did notplay an important role in the 1997crisis, partly because they were sim-ply too small to do so-and furtherundermined the credibility of thosewho claimed that capital market lib-eralization had little to do with the

instability. The sad fact, though, wasthat advocates of capital market lib-eralization had forced developingcountries to liberalize, without anyanalytic basis showing that it wouldbe good for growth, ignoring the the-ory and evidence that it imposedenormous risks, without a clear set ofguidelines for the circumstances inwhich countries might be able to bearthose risks, and without a set of pre-scriptions for how they might pre-pare themselves appropriately fordealing with them.

The purposes of interventions

Having established that, in princi-ple, some form of intervention isdesirable, the next natural questionis, are there forms of interventionwhere the benefits exceed the costs,that is, there are not ancillary coststhat more than offset the benefits, orin which evasion is not so large thatthe benefits are largely eroded?



233

Interventions by Chile, Malaysia,and China, among others, showedthat at least some countries could

manage such interventions well andthat they could take on a variety offorms. These countries demonstratedthat these interventions need nothinder economic growth-or eventhe ability of a country to attract for-eign funds-but they could helpensure economic stability. Before dis-cussing the alternative interven-tions, it is desirable to describe themultiple purposes that such inter-ventions might serve.

Stabilizing capital flows. The on-rush of short-term capital into acountry poses two problems: first, itcan result in inflationary pressures;and second, such money can leave acountry just as fast as it can enter,leaving in its wake economic devas-tation. On purpose of intervention isto stabilize the flows. Note that inter-ventions designed for this purpose(or several of the other purposes de-scribed below) do not have to be per-fect to be effective. Two metaphorsbring this point home: a leaky um-brella can still be useful in a thunder-

storm ; even if one gets damp, it isbetter than being drenched. The pur-pose of a dam is not to stop the flow ofwater from the melting of snow fromthe mountaintop to the ocean butmerely to stabilize it; without thedam, the onrush of water can causedeath and destruction; the dam canconvert this natural disaster into asource of water for food and suste-nance. Even with a good dam, therecan be spillage; some of the water cango around the dam. Even if it does not

stop every flood, it can contribute

greatly to increased well-being.Most of the well-known ways of

avoiding many forms of capital mar-ket controls (discussed below) do notundermine the ability of such inter-ventions to stabilize flows, for most ofthe evasion tactics (e.g., under- andoverinvoicing) work slowly. They aremore like the flows of water goingaround the dam; in the long run, theaggregate amounts may be signifi-cant, but in the short run, the flowsare still moderate, and it is the hugeflows that cause the problem.

Dampening the rush of capital outof a country. In the event of a crisis,there may be an irrational pessi-mism, matching the irrational exu-berance that brought the capital intothe country. Policies designed tomake it more difficult or more costlyfor capital to leave a country slow therush of capital out; and, like the cir-cuit breakers that have been put intostock markets, the extra &dquo;pause forreflection&dquo; can have large positive ef-fects. Before they fully work outmechanisms for avoiding the con-trols, matters are seen in a calmerlight-and markets themselves mayhave calmed down.

Designing more e ffective and lowercost stabilization measures for theeconomy. In simple models, wherethere is free flow of capital, there islittle scope for monetary policy. A de-crease in interest rates leads to arush of capital out of the country.Thus, governments must rely oncostly fiscal policy measures to stim-ulate the economy in the event of an



234

economic downturn. If, however,there are effective restrictions onshort-term capital movements, thenmonetary policy can be used. Thishas benefits both in the short run andthe long, as Malaysia so forcefullyshowed. Reliance on fiscal policyforces governments to have large def-icits, which can put a damper on fu-ture growth. Financing the deficit isalso problematic in countries withlimited access to foreign borrowing.Government borrowings crowd outprivate investment, with the net af-fect on recovery limited. At the highinterest rates, lending to firms be-comes particularly risky (Stiglitz-Weiss 1981), and banks prefer whatthey perceive to be relatively safeloans to government. In countrieswhere many firms have high lever-age, the high interest rates inducemassive corporate distress, weaken-ing the banking system and reducingits ability to lend even more. The costto the public of resolving the corpo-rate and financial stress is all the

larger, again with adverse effects onthe country’s future growth.

Providing greater scope forredistributive taxation. The irony isthat while short-term capital im-poses enormous costs on society, theability of a country to tax such capi-tal, should it become fixated on keep-ing it, is limited. This is a reflection ofa general principle in taxation: gov-ernments can impose only limitedtaxes as factors whose supply ishighly elastic. Capital market liber-alization, in effect, enhances the elas-ticity of supply, thereby lowering thescope for redistribution. As a second

irony, the fact that short-term capital

increases economic volatility meansthat the return that it must receive-to compensate it for the risks that ititself has caused-is higher. Hence,capital market liberalization candrive up the before-tax returns at thesame time that it reduces the scopefor taxation.

Preventing massive capital out-flows. The rush of capital out of Rus-sia has played an important role inthe economic demise of that country;China’s investing its huge savings in-side the country has similarly beencritical in its success. With open capi-tal markets, the oligarchs in Russiawere posed with a simple choice:where in the world to invest theirwealth-whether in Russia, whichwas going into a prolonged depres-sion of an almost unprecedentedscale, or in the United States (say),which was, at the time, experiencingone of the strongest expansions in itshistory, with a stock market boom tomatch? The fact that the wealth ofthe oligarchs was widely perceived tobe ill-gotten, based on political con-nections in a process of privatizationwithout political legitimacy, andtherefore (rightly perceived) subjectto be reversed in a subsequent ad-ministration only reinforced the wis-dom of taking the money out of thecountry. And as each oligarch (andsmaller investors) decided to do so, itmade it more desirable for others doso.

