The Effect of IMF Programs on Economic

Journal of Development EconomicsŽ .Vol. 62 2000 385–421

www.elsevier.comrlocatereconbase

The effect of IMF programs on economicgrowth

Adam Przeworski a, James Raymond Vreeland b,) ,1

a Department of Politics, New York UniÕersity, 715 Broadway, New York, NY 10003, USAb Department of Political Science, Yale UniÕersity, 124 Prospect Street, P.O. Box 208301, New HaÕen,

CT 06520-8301, USA

Abstract

Using a bivariate, dynamic version of the Heckman selection model, we estimate theŽ .effect of participation in International Monetary Fund IMF programs on economic growth.

We find evidence that governments enter into agreements with the IMF under the pressuresof a foreign reserves crisis but they also bring in the Fund to shield themselves from thepolitical costs of adjustment policies. Program participation lowers growth rates for as longas countries remain under a program. Once countries leave the program, they grow fasterthan if they had remained, but not faster than they would have without participation.q 2000 Elsevier Science B.V. All rights reserved.

JEL classification: C1; O4Keywords: Economic growth; IMF; Selection

Our primary objectiÕe is growth. In my Õiew, there is no longer any ambiguityabout this. It is toward growth that our programs and their conditionality areaimed. It is with a Õiew toward growth that we carry out our specialresponsibility of helping to correct balance of payments disequilibria and, moregenerally, to eliminate obstructiÕe macroeconomic imbalances.

) Corresponding author. Tel.: q1-203-432-6196; fax: q1-203-432-6196.Ž . Ž .E-mail addresses: [email protected] A. Przeworski , [email protected] J.R. Vreeland .

1 Tel.: q1-203-432-5230; fax: q1-203-432-6196.

0304-3878r00r$ - see front matter q2000 Elsevier Science B.V. All rights reserved.Ž .PII: S0304-3878 00 00090-0

( )A. Przeworski, J.R. VreelandrJournal of DeÕelopment Economics 62 2000 385–421386

ŽMichel Camdessus, IMF Managing Director Statement before the United Nations.Economic and Social Council in Geneva, July 11, 1990 .

1. Introduction

Ž .International Monetary Fund IMF programs are controversial. Governmentsthat enter into agreements with the IMF claim that it is for the better, thatopposition to them is uninformed or badly intentioned. Yet general strikes, riots,and ransacking of supermarkets manifest that IMF programs mobilize popularresistance. And scholarly opinion is also divided: statistical findings range all overthe spectrum of possible conclusions. Hence, our question: What is the effect ofIMF programs on growth?

The immediate goals of the IMF concern exchange rate stability and balance ofpayments, and evaluations of IMF programs tend to concentrate on these objec-

Ž . Ž .tives. Thus, Reichmann and Stillson 1978 and Connors 1979 found that IMFŽ .programs had no effect on balance of payments, while Pastor 1987b , Gylfason

Ž . Ž . Ž .1987 , Khan 1990 , and Bird 1996 reported improvements. Most studies findŽthat Fund programs have no effect on inflation Bird, 1996; Edwards and

.Santaella, 1993; Pastor, 1987b; Gylfason, 1987; Connors, 1979 , although Reich-Ž . Ž .mann and Stillson 1978 reported an unclear effect, and Killick 1995 reported

Ž . Ž . Ž .reduced inflation. And while Connors 1979 , Killick 1995 , and Pastor 1987bŽ .found no effect on current account, Khan 1990 and Edwards and Santaella

Ž .1993 discover that it improves.Yet whether or not the IMF programs have positive effects on these short-term

goals, what ultimately matters is whether they induce economic growth and do notconcentrate incomes.2 Indeed, the Articles of Agreement state that the mission ofthe IMF is to ‘‘facilitate the expansion and balanced growth of international trade,and to contribute thereby to the promotion and maintenance of high levels ofemployment and real income and to the development of the productive resourcesof all members as primary objectives of economic policy.’’ As we have seenabove, the former managing director of the Fund has placed economic growth asthe primary objective. But here again the results are ambivalent: while Reichmann

Ž . Ž . Ž . Ž .and Stillson 1978 , Connors 1979 , Pastor 1987b , and Gylfason 1987 reportedŽ . Ž .no effect, Killick 1995 found ambiguous effects, and Conway 1994 argued that

while growth declines in the first year of a program, the negative effects diminishthereafter.

2 Ž .According to Camdessus 1990 , the Fund seeks ‘‘high quality growth,’’ not merely ‘‘growth forŽ .the privileged few, leaving the poor with nothing but empty promises.’’ Pastor 1987a,b , however,

found that IMF programs sharply redistribute incomes from labor to capital in Latin America. Weintend to pursue the distributional effects in a separate paper.

( )A. Przeworski, J.R. VreelandrJournal of DeÕelopment Economics 62 2000 385–421 387

The standard difficulty in evaluating effects of any policy or program isŽ . 3nonrandom selection Heckman 1988 . What we observe in the real world are not

experiments, which would match the ‘‘treatment’’ and the ‘‘control’’ groups, thuspermitting direct inferences about the experimental effects. Indeed, one wouldhope that governments do not enter into the IMF programs as an experiment.These treatments are costly: at least in the short run, they limit national sovereigntyand inflict economic pain. In fact, governments often claim that they are ‘‘goingunder’’ only because the situation is dire and no choice is left but to ‘‘swallow thebitter pill,’’ ‘‘undergo radical surgery,’’ ‘‘take a horse treatment’’: the lexicon ismedical and the operation delicate. Going to the IMF is an act of courage, ademonstration of ‘‘political will.’’

But if countries enter and remain under agreements only when governmentsrecognize that the situation so demands and have the courage to swallow theconsequences, the conditions of countries participating in IMF programs are notthe same as of those which abstain. And if these conditions are not the same, thenthe effects of the programs may depend in some part on their inherent conse-quences and in some part on these conditions. To evaluate the consequences of theprograms, one must therefore distinguish the two components. The procedure isinextricably counterfactual: the task is to compare the performance if countries hadparticipated and not participated in the programs under the same conditions and, ifselection is nonrandom, we cannot always match the observed cases for theseconditions. Hence, we must proceed differently.

Moreover, while some of these conditions may be observable, as most eco-nomic circumstances are, some may not be. ‘‘Political will’’ is one example. Amethodology failing to account for this unobservable variable would overstate thevalue of participation by attributing the positive effects of ‘‘political will’’ to theIMF program. Furthermore, if such selection occurs, merely controlling for

Žobserved variables can increase the bias Achen ,1986; Przeworski and Limongi,.1996 .

Thus, we need to understand first why countries enter and leave the IMFŽprograms. Once we do, we can use the methods developed by Heckman 1976,

.1988 to estimate their effects independently of selection. In the appendices, wepresent the statistical model of selection and a method for analyzing the effect ofprograms on performance. These appendices are lengthy and technical but theissues are complex.

3 Evaluations of IMF programs have long paid attention to selection, from early before–after studiesŽ .Reichmann and Stillson, 1978; Connors, 1979; Pastor, 1987a,b; Killick, 1995 and work comparing

Ž .countries with and without IMF programs Gylfason, 1987; Edwards and Santaella, 1993 , to morerecent work which corrects for observable determinants of nonrandom selection of program countriesŽ .Khan, 1990, Conway, 1994 . Yet, none of these studies allowed for the possibility of unobservedfactors affecting selection and performance.


The article is organized as follows. The first part provides background informa-tion about our data. The second part analyzes why countries enter and stay underIMF programs. The third part provides the results concerning growth. A briefconclusion follows.

2. Background: the data set

In 1944, 44 countries signed the Bretton Woods agreement which establishedthe IMF to maintain exchange rates for international free trade. When the US wentoff the gold standard in 1973, the old exchange system collapsed. The new system

Ždid not need the IMF, and the organization faced a crisis of purpose Pastor,.1987a, 14; de Vries, 1986, 222 . However, the IMF’s purposes also included

w x‘‘providing members with opportunity to correct maladjustments in their balanceof payments without resorting to measures destructive of national or international

Ž .prosperity’’ Articles of Agreement . Fulfilling this function, the IMF transformeditself from a currency regulating institution to an international organizationinvolved in the national policies of much of the third world, particularly since theonset of the debt crisis in 1982. According to our data, the number of countriesunder one or another IMF program grew from 21 in 1974 to 52 by 1983.

ŽToday there are 182 IMF members who are eligible to take out loans or.technically make ‘‘repurchases’’ from the Fund. Membership requires a contribu-

tion to the Fund, called a country’s ‘‘quota,’’ the size of which depends on the sizeof the member’s economy. A member-country can freely draw up to 25% of itsquota to address balance of payments deficits. To draw on more than 25% requiresa special agreement with the Fund. The IMF attaches to these loans conditionswhich oblige countries to undertake specific policies in order to receive the loan

Žinstallments. These conditions typically entail fiscal austerity cutting government. Žexpenditures and increasing taxes , tight monetary policy raising interest rates and

.reducing credit creation , and currency devaluations. Upon entering an agreement,a member government collaborates with IMF officials to develop policy recom-

Ž .mendations, but the IMF ‘‘sets the tone’’ Taylor, 1993, 41–2 .Ž .There are four main types of IMF agreements: the stand-by arrangement SBA ,

Ž . Ž .the extended fund facility EFF , the structural adjustment facility SAF , and theŽ . Ž .enhanced structural adjustment facility ESAF . Polak 1991 describes the differ-

ences among these arrangements as they relate to the conditions, timing, and sizeof the loan disbursements, but he notes that the fundamental objectives of theseprograms do not differ. Thus, we consider only whether a country is under or notunder agreement without differentiation between type of agreement. Note that in

Ž .our full sample, which includes 678 agreements, 598 of them 88% are SBAs.While SBAs are designed to last 12 to 18 months and the other agreements are

Ž .supposed to cover 1 to 3 or at most 4 years, such time limits on participation arearbitrary. The vast experience of countries has been to sign consecutive agree-


ments. Out of the 226 separate spells of one or more consecutive IMF agreementsduring the period from 1951 to 1990, 111 of them spanned four or more years.South Korea, for example, spent 13 years from 1965 to 1977 under consecutive

Ž .agreements, Zaire spent 14 straight years in IMF programs 1976–1989 , LiberiaŽ .spent 15 years straight 1963–1977 , Peru was under consecutive agreements from

1954 to 1971, and Panama from 1968 to 1987. And, after a ‘‘brief’’ stint of onlyŽ .seven straight years of agreements 1961 to 1967 , Haiti went under again from

Ž .1970 to 1989 27 out of 29 years . During the entire period between 1951 and1990, an average completed spell lasted 4.7 years; between 1971 and 1990 thesespells were longer, lasting on average 5.3 years. Consecutive agreements are thusthe rule, not the exception. We consider such consecutive agreements as a singlespell of remaining ‘‘under.’’

ŽWhile our basic data set covers the period between 1951 and 1990 135.countries, for the total of 4126 country-years , some information is not available

for several countries and years.4 Most statistical analyses reported below are basedon 1024 annual observations of 79 countries, all for the post-1970 period. Thus,the availability of data is yet another source of nonrandom selection: it turns outthat countries for which information is available differ in some systematic waysfrom those for which it is not. To test the robustness of our findings, we examinedthem using various stripped versions of the selection model. These tests arereported in Appendix C.

3. Selection

Since a balance of payments deficit or a reserves crisis is the prerequisite forsigning an IMF agreement, one might not expect controversy in predictingparticipation. According to the ‘‘Conditions governing use of the Fund’s generalresources,’’ found in Article V, Section 3 of the IMF Articles of Agreement:

Ž .b A member shall be entitled to purchase the currencies of other membersfrom the Fund in exchange for an equivalent amount of its own currency

Ž .subject to the following conditions: . . . ii the member represents that it has aneed to make the purchase because of its balance of payments or its reserveposition or developments in its reserves.

Yet a balance of payments deficit is not sufficient to explain agreements. WhileŽ . Ž .Santaella 1996 , as well as Goldstein and Montiel 1986 , discovered that a

Ž .balance of payments deficit predicts participation, Knight and Santaella 1997found it not to be important. Other recent studies have also failed to report

4 We list the countries, years, and IMF agreements in our sample in Appendix D. Most of theeconomic data used in this article come from the World Bank. The exception is growth, measured astotal annual growth of output, taken from the Penn World Tables. The political data come from the

Ž .ACLP World PoliticalrEconomic Database see Alvarez et al., 1996 for details .


Žsignificant findings on balance of payments Bird, 1996; Conway, 1994; Edwards.and Santaella, 1993 . According to our data, while the IMF signs most of its

agreements with countries facing balance of payments deficits, it signed 32% ofthem with countries running an overall balance of payments surplus the yearpreceding the agreement. In 20% of agreements, countries had surpluses both theyear before and the year of the agreement.

We find the same with regard to the ratio of international reserves to monthlyimport requirements. Typically countries that participate in IMF programs havelow reserves, averaging 2.54 times monthly import requirements, and countrieswithout agreements hold stronger reserves, on the average 3.64 times monthlyimports. However, many countries enter and remain under agreements with strongreserves. Ghana, for example, entered an IMF agreement in 1983 when its foreignreserves reached their strongest position ever: 4.8 times monthly import require-ments. Uruguay entered a 1-year arrangement in 1990 when its reserves averaged7.7 times monthly imports, more than double the average reserves of the rest ofLatin America. Portugal entered into a 3-year IMF program in 1983 with foreignreserves averaging 9.6 times monthly imports. And after 8 years of consecutiveparticipation in IMF programs, Turkey signed an IMF agreement in 1968 withreserves averaging 9.7 times monthly imports. The Turkish government signedagain in 1969 when reserves were 11.7 times monthly imports, and then signedagain in 1970.5

Why do countries enter into agreements with the IMF when they do not needforeign exchange?