Although of all the objectives ofintervention listed, this may be themost difficult to achieve, the analysisabove suggests that even when thepurpose is discouraging long-termcapital outflows, the interventions



235

may be effective even when there are

ways of circumventing them. This isbecause there may be multiple equi-libria ; if most people keep theirmoney in the economy, it grows, andit becomes attractive for others to doso. Conversely, if most people pulltheir money out of the country, it

becomes attractive for others to do so.The restrictions on capital outflowscan &dquo;force&dquo; the economy to the &dquo;good&dquo;equilibrium, and once there, it is self-sustaining.

The forms and mechanismsof interventions

Intervention has taken on a num-ber of different forms and been

implemented through a number ofdifferent mechanisms. Some of these

rely more on &dquo;market mechanisms&dquo;that are both more flexible and thatavoid the opprobrium associatedwith the term controls

Taxes on capital inflows. For a longtime, Chile had an effective system ofwhat amounted in effect to a tax onshort-term capital inflows. (A third ofthe money coming into Chile had tobe deposited in the central bank for ayear at a zero-interest rate; hence,the first-year return was taxed, in ef-fect, at 33 percent.) The tax rate couldvary with economic circumstances-

discouraging inflows more when theyseemed to pose a greater problem. Inprinciple, a subsidy could be providedif the government wished to encour-age an inflow.At the same time, the tax on

inflows discouraged speculative out-flows : an investor worried about the

possibility of a small devaluationwould not find it attractive to take

his money out of the country over-night to bring it back in again thenext day, for there was a large effec-tive tax on such a roundtrip. Therewas little evidence that the tax dis-

couraged overall inflows, but it

lengthened the maturity of funds,thus stabilizing the economy. Chilewas, of course, adversely affected bythe global financial crisis, as were allcountries, and especially countriesheavily dependent on commodityexports. No one believed that it

would eliminate all sources of insta-

bility. In the aftermath of the crisis,as funds for emerging markets driedup everywhere, Chile decided thatthe problem was not a surfeit of fundsbut a lack of funds, and hence the taxwas reduced to zero.

Controls on capital outflows. Inthe uncertainty of the early days ofthe global financial crisis followingRussia’s default, Malaysia respondedby imposing controls on capital out-flows. It noted high levels of specula-tive activity on the ringit, especiallyoccurring in Singapore, and worriedthat such speculation would de-stabilize the economy. The controlswere carefully designed to ensurethat those who had invested longterm in the country would be able totake their profits out. And they wereannounced as short term, to be re-moved within a year. The vituper-ativeness with which these controlswere greeted was almost unprece-dented, not only from the IMF, theself-appointed guardian of capitalmarket liberalization, but also fromthe U.S. Treasury. They forecast thatthe controls would be ineffective;that the country would never be able



236

to attract capital; that they would becounterproductive, exacerbating theeconomic downturn; that they wouldnever be removed; and that withoutthe discipline of capital markets, thecountry would never address its

problems. Their forecasts were as farfrom the mark and as based on ideol-

ogy and interests as their earlier ad-

vocacy of capital market liberaliza-tion had been. Just as Malaysia hadonce before imposed capital controlsin an emergency and removed themas promised, so too did it fulfill itscommitment. Its downturn wasshorter than any of the other coun-

tries, 16 and the country was left withless of a legacy of indebtedness. (Seethe discussion in the preceding sec-tion.) It did use the time well to re-structure, with a program that wasfar more effective than that of its

neighbor, who remained under IMFtutelage. Capital (foreign direct in-vestment) continued to flow into the

country 1’The World Bank worked with

Malaysia to convert the controls intoan exit tax, with the tax rate gradu-ally lowered. The result was thatwhen the controls (taxes) werefinally removed, there was no distur-bance to the market: it was a virtu-

ally seamless change.

Bank regulations. Today, increas-ingly, capital controls are imple-mented through bank regulations,which limit not only the (uncovered)foreign exchange exposure of banksbut also of the firms to which theylend. Since most financial transac-tions are intermediated throughbanks and most domestic firms bor-

row at home as well as abroad, theseregulations can be highly effective.Even before the crisis, Malaysia hadsucceeded in limiting the foreign ex-change exposure of its banks.

Such changes can be implementedeither through direct regulations orthrough more price-based mecha-nisms, for example, through depositinsurance systems where the premiaincrease with risk and where the for-

eign exchange exposure of the bank(direct and indirect) is included inthe risk measure, or through riskadjusted capital adequacy require-ments, again where foreign exchangeexposure is included in the riskmeasure.

Taxes. Many countries have indi-vidual and corporate income tax sys-tems that allow the deduction ofinterest payments. By disallowingthe deduction of interest on short-term foreign denominated debt,households and firms would be pro-vided with an incentive not to under-take such debt.&dquo;

EXCHANGE RATEREGIMES AND CAPITAL

MARKET CONTROLS

So far, we have argued that capitalcontrols increase risk but do notincrease growth. We have elided thequestion of the extent to which theseresults are dependent on the ex-change rate regime. In this section,we shall argue that in the absence of

capital controls, the only exchangerate regimes that, in practice, canwork effectively are floating ex-change rates or dollarization, but



237

that even with floating exchange ratesystems, capital controls can en-hance economic stability.

Why fixed exchangerate systems failwithout capital controls

A major failing of fixed exchangerate systems is their vulnerability tospeculative attack, especially whenthere is a perception that theexchange rate is overvalued and can-not be sustained. Countries are not

given the grace of a gradual adjust-ment. When there is not sufficientreserves to back up the demands for

dollars, then there can be a run on thecurrency, just as there can be a run ona bank when there is not sufficientreserves available to meet its liabili-ties (see Diamond and Dybvig 1983).If all creditors and potential claim-ants believed that the country couldmeet its obligations, then, of course,they would not wish to &dquo;cash in,&dquo; topull their funds out of the country;but if they believe that the exchangerate is not sustainable and will crash,the returns to doing so are enormous.They can pull their funds out today,putting them back in tomorrow, andmake an enormous return from the

speculative activity.The problem is that the amount of

reserves required to ensure that acountry can meet its commitment tomaintain the exchange rate is enor-mous under full capital market liber-alization : it equals the total value ofthe money supply plus short-term,foreign-denominated credit, for anydomestic currency can be convertedinto dollars on demand. In short, thecountry has to have a fully backedcurrency-equivalent to the

elimination of fiat money. In effect,then, a country with full capital mar-ket liberalization surrenders controlover its money supply and monetarypolicy. The consequences are not onlythat the government cannot engagein stabilizing macro-policy, but alsothere may actually be a destabilizingdynamic put into place.

Assume, for instance, its firmsdecide to borrow more abroad. Itmust then increase reserves or take

strong actions to discourage such for-eign borrowing. Assume that themantra is that the country not onlycannot control the capital inflows,but it cannot tax them or the uses towhich the funds are put. Then, if itwishes to add to reserves, it will putdownward pressure on the exchangerate. But given that it is committed tomaintaining the exchange rate at itsfixed level, it must offset this pres-sure by increasing the interest rate.It might well justify that measurefurther by noting that it dampens theinflationary pressures that are oftenassociated with the capital inflows.But this induces domestic firms that

expect the government to fulfill itscommitment to a fixed exchange rateto borrow even more abroad, espe-cially if foreign borrowing is being, ineffect, encouraged by the foreignlenders, as was the case in East Asia.(The Basle capital adequacy stan-dards provide preferential risk treat-ment for short-term lending, and theunderregulation of the undercapital-ized Japanese banks provided themwith an incentive to engage in riskylending.’9) There is thus a vicious cir-cle, one that can easily lead to mas-sive distortions of resource alloca-tion, as in the case of Thailand: the



238

foreign lending goes into areas thatare collateralizable, feeds a specula-tive real estate boom, while the highinterest rates meanwhile choke offmore valuable domestic invest-ment.2° It made little sense to buildempty office buildings in Bangkokand Jakarta when there wereother investments that would have

enhanced the growth prospects andjob opportunities. Yet, that is whathappened under deregulation.

There is an alternative strategythat few countries have followed. To

discourage foreign borrowing, thecountry can make loans available atmore attractive terms to domesticborrowers. One argument holds thatlowering interest rates to domesticborrowers will lead to lower depositrates and a possible flow of moneyout of the country This argument,while often mentioned, is unpersua-sive : the objective was to stem anexcess flow of foreign borrowing, andthere must exist a &dquo;fixed point,&dquo; alevel of interest rates that attractsthe desired level of capital.

But there is another argumentthat is somewhat more compelling.The lower interest rate will normallylead to an increased level of domesticinvestment and hence possibly con-tribute to inflationary pressures. Ineffect, a country in such a situation isforced to cut back on public expendi-tures or increase taxes in response toan onslaught of foreign capital, nomatter whether that onslaught isbased on irrational exuberance ornot. If such policies had been pursuedin Thailand, investments in emptyoffice buildings would have crowdedout higher return investments ineducation or infrastructure or forced

politically unpalatable increases intaxes-in a country already runninga fiscal surplus.

(There is still a third alternative

strategy, from which the countrieswere discouraged under the doc-trines of liberalization, and that ismicro-economic interventions, suchas taxing real estate capital gains.)

The dynamic is worse than justdescribed: If the monetary and fiscalmeasures designed to maintain theexchange rate do not, at the sametime, perfectly offset any inflationarypressures, then the capital inflowmay well lead to a real appreciation,leading to a trade deficit. (In somesense, the trade deficit is the inevita-ble accompaniment of the capitalinflow, and the trade deficit willtypically&dquo; be generated by a realappreciation.) But the increasinglylarge trade deficit, under standarddoctrines and models, will be viewedas &dquo;unsustainable.&dquo; That is, marketsmay increasingly anticipate a correc-tion in the overvalued exchange rate,a correction more likely to occurthrough a sudden change in theexchange rate than in adjustments inwages and prices.

The fact of the matter is that few

governments have been willing tomaintain high-cost reserves equal totheir money supply and foreign-denominated short-term indebted-

ness, and short of that, the countrieswill be vulnerable to a speculativeattack. Even the massive amount of

money that the IMF has been able tomobilize in recent crises is not

enough to fill the gap and therefore tomaintain confidence in the overval-ued exchange rate.