It is often assumed that, because IMF agreements limit national sovereignty byimposing specific conditions, governments will avoid the Fund if they do not needa loan. As Stanley Fischer notes, ‘‘Policy conditionality can be interpreted as a. . . penalty, as seen from the viewpoint of the borrower country’s policy makers’’Ž .Fischer 1999 . In this view, governments accept conditions only when they needthe IMF loan.

Yet governments may enter into an IMF agreement not necessarily for a loan,Žbut because they want conditions to be imposed Spaventa, 1983; Remmer, 1986;

Vaubel, 1986, 1991; Putnam, 1988; Stein, 1992; Edwards and Santaella, 1993;

5 Other economic variables have also failed to provide a complete picture. Regarding inflation,Ž . Ž .Edwards and Santaella 1993 and Goldstein and Montiel 1986 found that countries entering IMF

Ž . Ž .programs had experienced increasing inflation, but Santaella 1996 , Knight and Santaella 1997 , andŽ .Conway 1994 found inflation not to be a significant predictor of agreements. The importance of terms

Ž . Ž .of trade is also a subject of divergent findings, with Conway 1994 and Santaella 1996 showing thatŽ . Ž .it is significant, but Knight and Santaella 1997 finding it not important. Bird 1996 summarizes the

consensus in the literature about variables that increase the likelihood of an IMF program: low levels ofdevelopment, an overvalued exchange rate, and low GDP growth. Yet, as reported below, we do notfind these variables to have significant effects. We do, however, find that high debt plays a role in

Ž . Ž . Ž .entering programs as do Knight and Santaella 1997 , Santaella 1996 , and Conway 1994 .


.Bjork, 1995; Dixit, 1996 . Suppose that a government wants to restructure publicŽ .finances but faces tough opposition see Vreeland, 1999 for details . Tying the

government’s budget proposal to the conditions of an IMF agreement raises thecosts for the domestic constituencies of rejecting the proposal. Turning down thepolicy is no longer a rejection of the government, but a rejection of the IMF,which is costly because it sends a negative signal to creditors and investors. Eventhough the government risks that the opposition will accuse it of ‘‘selling out’’ tothe Fund and thus faces political – ‘‘sovereignty’’ – costs, the IMF agreementmay enable the government to push through policies that otherwise would havebeen rejected.

Note that in order for this strategy to be effective, the IMF’s decision must beindependent from the government. If the government and the Fund always collude,the threat that the IMF will impose rejection costs upon a country is not credible,and without this threat, the government gains no bargaining leverage. Thus, theIMF must make decisions separately from governments. We see the Fund asmotivated by both technocratic and bureaucratic concerns. We agree with VaubelŽ .1986, 1991 , who argues that the Fund seeks to maximize its budget, but we donot neglect the fact that the IMF is constrained by its mandate to promoteinternational financial stability.

Because the decisions of the government and the Fund are separate, participa-Ž .tion in an IMF program is a joint decision Knight and Santaella 1997 . For a

country to enter and to remain under IMF agreements, both the government andthe IMF must want to do so, and they may have different reasons.

A second complication is that the decisions to enter an agreement and thedecisions to terminate it may have different determinants. As we noted above,contrary to what the IMF advertises — that SBAs are intended to last 1 to 2 yearsand the other programs 3 to 4 — countries remain under agreements an average of4 to 5 years. Why does participation continue and why does it end? To distinguishbetween these questions, we study participation in IMF programs as a Markov

Ž .process Amemiya, 1985 .Hence, four decisions are entailed in the interaction between a particular

Ž . Ž .country and the Fund: 1 government’s decision to seek an agreement, 2 IMF’sŽ .decision to seek an agreement, 3 government’s decision to terminate an agree-

Ž .ment, and 4 IMF’s decision to terminate it. Our statistical model, presented inAppendix A, identifies the parameters of the variables that affect these decisionsŽ .Table 1 .

To measure a government’s need for a loan, we include RESERVES, the averageannual foreign reserves in terms of monthly imports. RESERVES has a significantnegative effect on the probability that the government will enter into an IMFagreement. Governments are more likely to participate in IMF arrangements whenforeign reserves are low.

To test whether governments turn to the Fund to lower the budget deficit, weinclude DEFICIT, the annual government budget surplus as a proportion of GDP.

()

A.P

rzeworski,J.R

.Vreeland

rJournalof

DeÕelopm

entEconom

ics62

2000385

–421

392

Table 1Determinants of program participation

Determinants of Determinants ofentering remaining

GoÕernment

Variable Coefficient S.E. Coefficient S.E.

CONSTANT y2.27 0.611 y0.01 0.592Loan RESERVES y0.83 0.424 y0.26 0.475Conditions DEFICIT y0.95 0.277 y0.29 0.329Rejection costs DEBT SERVICE 1.38 0.516 0.65 0.678

INVESTMENT y6.06 1.789 y0.17 1.922Sovereignty costs YEARS UNDER 0.36 0.212 y0.36 0.266

NUMBER UNDER 0.44 0.178 0.38 0.190LAGGED ELECTION 0.87 0.288 y0.01 0.352

IMF

Variable Coefficient S.E. Coefficient S.E.

CONSTANT 2.14 1.241 2.84 2.016Mandate BOP y0.91 0.370 y0.41 0.230Budget NUMBER UNDER y0.73 0.268 y0.39 0.429Negotiation costs REGIME 0.43 0.260 0.33 0.273

()

A.P

rzeworski,J.R

.Vreeland

rJournalof

DeÕelopm

entEconom

ics62

2000385

–421

393

Frequencies of actual and predicted outcomes

Actual Predicted

0 1 Total

0 484 75 5591 66 399 465TOTAL 550 474 1024

Number of observations 1024Log likelihood function y355.77Restricted log likelihood y705.46Chi-squared 699.39Degrees of freedom 23Significance level 0.0000

Variables: CONSTANT: to enter, 1; to remain, dummy variable lagged participation status; RESERVES: average annual foreign reserves in terms of monthly importrequirements; DEFICIT: measured as surplus of government budget as a proportion of GDP; DEBT SERVICE: debt service as a proportion of GDP; INVESMENT: real

Ž .gross domestic investment private and public as a proportion of GDP; LAGGED ELECTION: dummy variable coded 1 for a lagged legislative election and 0otherwise; YEARS UNDER: sum of past years under agreements for a country; NUMBER UNDER: number of other countries currently under IMF agreements; BOP:overall balance of trade in constant 1987 US$; REGIME: Dummy variable coded 1 for dictatorships and 0 for democracies.All variables are lagged 1 year.