239

But the fact that the IMF has engi-neered massive (and typically unsuc-cessful) bailouts has exacerbated theoverall problem in three distinctways. First, in effect, the IMF has fedthe speculative sharks (though thecost is borne by the taxpayers in thedeveloping country). In the absenceof outside funds, speculation is azero-sum game, with some specula-tors gaining at the expense of others.The IMF engineered bailouts convertwhat would be a zero-sum game intoa positive-sum game for speculators:they stand to gain at the expense oftaxpayers, and they have indeedgained handsomely. Second, thefunds have facilitated the bailout of

creditors, generating the moral-hazard problem-lenders do not bearthe full costs of their lending deci-sions (even ignoring the macro-economic externality). IMF econo-mists (including their chief econo-mist, Michael Mussa) have arguedthat the lenders have borne some

costs, but that is not the point: moral-hazard problems arise wheneverthey do not bear the full costs, andthis they clearly have not. Third, theIMF efforts to sustain the exchangerate (even if only partially success-ful), and the rhetoric that, otherwise,borrowers who have borrowed in for-

eign denominations will be hurt,have led to a &dquo;foreign exchange covermoral hazard.&dquo; Private borrowershave felt that they do not need to buyinsurance against the risk of devalu-ation or as much insurance as theyotherwise would, and in this they areright: when enough of them take thatposition, the IMF will use that fact tohelp support the exchange rate. Theyare willing to force small firms to

bear the costs through high interestrates to save those that should havepurchased insurance. Doing so, itcould be argued, is not only contrib-uting to a moral-hazard problem butto a moral problem.

The arguments put forward forfixed exchange rate systems applywith equal force to controlledexchange rate systems, for example,where the exchange rate is allowed tomove within a well-defined band.When it becomes apparent that theband cannot be sustained, there willbe a speculative attack.

Dollarization. While capital mar-ket liberalization has thus madefixed exchange rate systems of theconventional kind untenable, it hasenhanced the argument for

dollarization, in which the countrygives up control over its money sup-ply. Under standard criteria, Argen-tina and the United States, orEcuador and the United States, donot constitute an optimal currencyarea (see Mundell 1961). ButMundell (1961) wrote his classic arti-cle before capital market liberaliza-tion was the vogue. The shocks facingEcuador and the United States are

markedly different, and giving upconventional monetary policy in-struments will impair the abilityto stabilize the economy. But the

alternative-allowing the country tobe buffeted by speculative exchangerate movements-may be evenworse, and even with dollarization,there may be some scope for mone-

tary policy (see Stiglitz 2001b).

Yolatility among the major cur-rency areas. Dollarization, however,



240

is not really a viable solution forcountries engaged in trade withmany different countries, for exam-ple, Japan, Europe, and the UnitedStates, simply because of the hugevolatility of the exchange ratesamong their currencies. Fixing theexchange rate to the dollar meansthat firms face enormous risks in the

exchange rate with Japan and Eu-rope. In short, there really is no suchthing as a fixed exchange rate, onlyan exchange rate that is fixed interms of one of the many currenciesthat the country interacts with. Sta-bilizing on a basket of exchange ratesdoes not solve the problem of particu-lar firms. It leaves a firm that exportsin the dollar zone, or imports fromthe yen zone, or imports from onezone and exports to another, bearingenormous risk. It does a firm little

good to know that on average, ex-change rates are stable, if it facesbankruptcy, because imports haveundermined it. Today, small coun-tries around the world have to facethis challenge in risk management:there is nothing they can do aboutthis volatility. But given the highlevel of risk that they have to managein any case, the burden of managingthe additional risk posed by short-term speculative capital flows in theabsence of capital controls is all thegreater.

Flexible exchange rates

Some critics of Thailand suggestthat the problems it faced in the EastAsia crisis lie with the fixed exchangerate system, but that contention iswrong (See Furman and Stiglitz1998a). Had the exchange rate beenallowed to adjust, it would have

appreciated, increasing the tradedeficit, distorting the economythrough that channel. When the col-lapse came, it might have even beenworse, simply because the fall in theexchange rate would have been froma higher, more overvalued level.Some argue that investors werelulled by the seemingly fixedexchange rate to take a more exposedposition than they otherwise wouldhave, but this argument is unpersua-sive on two grounds. First, prudentbehavior required the purchase ofinsurance, and insurance marketsare particularly well designed (inprinciple) for handling the risks asso-ciated with fixed exchange rate sys-tems-small probability events withlarge consequences. There never hasbeen a truly fixed exchange rate sys-tem ; fixed exchange rate systemsonly mean that adjustments occur inlarge steps but infrequently. If themarket shared the investors’ percep-tions that the probability of anadjustment was small, then theinsurance premium would have beencorrespondingly small. Thus, the fail-ure to obtain cover, exposing them-selves and the country to large risk,was as much, or more, a case of mar-ket failure than of government fail-ure, of markets either being irratio-nal (investors believing that theyknow better than the market, asreflected in insurance premia, aboutthe future course of exchange rates)or inefficient, with the cost of coverbeing excessively high relative to therisk being divested. Second, therehave been &dquo;crashes&dquo;-rapid changesin asset prices-in &dquo;flexible&dquo;-pricemarkets as there have been in fixed-

price markets.



241

With flexible exchange rates,there are high costs of the absence ofcapital controls, especially given theimperfections of risk markets.&dquo;Changes in speculative attitudes,say, toward the exchange rate, canforce the exchange rate up or down,imposing huge problems for export-ers and those in import competingsectors, or even to domestic produc-ers relying on imported inputs.Typically, neither the producers norconsumers can divest themselves ofthe resulting large risks. Especiallyin the presence of imperfections ofcapital markets,23 the costs of suchrisks can be enormous: workers,underprotected by social safety nets,cannot borrow against the prospectof future income; firms may be forcedto shut down, with an enormous lossof firm-specific human capital andorganizational capital. And theanticipation of such costs will makeinvestment in the country lessattractive. Even when there arefutures and forward markets, theyextend only to a limited extent intothe future, not enough to deal withthe risks associated with long-termreal investments.

Macro-stability. The huge volatil-ity in exchange rates provides realchallenges (to put it mildly) on thoseresponsible for macro-stability, espe-cially if traditional IMF/central bankresponses are employed. A loss ofconfidence in the currency will lead

depreciation. If true free-market

principles were adhered to, so thatthe government simply allowed theexchange rate to be whatever themarket determined, then the govern-

ment would simply have to apprisethe adverse real balance effect offirms and households that were net

foreign debtors, the positive real bal-ance effect of those who were net

creditors, and the positive effect onnet exports (taking into account thedynamics of adjustment). Fiscal andmonetary policy could freely be usedto adjust the level and composition ofoutput, either increasing or decreas-ing the extent of devaluation. (To besure, the calculations of the appropri-ate policies would be complicated notonly by the dynamics of adjustmentbut also by the complexity of expecta-tion formation. But these are detailsthat need not detain us here.)

In practice, the IMF has seldomallowed governments the freedomjust described. It has worried thatdevaluations would lead to inflation,it has inveighed against what itdescribes as competitive devalua-tions (never mind that such devalua-tions are effectively typically aimedat changing the exchange ratesagainst the dollar, not just at gaininga competitive edge over similar coun-tries), and it has worried that withdevaluation, those who owe moneyabroad in dollars would not be able tomeet their obligations or lead to con-tagion. While in other spheres, it hastaken a strong promarket line, in thisarea, it has talked about overshoot-

ing ;24 but it has never provided acoherent explanation for why over-shooting should be more prevalent inthis market than in other asset mar-

kets, why government interven-tion-and, in effect, government sub-sidies-should be more acceptable inthis market than in other markets, or



242

why the interventions should be lim-ited to the particular kinds of inter-vention (high interest rates, fiscalcontraction, direct exchange ratesupport) that it favors. Their argu-ments have rung an increasingly hol-low note: the large devaluations asso-ciated the global financial crisis didnot set off inflationary spirals;Brazil’s large devaluation did notlead to contagion; the bankruptciesthat marked East Asia were as mucha result of the high interest rate andcontractionary fiscal policies put inplace to stave off a devaluation thanto the devaluation itself (though to besure, the devaluation had a greaterimpact on the foreign creditors, theclientele of the IMF, while the highinterest rates had a greater impacton domestic creditors, which seem-ingly were of little direct concern).Adjustments in exchange rates inother noncrisis countries, like Tai-wan, followed along the lines of thecrisis countries: there was no compet-itive devaluation, just an exchangerate adjustment. Most tellingly, care-ful micro-studies, for example, ofThailand, showed that the seemingworry about the impact of devalua-tion on the economy was largelybogus and certainly of second ordercompared to the adverse conse-quences of the high interest ratesand excessive fiscal contraction.Those with large foreign indebted-ness were largely in the real estatesector and already dead; furtherdevaluation would not make them

any deader, and arresting the devalu-ation would not lead to a revival ofthis sector. The second most heavilyindebted group were exporters, whowould, on the whole, gain more from

the devaluation in terms of exportsthan they would lose on their balancesheets.25

But those who come under the

sway of the IMF have to respond tothe devaluation by interest rateincreases and fiscal contractions,which lead to recession and, in somecases, depression. Indeed, the basicframework, which has come to becalled &dquo;beggar-thy-self policies,&dquo; is

designed to bring about an economicdownturn-and to bring with itadverse contagion to neighbors. Acommon (but not universal) charac-teristic of the precrisis situation is atrade deficit.26 Countries are told toredress the deficit; the resulting sur-plus facilitates the ability to repay itscreditors. But given that devalua-tions are discouraged, tariff andother barriers to imports are notallowed, and exports cannot beincreased overnight; the only way todo so is to decrease incomes-cause arecession-which reduces imports.Trading partners, of course, do notcare much about why their exportsare down; all they know is that theyhave fallen. The downturn in one

country is thus transmitted to its

neighbors, just as in the beggar-thy-neighbor policies that played such animportant role in the propagation ofthe Great Depression. But these poli-cies do not even have the savinggrace of helping the domestic econ-omy as they hurt those of tradingpartners. Thus, countries with capi-tal market liberalization (underfixed or flexible exchange rates) thathave their responses to large varia-tions in capital flows dictated by theIMF are likely to fmd themselves



243

confronting the consequences oflarge economic downturns.

In short, with fixed exchange ratesand full capital market liberalzation,the government absorbs some of thecosts that the huge movements inshort-term capital impose underflexible exchange rates; but the gov-ernment’s ability to do so is limited. Ifit wishes to do so, it must bear hugecosts, both in terms of the size ofreserves that have to be maintainedand in terms of its loss of ability tomaintain macro-economic stability.But with flexible exchange rates, fullcapital market liberalization im-poses enormous risks on firms, andwhile there are macro-policies thatmay do a reasonable job of offsettingthe effects, ensuring a modicum ofmacro-stability, in practice, countriesare likely to face significant macro-instability under these exchangerate regimes as well.