ŽDEFICIT which is measured as budget surplus, following the World Bank conven-.tion has a significant effect on a government’s decision to enter. Governments

with high deficits are more likely to turn to the Fund.To capture rejection costs, the sensitivity of a country to the decisions of

creditors and investors, we include DEBT SERVICE, the annual debt service as aproportion of GDP, and INVESTMENT, real gross domestic investment as a propor-tion of GDP. Rejection costs also matter. When DEBT SERVICE is high, a govern-ment is more likely to enter an IMF arrangement, and governments typically turnto the IMF when INVESTMENT is low.

By bringing in the IMF, a government risks being accused of ‘‘selling out.’’Thus we include ‘‘sovereignty cost variables’’: YEARS UNDER, the number of pastyears a country has spent under IMF agreements; NUMBER UNDER, the number ofother countries around the world currently participating in an IMF agreement, andLAGGED ELECTION, a dummy variable coded 1 if the previous year had a legislativeelection and 0 otherwise. When many governments in a country’s past have beenunder IMF agreements and when many other countries around the world areparticipating, a government is more likely to enter into an IMF program. Govern-ments are more likely to enter agreements early in their electoral terms, hopingthat the stigma of signing an agreement will be forgiven or forgotten before thenext elections.

The effect of the above variables on the government’s decision to remain underagreements is surprising. None of the economic variables matter. The onlyvariable which has a significant effect on the government’s decision to continueparticipation in an IMF agreement is NUMBER UNDER. It has a positive effect. Agovernment is more likely to continue participation when others do so.6

For the IMF, we include three variables: BOP, NUMBER UNDER, and REGIME. BOP,a country’s overall balance of payments in constant 1987 US dollars, is included totest whether the Fund is concerned with fulfilling its mandate of promoting worldfinancial stability. The larger a balance of payments deficit in absolute terms, themore likely the Fund will provide a country with an arrangement.

NUMBER UNDER is a proxy for the IMF budget.7 The significant effect of thisvariable shows that the IMF seeks to sign more countries when few are participat-ing, when the budget constraint is less binding.

REGIME, a dummy variable coded 1 for dictatorships and 0 for democracies,indicates the IMF is more likely to sign with dictatorships. This may be becausethey are easier to negotiate with. Governments have private information about the

6 In this specification, no other variables we tested affected the government’s decisions to enter orŽleave: TERMS OF TRADE, BOP, CURRENT ACCOUNT, INFLATION, FOREIGN DIRECT INVESTMENT, EXPORTS as a

.capacity to import , EXCHANGE RATE, PER CAPITA INCOME, WORLD PER CAPITA INCOME GROWTH, OUTPUT

GROWTH, CAPITAL STOCK GROWTH, LABOR FORCE GROWTH, and a dummy variable for post-1982 years.7 There are, of course, better measures for the IMF budget, but the Fund keeps such information

secret.


policies domestic constituencies will accept and can use this information in IMFnegotiations to have their preferred policies imposed through the agreement. Sucha strategy is more effective in democracies, where opposition may have more

Ž .power. As Schelling 1960, 28 argues, ‘‘the ability of a democratic government toget itself tied by public opinion may be different from the ability of a totalitariangovernment to incur such a commitment.’’ Dictatorships are less constrained bypublic opinion and competitive elections, and thus they make easier negotiationpartners.

Regarding the decision of the Fund to retain countries under its programs, onlyBOP has a significant effect. This is not surprising. Once negotiations have beenconcluded, the IMF’s costs of negotiation have been met. All that matters iswhether balance of payments deficit continues to be large.8

Hence, governments typically seek IMF programs when they need a loan. Butsome governments also want conditions. Governments with high deficits enter intoIMF arrangements to have fiscal discipline imposed upon them. They are likely topursue this strategy when the costs of rejecting the IMF are high — when they aresensitive to decisions of creditors and investors — provided sovereignty costs arelow. The Fund cares about the absolute size of a balance of payments deficit.Beyond this, it seeks to maximize its budget by signing as many countries as itsresources allow.

4. The effect of programs on growth

The average rate of growth of total income among the 1,024 annual observa-tions for which we can specify the full selection model was 3.33%. Countries

Ž .participating in IMF programs grew at the rate of 2.04% Ns465 , whileŽ .countries not under the programs at 4.39% Ns559 : a difference of –2.35%.

The question is whether this difference is due to the conditions the countries facedor to program participation.

Growth observed under IMF programs was lower regardless of the conditionsunder which countries participated. In Table 2 we classify these conditions by thesize of the domestic deficit and of foreign reserves. Note that in 97 cases countriesremained under IMF programs even though they had decent reserves and lowdeficits: those countries grew by 1.02% slower than countries which enjoyed thesame conditions while not being subject to these programs. But even countries

8 Other variables that we tested were not significant in this specification: TERMS OF TRADE, CURRENT

ACCOUNT, INFLATION, DEBT, FOREIGN DIRECT INVESTMENT, EXPORTS, EXCHANGE RATE, PER CAPITA INCOME,WORLD ECONOMIC GROWTH, CAPITAL STOCK GROWTH, LABOR FORCE GROWTH, and the post-1982 dummyvariable.

()

A.P

rzeworski,J.R

.Vreeland

rJournalof

DeÕelopm

entEconom

ics62

2000385

–421

396

Table 2Ž .Growth according to observable conditions reserves and deficit

Reserves, Not under Examplesdeficit Growth Deficit Reserves Debt service N Typical Extreme

Good, good 5.22 y0.84 5.47 3.50 248 Columbia 1988, Iran 1977 Botswana 1987, Malta 1975Good, bad 4.65 y11.99 4.36 3.83 121 Rwanda 1988, Nepal 1984 C. Verde 1984, Iran 1981Bad, good 4.00 y1.96 1.19 3.76 102 Nigeria 1979, Dom. Rep. 1975 Cameroon 1987, Benin 1979Bad, bad 2.19 y12.51 1.09 5.29 88 Nicaragua 1982, Zambia 1989 Guyana 1986, Congo 1989Total 4.39 y5.30 3.76 3.90 559

Under

Good, good 4.20 y2.25 4.26 6.08 97 Philippines 1994, Ecuador 1970 Togo 1989, Uruguay 1980Good, bad 3.14 y9.35 3.34 5.46 89 Solomon 1982, Ghana 1979 Mexico 1988, India 1981Bad, good 1.95 y2.07 1.06 6.65 97 Mauritania 1988, Uganda 1988 Ivory Cst 1989, Gabon 1981Bad, bad 0.40 y11.87 0.89 7.54 182 Jordan 1989, Panama 1977 G. Bisau 1987, Guyana 1982Total 2.04 y7.34 2.10 6.65 465

‘‘Good’’ reserves: foreign reserves)2 times monthly imports.‘‘Bad’’ reserves: foreign reservesF2 times monthly imports.‘‘Good’’ deficit: government budget surplus)y5% of GDP.‘‘Bad’’ deficit: government budget surplusFy5% of GDP.


with low reserves and high deficits did better if they did not participate: theirgrowth was 1.79% faster. Thus, while countries facing bad conditions grewslower, participation in IMF programs lowered growth under all conditions.