CONCLUDING REMARKS

Developing countries differ frommore developed countries in manyways, besides the lower level ofincomes. In particular, they facegreater economic volatility and alower ability to manage that volatil-ity, even though they may have moreflexible wages and labor marketsthan more developed countries (seeEasterly, Islam, and Stiglitz 2000).Capital market liberalization in-creases the risks they face-underany exchange rate regime, althoughit may enhance the arguments forflexible exchange rates. Given theabsence of evidence that it promotesgrowth, given the compelling theo-retical arguments that it may

actually have adverse effects ongrowth, and given the theory and evi-dence that it enhances economic

instability, one might well ask, howcould an international body, the IMF,founded to promote global economicstability be so active in promoting it,going so far as to seek a change in itscharter to mandate it?

A full answer to this would take uswell beyond the scope of this article: amixture of bad economics (using oldmacro-economic models that simplyfailed to incorporate in a meaningfulway finance, although this has beenone of the major areas of advance ineconomic theory in the past quartercentury27), ideology, and specialinterests: financial markets would

gain from the opening up of new mar-kets, and American financial mar-kets wanted them to be opened upquickly, before others were in a posi-tion to take advantage of these newopportunities; and the free-marketideology served these interests well(even if there was a note of intellec-tual incoherence in free-marketers

asking the government to use itspower to force others to open up theirmarkets and in defending multi-billion dollar bailouts for Western

creditors). But the lack of transpar-ency with which the IMF operatesexacerbates these problems: its poli-cies were not subject to the kinds ofintensive scrutiny that should be thehallmark of democratic processes,simply because much of what it doesgoes on behind closed doors, withpublic announcements coming toolate for meaningful inputs from otherstakeholders. Secrecy is the hall-mark of financial markets, and theIMF has borrowed its culture from



244

those it has sought to serve, withwhom it interacts constantly, andfrom whom it draws so much of its

personnel.But underlying all of these prob-

lems is governance: to whom theinstitution is accountable. With vot-

ing rights allocated according to mar-ket power at the end of World War II,with some adjustments since then,with finance ministries and centralbanks speaking for the governments,with other stakeholders precluded fromhaving a seat at the table, the policiespushed by the IMF become under-standable but no more acceptable.28

Capital market liberalization rep-resents a major change in the rules ofthe game. It was a change in the rulesthat did not serve the interests of the

developing countries well. The fun-damental problem facing globaliza-tion is how these rules of the gameare made. Dissatisfaction with thecurrent system is well deserved.

Notes

1. Interestingly, many in the Interna-tional Monetary Fund (IMF) claimed that theynever had really pushed for capital market lib-eralization for countries that were unpre-pared. At best, this was a semantic quibble: al-though they had often accompanied theirdemands for capital market liberalization bydemands for other reforms, they had neversaid not to go forward with capital market lib-eralization until those other reforms weremade.

2. With even the U.S. secretary of treasuryjoining the attack, it was not left just to thenormal bureaucratic processes.

3. The most notorious example involvedthe newly appointed managing director givinga speech in Bangkok, in which he reflectedsome of the new thinking in the fund concern-

ing the risks of capital market liberalization.By the time he reached Jakarta, the Indone-sians already were discussing ways by whichthe new thinking might be reflected in prac-tice. But by then, the IMF staff had reportedlygotten to their new managing director, andthere was quick backtracking: it was put to theIndonesians in no uncertain terms that goingback on capital market liberalization was notto be part of their economic agenda.

4. See below for a more extended discus-sion of the Thai case.

5. The only possible justification might bethat banks in the United States do a better jobat allocating scarce funds in the developingcountry, so much better that income in the

country is higher than it would otherwise havebeen. There is no evidence in support of thisposition.

6. See Furman and Stiglitz (1998a) andRodrik and Velasco (2000). To be sure, otherfactors played a role, some of which are de-scribed below.

7. There were deeper failings in their ar-guments : they seemed to believe that a tempo-rary intervention in the market would lead toa permanent shift in the demand functions, forexample, for investment. The mechanism bywhich this might occur, other than throughsome vague appeal to the intervention result-ing in a restoration of confidence, were neverspelled out. There was no systematic analysisof investor psychology (with empirical supportfor the maintained hypotheses), nor was thereany appeal to rational expectations modelsthat are the normal staple of much of modemmacroeconomics. Krugman (1998) criticizedthe IMF for playing the role of armchair mar-ket psychologists, a role for which they wereeminently unqualified, with a commensu-rately weak track record. Stiglitz (1999) pro-vided a more detailed critique of the underly-ing theories.

8. For an econometric analysis for theUnited States, see Furman and Stiglitz(1998b).

9. In the jargon of standard economics,low-income individuals have a high level ofrisk aversion and have little access to mecha-nisms with which to divest themselves of therisks they face. This can be viewed as an im-portant instance of market failure.



245

10. That is, taking into account the imper-fections of information and the costs of trans-

actions, for example, associated with creatingmarkets. See, for example, Greenwald andStiglitz (1986).

11. But even this needs to be qualified.Newbery and Stiglitz showed that when riskmarkets are imperfect—which they alwaysare in practice—free trade may actually makeeveryone worse off. See Newbery and Stiglitz(1984).

12. In particular, Murdock and Stiglitz(1993) and Hellmann, Murdock, and Stiglitz(1996, 2000) showed that restrictions on finan-cial markets may be welfare enhancing.

13. This is the essential point of the largeliterature on "local public goods." See Tiebout(1956) and Stiglitz (1983a, 1983b).

14. For an analysis of the Latin Americancase, refer to Galvin and Hausman (1996).

15. See, for example, Weitzman (1974) or astandard public sector textbook, such asStiglitz (2000b).

16. See Kaplan and Rodrik (2001). To besure, its downturn was somewhat longer thanit might otherwise have been because FinanceMinister Anwar at first tried the standardIMF recipes (in what was called an IMF pro-gram without the IMF), raising interest ratesand cutting back on public expenditures.Recovery only began when these policies werereversed.

17. There remains some controversy overwhether in 2000 Thailand was more successfulin attracting foreign investment than Malay-sia, with disputes both about data and their in-terpretation. What matters for growth, ofcourse, are greenfield investments, not simplyforeigners buying already existing assets, un-less the funds they provide to the country indoing so are themselves turned into invest-ments. Large fire sales in Thailand might tem-porarily succeed in diverting funds from Ma-laysia but are hardly indicative of a better"strategy."

18. There are certain practical problems inthe implementation of such provisions, whichcan easily be overcome. Because firms will betempted to use derivatives to subvert the in-tent of these tax provisions, there will have tobe netting provisions, with full disclosure ofderivative positions (enforced, e.g., by lawsthat limit the enforceability of derivative posi-

tions that are not disclosed, or giving them ju-nior positions in the event of bankruptcy).Similarly, debt covenants making debt imme-diately callable in the event of certain circum-stances (such as those associated with a crisis)should either be made not enforceable orbonds with those provisions not be given favor-able tax treatment.

19. A situation analogous to that encoun-tered in the United States in the savings andloan debacle.

20. Proponents of capital market liberal-ization (including those in the IMF) underesti-mated these distortions, simply because theydid not understand the functioning of capitalmarkets and the ways that such markets differfrom ordinary markets for goods and services.In their simplistic models, capital in a well-functioning economy (which most developingcountries are not) is allocated (as if) by an auc-tion process to the borrower offering the high-est interest rate, just like any other good is al-located to the buyer offering the highest price.Thus, the fact that real estate was offering thebest interest rates meant that that had to bethe highest return activity. And it was simplyassumed that if there were mistakes in judg-ment, only the lender would bear the cost ofsuch mistakes. In fact, capital is not allocatedby an auction process (see Stiglitz and Weiss1981; Greenwald and Stiglitz forthcoming) butby a screening process, and for an obvious rea-son: those offering the highest interest ratesmay not be those most likely to repay. More-over, we have seen that when large numbers ofdebtors cannot repay the loans, there aremacro-economic consequences, with others be-sides those who have borrowed and lent hav-

ing to bear the costs.21. Although not necessarily: the availabil-

ity of new sources of credit can increase the de-mand for imports, even if relative prices re-main relatively unchanged. This seems tohave been the case recently in Iceland. SeeStiglitz (2001a).

22. Many advocates of capital market liber-alization, especially those that appeal to theanalogy of the benefits of free markets forgoods, fail to appreciate these market failuresand their consequences.

23. Themselves explicable in terms of im-perfect information. See Stiglitz (2000b).



246

24. That is, if the initial exchange rate is140 to the dollar, the true equilibrium is 190 tothe dollar; in the initial adjustment, the ex-change rate may overshoot to 210 to the dollar.

25. Moreover, while the IMF tried to char-acterize there being a trade-off, as we saw ear-lier there was none: the high interest rates in-tended to prevent a devaluation simplypushed the economy deeper into recession,weakening confidence in the country and itscurrency.

26. For instance, Korea, at the time that thecrisis struck, did not have a balance of tradedeficit. In the old world, before capital marketliberalization, the link between trade deficitsand crises was closer.

27. Highlighted by the fact that in the typi-cal macro-models employed by the IMF, bank-ruptcy and default were not modeled, althoughbankruptcy and default were at the center ofthe global financial crisis.

28. For a more extensive discussion of these

points, see Stiglitz (forthcoming).

References

Caprio, Gerard, and Daniela Klingebiel.1996. Bank insolvencies: Cross-

country experience. Policy researchworking paper no. 1620. Washington,DC: World Bank.

Collier, Paul, and Anke Hoeffler. 1998. Oneconomic causes of civil war. OxfordEconomic Papers 50 (4): 563-73.

Demirgüç-Kunt, Asli, and EnricaDetragiache. 1997. The determinantsof banking crises: Evidence from in-dustrial and developing countries.World Bank working paper no. 1828.Washington, DC: World Bank.

—. 1999. Financial liberalizationand financial fragility. In AnnualWorld Bank conference on develop-ment economics 1998, edited by BorisPleskovic and Joseph E. Stiglitz.Washington, DC: World Bank.

Diamond, Douglas W., and Philip H.Dybvig. 1983. Bank runs, deposit in-surance, and liquidity. Journal of Po-litical Economy 91 (3): 401-19.

Easterly, William Russel, Roumeen Is-lam, and Joseph E. Stiglitz. 2000.Shaken and stirred: Volatility andmacroeconomic paradigms for richand poor countries. Michael BrunoMemorial Lecture. Washington, DC:World Bank.