Yet matching the observed cases can be misleading. For one, the distribution ofconditions is not the same for cases observed under and not IMF agreements andthose observed not under. For two, unobserved variables may be responsible forthe observed differences in performance. Hence, we proceed as follows. First, agrowth model is estimated separately for countries observed under and those notunder the agreements, with instruments to control for the effects of selection. Theparameters resulting from this estimation are not biased by selection. Second, thevector of independent variables characterizing each country at each time ismultiplied by the parameters characterizing the ‘‘under’’ and the ‘‘not under’’samples. This procedure generates values of growth expected in the two situationsindependently of selection, for each country during each year. Finally, theseselection-unbiased values are averaged over all countries, so that the differencebetween them is the net effect of IMF programs. A technical justification of these

Ž .procedures, due originally to Heckman 1976, 1988 , is included in Appendix B.The growth model used is a simple production function of the form YsŽ . a b Ž .AF K , L sAK L , to which we refer as ‘‘barebones’’ Table 3 . The model is

estimated in growth form separately for js1 if the country is observed underagreement and js0 otherwise,

˙ ˙ ˙ G IYrYsCONSTANT qa KrK qb LrL qu l qu l .Ž . Ž .j j j G j j I j j

G I Ž .where l and l called hazard rates are the instruments used to control for thei 1

effects of selection.When selection is taken into account, IMF programs still appear to reduce

growth. The average effect of program participation predicted by the full selectionmodel and the barebones growth model in the 1,024 sample is y1.53. If allcountries in this sample were under IMF programs every year, they would havegrown at the average rate of 2.00. If none of the countries ever had an agreement,they would have grown at the rate of 3.53. Hence, of the observed difference ofy2.35, y0.82 is due to the economic conditions and y1.53 to the effect of IMFprograms. The countries that had IMF agreements experienced less favorableconditions, but the effect of IMF programs was adverse to growth independentlyof these conditions. Indeed, this difference is very close for the cases actuallyobserved under and not under:

Observed as Hypothetically as

Under Not Program effect

Under 0.70 2.33 y1.63Not 3.08 4.52 y1.44All 2.00 3.53 y1.53


Table 3‘‘Barebones’’ growth regression by participation status

Under Not under

Selection-corrected Observed means Selection-corrected Observed means

CONSTANT y1.73 1.00 y0.13 1.00Standard error 0.44 0.38

K̇ rK 0.47 2.01 0.44 7.15Standard error 0.01 0.02

L̇rL 0.53 2.80 0.56 2.69Standard error 0.01 0.02

Gl 4.31 0.14 0.07 y0.73Standard error 1.48 0.23

Il 6.17 0.12 0.09 y0.82Standard error 2.23 0.29

˙Ž .E Yr Y 2.00 2.04 3.53 4.39Standard deviation 5.93 6.68 5.50 7.15

Observations 465 559Durbin–Watson 1.75 1.89

2Adjusted R 0.71 0.59Ž . Ž .F-test 0.00, ps 0.99 1.68, ps 0.19

Note: F-test is for the restriction a q b s1.

Experiments with other specifications and different samples, reported in AppendixC, generate the same qualitative results. Under different selection mechanisms,model specifications and samples, the effect ranges from y1.07 to y3.88.

The difference in the predicted growth rates is due almost exclusively to theconstants, indicating that the effects of unobserved variables operate through

Ž .channels other than the supply of capital and labor see Heckman, 1979, 155 .When growth equations are estimated without selection instruments, the constants

Ž .are almost the same for the observations ‘‘not under’’ y0.26 and ‘‘under’’Ž .y0.38 . Yet, when countries are observed under agreements, the coefficients onthe selection instruments, lG and lI , are large and significant. Hence, omitting1 1

these instruments strongly biases the constant. When corrected for selection, thedifference between the constants becomes large.

Hence, there are grounds to suspect that some unobserved factors influenceboth program participation and growth. One candidate is the ‘‘political will’’ ofcountries participating in IMF programs.9 Several students of the IMF have argued

9 ‘‘Political will’’ is merely one possible unobserved variable that affects both selection andperformance. The reputation of a government’s officials or advisors, its negotiation posture, or thepopulation’s trust in government may all lead a government to seek an IMF agreement and also affectgrowth.


Table 4Growth according to transition type

Observations of countries N YG KSG LFG Reserves Deficit

Never under a spell 82 5.973 9.619 2.356 5.895 y6.979Before spells 142 4.135 7.270 2.654 2.786 y4.792Before and between spells 346 4.301 7.107 2.711 3.219 y4.943Between spells 204 4.416 6.994 2.750 3.521 y5.048Between and after spells 335 4.118 6.493 2.793 3.653 y5.100After spells 131 3.654 5.714 2.860 3.858 y5.180During spells 465 2.042 2.010 2.802 2.101 y7.339

Žthat agreements may break down because of a lack of ‘‘political will’’ Edwards,.1999; Killick, 1995; Nelson, 1991 . Political will may make countries more likely

to continue participation and to grow. Moreover, not only are highly motivatedgovernments more likely to continue IMF programs but the IMF is also predis-posed to keep them. Yet while political will makes countries that participate growfaster than would participating countries not endowed with this will, or whateverelse the unobserved factors may be, the same countries would have grown fasternot participating in the IMF programs.

Participation in IMF programs calls for several measures that reduce growth inthe short run. But the standard argument in favor of these programs is that oncethe economy is stabilized, deficits are eliminated or reduced, and the balance ofpayments is improved, growth will resume. This argument is often only implicitand it is ideological: there are no good theoretical reasons to believe that abalanced budget and foreign account are sufficient for growth to occur. And it

Ž .appears to be false. Contrary to Conway 1994 , countries that emerge from IMF

Fig. 1. Observed growth before and after participation.