Furman, Jason, and Joseph E. Stiglitz.1998a. Economic crises: Evidence and

insights from East Asia. Brookings Pa-pers on Economic Activity 2:1-114.

—. 1998b. On the economic causes of

inequality Paper presented at IncomeInequality: Issues and Policy Options,a symposium sponsored by the Fed-eral Reserve Bank of Kansas City,Jackson Hole, Wyoming, 27-29 Au-gust, Kansas City, Missouri.

Galvin, Michael, and Ricardo Hausman.1996. The roots of banking crises: Themacroeconomic context. In Bankingcrises in Latin America, edited byRicardo Hausmann and Liliana Rojas-Suarez. Washington, DC: Inter-american Development Bank.

Greenwald, Bruce, and Joseph E. Stiglitz.1986. Externalities in economies with

imperfect information and incompletemarkets. Quarterly Journal of Eco-nomics 101 (2): 229-64.

—. Forthcoming. Towards a new par-adigm for monetary economics.Mattioli lecture, Milan.

Hellman, Thomas F., Kevin C. Murdock,and Joseph E. Stiglitz. 1996. Depositmobilization through financial re-straint. In Financial development andeconomic growth: Theory and experi-ences from developing economies, ed-ited by N. Hermes and R. Lensink.London: Routledge.

—. 2000. Liberalization, moral haz-ard in banking and prudential regula-tion : Are capital requirementsenough? American Economic Review90 (1): 147-65.

Kaplan, Ethan, and Dani Rodrik. 2001.Did the Malaysian capital controls



247

work? NBER working paper no.W8142.

Krugman, Paul. 1998. The confidencegame. The New Republic, October, 5.

Lindgren, Carl Johann, Gillian Garcia,and Matthew I. Saal. 1996. Bankingsoundness and macroeconomic policy.Washington, DC: International Mone-tary Fund.

Miller, Marcus, and Joseph E. Stiglitz.1999. Bankruptcy protection againstmacroeconomic shocks: The case for a

"super chapter 11." World Bank Con-ference on Capital Flows, FinancialCrises, and Policies, 15 April.

Murdock, Kevin C., and Joeseph E.Stiglitz. 1993. The effect of financialrepression in an economy with posi-tive real interest rates: Theory and ev-idence. Mimeo, Stanford University.

Mussa, Michael. 2000. Factors drivingglobal economic integration. Paperpresented at Global Economic Inte-gration : Opportunities and Chal-lenges, a symposium sponsored by theFederal Reserve Bank of Kansas City,Jackson Hole, Wyoming, 24-26 Au-gust, Kansas City, Missouri.

Newbery, David M. G., and Joseph E.Stiglitz. 1984. Pareto inferior trade.Review of Economic Studies 51 (1): 1-12.

Rodriguez, Francisco, and Dani Rodrik.1999. Trade policy and economicgrowth: A skeptic’s guide to cross-na-tional evidence. NBER working paperno. W7081.

Rodrik, Dani. 1998. Who needs capital-account convertibility? Essays in In-ternational Finance 207, 55-65.Princeton, NJ: Department of Eco-nomics, Princeton University.

Rodrik, Dani, and Andres Velasco. 2000.Short-term capital flows In AnnualWorld Bank Conference on Develop-ment Economics 1999, edited by BorisPleskovic and Joseph E. Stiglitz.Washington, DC: World Bank.

Sachs, Jeffrey, and Andrew Warner. 1995.Economic reform and the process of

global integration. Brookings Paperson Economic Activity 1:1-118.

Stiglitz, Joseph E. 1983a. Public goods inopen economies with heterogeneousindividuals. In Locational analysis ofpublic facilities, edited by J. F. Thisseand H. G. Zoller, 55-78. Amsterdam,the Netherlands: North-Holland.

—. 1983b. The theory of local publicgoods twenty-five years after Tiebout:A perspective. In Local provision ofpublic services: The Tiebout model af-ter twenty-five years, edited by G. R.Zodrow. New York: Academic Press.

—. 1999. Knowledge for develop-ment : Economic science, economic pol-icy, and economic advice. In AnnualWorld Bank Conference on Develop-ment Economics 1998, edited by BorisPleskovic and Joseph E. Stiglitz.Washington, DC: World Bank.

—. 2000a. The contributions of theeconomics of information to twentieth

century economics. Quarterly Journalof Economics 115 (4): 1441-78.

—. 2000b. Economics of the publicsector. 3d ed. New York: Norton.

—. 2001a. (with Sebago Associates).Monetary and exchange rate policy insmall open economies: The case of Ice-land. Paper prepared for the CentralBank of Iceland, February.

—. 2001b. The role of the centralbank and monetary policy underdollarization. Paper presented at theCentral Bank of Ecuador, March,Quito.

—. Forthcoming. Democratizing theIMF and World Bank: Governance and

accountability.Stiglitz, Joseph. E., and Andrew Weiss.

1981. Credit rationing in marketswith imperfect information. AmericanEconomic Review 71 (3): 393-410.



248

Tiebout, Charles M. 1956. A pure theoryof local expenditure. Journal of Politi-cal Economy 64 (5): 416-24.

Weitzman, M. L. 1974. Prices vs. quanti-ties. Review of Economic Studies 41(4): 477-91.

World Bank. 2000. World Development re-port 2000/2001: Attacking poverty.New York: Oxford University Press.



initiative for policy dialoguepolicydialogue.org/files/publications/capital_mrkt_lib_s... €...

Documents