()

A.P

rzeworski,J.R

.Vreeland

rJournalof

DeÕelopm

entEconom

ics62

2000385

–421

400

Table 5The experience over time of countries that participated in IMF programs

Consecutive years Years since leaving the IMF program Cumulative growth over next 4 yearsŽ .participating weighted average0 1 2 3 4

Growth N Growth N Growth N Growth N Growth N Observed If participation had continued

Ž . Ž . Ž . Ž . Ž .First 0.92 82 0 0 0 0 2.09y1.63 y1.63

Ž . Ž . Ž . Ž . Ž .Second 1.90 82 5.81 14 3.96 16 3.71 23 5.48 24 4.70 2.42y1.62 y1.39 y1.52 y1.50 y1.42 y1.46 y1.63

Ž . Ž . Ž . Ž . Ž .Third 2.61 66 4.80 7 5.19 17 5.60 18 4.81 18 5.15 2.60y1.63 y1.61 y1.49 y1.42 y1.50 y1.49 y1.61

Ž . Ž . Ž . Ž . Ž .Fourth 2.50 49 4.26 11 5.99 14 4.45 15 3.51 14 4.57 2.40y1.59 y1.56 y1.44 y1.50 y1.47 y1.49 y1.64

Ž . Ž . Ž . Ž . Ž .Fifth 1.06 37 6.29 7 5.18 9 4.80 9 6.66 15 5.84 2.82y1.72 y1.50 y1.39 y1.43 y1.48 y1.45 y1.59

Ž . Ž . Ž . Ž . Ž .Sixth 3.62 29 5.51 4 5.73 4 6.02 9 4.78 10 5.44 2.38y1.62 y1.29 y1.37 y1.56 y1.46 y1.46 y1.59

Ž . Ž . Ž . Ž . Ž .Seventh 3.90 25 9.76 2 5.38 7 3.47 8 5.39 6 5.10 1.76y1.50 y1.04 y1.57 y1.50 y1.49 y1.48 y1.64

Ž . Ž . Ž . Ž . Ž .Eighth 1.47 25 6.53 5 3.85 5 2.22 3 10.42 5 6.15 1.93y1.66 y1.55 y1.47 y1.82 y0.85 y1.38 y1.65

Cumulative annual 2.04 5.60 5.03 4.53 5.43growth for 8 years y1.63 y1.47 y1.48 y1.50 y1.43Ž .weighted average

()

A.P

rzeworski,J.R

.Vreeland

rJournalof

DeÕelopm

entEconom

ics62

2000385

–421

401

Ž . Ž . Ž . Ž . Ž .Ninth 1.64 14 0 0 0 0 2.17y1.56 y1.70

Ž . Ž . Ž . Ž . Ž .Tenth 1.95 10 0 0 0 0 2.37y1.63 y1.71

Ž . Ž . Ž . Ž . Ž .Eleventh 2.56 9 0 0 0 0 2.73y1.70 y1.66

Ž . Ž . Ž . Ž . Ž .Twelfth 1.67 7 0 0 0 0y1.81

Ž . Ž . Ž . Ž . Ž .Thirteenth 2.75 4 0 0 0 0y1.71

Ž . Ž . Ž . Ž . Ž .Fourteenth 2.68 6 0 0 0 0y1.60

Ž . Ž . Ž . Ž . Ž .Fifteenth 4.25 5 0 0 0 0y1.47

Ž . Ž . Ž . Ž . Ž .Cumulative annual 2.04 465 4.90 61 4.06 43 3.16 38 6.76 32growth for 15 years y1.63 y1.49 y1.50 y1.55 y1.27Ž .weighted average

Top row for each consecutive year of participation: the observed rate of growth.Bottom row for each consecutive year of participation: the estimated effect of IMF programs on growth if the country participated.


programs do not grow faster than they did before having entered them or fasterthan countries that never entered.

Again, consider first the observed patterns. In Table 4, we classify theobservations according to their experience of participation. Note that these obser-vations are right-hand, but not left-hand, censored: they reflect all the priorexperience but end in 1990. Countries which never experienced IMF programsgrew the fastest. But the most relevant comparison is of growth before and afterprogram participation. And, whether or not we include cases in which a country

Ž .would turn to the IMF again those cases observed ‘‘between spells’’ programparticipation certainly does not accelerate growth. Indeed, for any combination ofthe ‘‘before’’ and ‘‘after’’ growth rates, the latter are somewhat lower.

Neither can we detect any trend once countries leave a program. Fig. 1 shows astylized picture of the experience with participation. The three horizontal linesshow the average growth rates of countries that never participated, countries that

Žwere currently not participating, and countries currently participating which we.stylized to last 5 years, about the average spell . Both ‘‘before’’ and ‘‘after’’

growth rates exhibit wide swings and no trend emerges once countries leaveprograms. Statistical analysis confirms these visual impressions.

Does it matter for future growth how long countries stay under the programs?Table 5 shows the observed rates of growth of all countries that ever entered andleft IMF programs and the hypothetical differences described above for eachconsecutive year during and after program participation. Since this table issomewhat complicated, it bears an explanation. Take the second year under. Therewere 82 episodes of participating at least 2 years, the mean rate of growth duringthe second year of participation was 1.90, and if these countries had not partici-pated, they would have grown 1.62 faster during this year. The next four columnsof this row show the rates of growth observed during the subsequent 4 years forcountries that left the program after 2 years and the hypothetical effect ifparticipation had continued.

Eye inspection of these numbers appears to indicate that countries that emergefrom longer spells grow somewhat faster. But their observations are few, so thatlittle can be inferred from these numbers. When a variable measuring the length ofthe past participation spell is added to the growth model for countries which arenot under, it is never significant. Hence, it seems that duration of programs doesnot affect growth once a country leaves them.

Perhaps the best way to conclude is to point to the last two columns of Table 5.They show the weighted average rate of growth during the subsequent 4 years ofcountries that left the program after different numbers of years with that ofcountries that continued to participate four more years. Their comparison showsthat for any number of years already under the program, one through eight, theobserved rates of growth during the next four years were lower if a countryremained under than if it left. The selection-corrected, hypothetical effect of theprogram is also always negative. Thus both the observed and the hypothetical


differences indicate that countries are better off, at least during the next 4 years,leaving the program rather than continuing to participate.

Hence, our conclusions are the following. When matched for exogenousconditions, participation in IMF programs reduces growth while a country remainsunder and has no salutary effect once a country leaves.

5. Conclusions

Our findings appear robust to the selection mechanism, specification of growthequations, and samples. Governments facing economic difficulties adopt IMFprograms either because they are desperate for foreign reserves, or because theywant to use the IMF as a foil to reduce budget deficits, or both. These programsreduce growth while countries remain under and do not return benefits that wouldcompensate the losses once they leave.

These conclusions are not intended as a blanket indictment of the IMF. Balanceof payment crises and exchange instability are facts of life, so the IMF has animportant role to play. The question is whether coping with these crises mustnecessarily reduce growth. This is a broader, and more fundamental, question thanthe effect of IMF programs. Our results indicate that countries that do not enterinto IMF programs grow faster than those that do even when both groups facehigh domestic deficits or foreign reserves crises. Hence, if soliciting IMF condi-tionality is just a way to impose domestically unpopular austerity policies, then theculprit is austerity policies per se, rather than the fact that they result from IMFagreements. ‘‘Political will’’ may just lead governments astray.

The question whether or not austerity policies are necessary to restore growth isbeyond the scope of this paper. We believe to have shown, however, that if growthis the primary objective then IMF programs are badly designed. Indeed, some of

Ž .the sharpest critiques of these programs are intramural. Tanzi 1989 argued thatŽIMF programs induce governments to save on public investment see also Tanzi

.and Davoodi, 1998 , with nefarious consequences for growth. Blejer and CheastyŽ .1989 pointed out that the high real interest rates induce good firms to shut downalong with bad ones. These effects seem to be avoidable. However, it appears thatprograms that stabilize inflation, reduce deficits, improve balance of payments,and at the same time do not hurt growth, are yet to be developed.

Since our analysis ends in 1990, one may wonder whether the IMF hasmodified the designs of its programs since then. We could find no evidence that it

Žhas. In the last official statement we discovered, dated October 1993 IMF SurÕey:.Supplement on the IMF 1993, 13 , the Fund reported that ‘‘The Executive Board

undertakes periodic reviews of conditionality and, on many occasions, it hasadjusted the policies and practices relating to the use of the IMF’s resources. In itsmost recent discussion of issues related to conditionality and program design, in


July 1991, the Executive Board affirmed that the current guidelines on conditional-ity, which the Board adopted in 1979, remain broadly appropriate.’’

Acknowledgements

We appreciate comments by Jess Benhabib, David Denoon, Martin Edwards,Jennifer Gandhi, Manuel Pastor, David Rowe, Alastair Smith, Elisabeth Wood,and the anonymous reviewer.

Appendix A. Dynamic bivariate probit model with partial observability

Let S denote participation status of country i at time t, S s‘‘Under’’ ifi t i t

country i is under agreement at time t, and S s‘‘Not Under’’ otherwise. Toi tŽ .simplify notation, write the probability that S sUnder as p U and the probabil-i t i t

Ž .ity that S sNot Under as p N .i t i t

Let p denote the probability that country i enters an agreement at time tN U, i tŽ .i.e., that it goes from ‘‘not under’’ at time ty1 to ‘‘under’’ at time t . Theprobability that country i does not sign at time t is p s1yp . Similarly,NN, i t N U, i t

p denotes the probability that country i stays under at time t. The probabilityU U, i t

that participation ends at time t is thus p s1yp . These transitionUN, i t U U, i t

probabilities vary across time and country, and thus are subscripted i,t todesignate individual country-years.

Ž .With lagged participation status S and these transition probabilities, onei, ty1

can describe participation as a first-order Markov process:10

p U NS NŽ . p pit i , ty1 i , ty1N U , i t U U , i ts p p Up N NSŽ . NN , i t UN , i t i , ty1i t i , ty1

where U is an indicator variable set equal to 1 if country i was under at timei, ty1

ty1, and 0 otherwise. Similarly, N is a dichotomous variable set to 1 ifi, ty1

country i was not under at time ty1 and 0 otherwise, N s1yU .i, ty1 i, ty1Ž .The probability of participating at time t, p U , is thusi t

p U NS sp N qp UŽ .i t i , ty1 N U , i t i , ty1 U U , i t i , ty1

sp q p yp U , A1Ž . Ž .N U , i t U U , i t N U , i t i , ty1

10 This characterization of a first-order Markov process follows Amemiya 1985, Chapter 11.


and the probability of not being under agreement at time t is the complement ofthe above to 1.

The decisions to enter and to continue IMF agreements are joint decisionsbetween a government and the Fund. For the government’s decision, we write thevalue of participation as the latent regression:

dG ) sgX xG qk

X xG U qÕG , A2Ž .i t i , ty1 i , ty1 i , ty1 i t

where the effects of the vector of variables determining the value of participation,xG , are captured by g if the country was not under an agreement at time ty1i, ty1

Ž .and by if the country was under. We write gqa as a shift for convenience, aswill be seen below. The effect of unobserved variables determining the value ofparticipation for the government is captured by ÕG, which we assume to bei t

normally distributed. The government wants to be under an IMF agreement if andonly if the value of participation is positive, dG ) )0.i t

We define the value of an agreement to the IMF with a similar equation,

d I ) smX x I qbX x I U qÕ I , A3Ž .i t i , ty1 i , ty1 i , ty1 i t

where the effect of the vector of variables x I is captured by the vector m ifi, ty1Ž . IU s0 and by mqb if U s1. Unobserved effects are captured by Õ ,i, ty1 i, ty1 i t

also assumed to be normally distributed.If we assume the unobserved variables that influence the government and the

IMF are independent, we can write the probability of entering an IMF agreementas

p sF gX xG F mX x I , A4Ž .Ž . Ž .N U , i t i , ty1 i , ty1

Ž .where F P is the CDF of the standard normal distribution.The probability of continued participation can be written as:

X XG Ip sF gqk x F mqh x . A5Ž . Ž . Ž .U U , i t i , ty1 i , ty1

Ž . Ž . Ž . Ž .Substituting Eqs. A4 and A5 into Eq. A1 yields Eq. A6 :

p U NS sF gX xG F mX x IŽ . Ž . Ž .i t i , ty1 i , ty1 i , ty1

X XG Iq F gqk x F mqh xŽ . Ž .� i , ty1 i , ty1

yF gX xG F mX x I U A6Ž .4Ž . Ž .i , ty1 i , ty1 i , ty1

Ž .The likelihood function can be formed by using the above Eq. A6 . However,Ž .after some tedious algebra, Eq. A6 can be rewritten in a simplified form:

p U NS sF gX xG qU k

X xGŽ . Ž .i t i , ty1 i , ty1 i , ty1 i , ty1

=F mX x I qU hX x I . A7Ž .Ž .i , ty1 i , ty1 i , ty1


The likelihood function is thus

UitX X X XG G I ILs F g x qU k x F m x qU h xŽ . Ž .Ł i , ty1 i , ty1 i , ty1 i , ty1 i , ty1 i , ty1i , t

X XG G= 1yF g x qU k xŽ .i , ty1 i , ty1 i , ty1

Ž .1yUX X i tI I=F m x qU h x , A8Ž .Ž .i , ty1 i , ty1 i , ty1

Ž .which can be estimated by a canned version of the Abowd and Farber 1982Ž .bivariate probit model with partial observability as is found in LIMDEP .

If the vectors X G and X I contain exactly the same set of variables, then themodel cannot be identified. To identify the parameters g , m, k , and h, it issufficient that there be at least one variable not in common between X G and X I.

Appendix B. Correcting for selection bias

Let Y stand for the rate of growth and X for some vector of observableŽ .variables. Assume that y is a function of x, with the population density f yNx .

Ž .Write the regression equations as suppressing the i, t subscripts

y sbX xqe , B1Ž .j j j

� 4where js1 if the country-year observation is ‘‘under’’ agreement, js 2,3,4otherwise.

The expected value of y for the observed sample is1

E y Nx , Y is observed sE y Nx ,dG ))0,d I ))0Ž . Ž .u 1 1

sbX xqE e NdG ) )0,d I ) )0 , B2Ž .Ž .1 1

Ž G ) I ) .where E e Nd )0, d )0 is the conditional expectation of e given that y1 1 1

is observed.On the other hand, the expected value for the population is

E y Nx sbX x . B3Ž . Ž .1 1

Hence, the value expected in the sample differs from the population value bythe conditional expectations of the error term given what we observe. If selection

Ž . G Iis not random, then e is correlated with either or both Õ or Õ , the effects of1 i t i t

unobserved variables on the respective decisions of the government and the IMFŽ Ž . Ž . .to participate from Eqs. A2 and A3 in Appendix A .


ŽThe expected value for the selected sample can be derived as follows see. G IPoirier, 1980, 216 and Greene, 1993 . Assuming that V and V are independent,Ž G ) I ) .we can write the expected value of E e Nd )0,d )0 as:1

E e NdG ) )0,d I ) )0 suX E ÕG NdG ) )0 qu

X E Õ I Nd I ) )0 ,Ž . Ž .Ž .1 1G 1 I

B4Ž .where

X Gf gqk xŽ . i , ty1G G ) GE Õ Nd )0 sl sŽ . X1 GF gqk xŽ . i , ty1

andX If mqh xŽ . i , ty1I I ) IE Õ Nd )0 sl s .Ž . X1 IF mqh xŽ . i , ty1

G I Ž .Hence the variables l and l are the hazard rates from Eq. A8 .1 1

We can thus write the expected value in the observed sample as

E y NdG ) )0,d I ) )0 sbX xqu

XlG qu

XlI . B5Ž .Ž .1 1 1G 1 1 I 1

Ž .Note that if B5 is estimated on the basis of the observed sample, the variableslG and lI are omitted from the specification. Hence, selection is a source of

Ž .omitted variable bias Heckman, 1979 .Ž .We can now also understand why controlling ‘‘matching’’ for the variables

that enter both into selection and outcome equations may in fact exacerbate, ratherŽ . Ž .than attenuate, the selection bias Achen, 1986 . Following Heckman 1988 ,

distinguish first between selection on observables and on unobservables. SelectionŽ j. Ž � 4.on observables occurs when the expected covariance E e Õ /0 jg G, I , but1

once the observed variables X j are controlled the covariance vanishes, so thatŽ j j. Ž j j.E e Õ NX s0. Selection is on unobservables when E e Õ NX /0, so that1 1

controlling the factors observed by the investigator does not remove the covari-ance between the errors in the outcome and the selection equations. Now, note that

Ž j. Ž .the regression coefficient u scov e ,Õ rvar e . If selection is on unobserv-1 1 1

ables, controlling for some variable x in the outcome equation may reduce theerror variance e without equally reducing the covariance e and Õ j. Hence, the1 1

coefficient on the omitted variable will be larger and the bias will be exacerbated.Note that correcting for selection bias for country-years observed ‘‘not under’’

Ž .is not straightforward. There are a total of four states of the world. In state 1 ,where country-years observed under an IMF agreement, correction is straightfor-ward:

Ž .1. E y Nx, observe an agreement s1

E y Nx , dG ) )0, d I ) )0 sbX xqE e NdG ) )0, d I ) )0 ,Ž . Ž .1 1 1

where both the government and the IMF want the agreement.


For country-years observed not under, however, there are three possible statesof the world:Ž . Ž G ) I ) . X Ž2. E y Nx, observe no agreement sE y Nx, d -0, d -0 sb xqE e N2 2 2 2G ) I ) .d -0, d -0 ,Ž . Ž G ) I ) . X Ž3. E y Nx, observe no agreement sE y Nx, d -0, d )0 sb xqE e N3 3 3 3G ) I ) .d -0, d )0 ,Ž . Ž G ) I ) . X Ž4. E y Nx, observe no agreement sE y Nx, d )0, d -0 sb xqE e N4 4 4 4G ) I ) .d )0, d -0 .

Ž . Ž .In 2 , neither the IMF nor the government wants to be under. In 3 , the IMFŽ .wants an agreement, but the government does not want to be under. In 4 , the

government wants the agreement, and the IMF does not. The problem is that sincewe do not observe the actors’ individual decisions, we do not know how to break

Ž . Ž .up the not under observations respectively between states 2 through 4 . Weknow that

yf gX xGŽ .i , ty1G G ) GE Õ Nd -0 sl sŽ . X0 G1yF g xŽ .i , ty1

and

yf mX x IŽ .i , ty1I I ) IE Õ Nd -0 sl s ,Ž . X0 I1yF m xŽ .i , ty1

but we do not observe when to use which pair of instruments when an observationis ‘‘not under’’:

lG ,lI , lG ,lI , or lG ,lI .Ž . Ž . Ž .0 0 0 1 1 0

We proceed by trying three different assumptions, assuming that all observationsŽ . Ž .belong to state 2 , all observations belong to state 3 , or all observations belong

Ž .to state 4 :

E y Nds1 sbX xqu

XlG qu

XlI ,Ž .1 1 G1 1 I1 1

E y Nds0 sbX xqu

XlG qu

XlI ,Ž .2 2 G2 0 I 2 0

E y Nds0 sbX xqu

XlG qu

XlI ,Ž .3 3 G3 0 I 3 1

E y Nds0 sbX xqu

XlG qu

XlI .Ž .4 4 G4 1 I4 0

This produces four sets of b ’s with which to estimate growth according to fourpossible states of the world. The expected average rates of growth are


Ž . Ž .1 Growth ‘‘under’’ both want agreement 2.00Ž .2 Growth ‘‘not under’’ because both 3.53Ž .3 Growth ‘‘not under’’ because of government 3.66Ž .4 Growth ‘‘not under’’ because of IMF 3.19

The effect of the IMF ranges from y1.19 to y1.66. In the main body of theŽ .paper, we report the effect of IMF programs as the difference between state 1

Ž .and state 2 .

Appendix C. Tests for robustness

Our robustness tests are best presented as a table, in which the entries are theselection-corrected differences between average growth expected under and notunder agreements:

Growth model Selection modelrsample

Full Stripped

Exogenous Endogenous Post-1970 Large

Ns1024 Ns2607 Ns3991

Barebones y1.53 y1.33 y1.33 y1.95˙With endogenous KrK y3.88 y3.68 y2.67 y3.05

Ž .Panel one-way fixed y2.29 y1.96 y1.07 y1.29

The ‘‘Full’’ selection model is the one presented in Table 1 in the main body ofthe text. Since several observations necessary to estimate this model are notavailable, we also estimate a ‘‘Stripped’’ selection model, which includes onlythose variables that are available for the entire sample. RESERVES, DEBT SERVICE

and BOP are not available and instead of DEFICIT we use government consumptionas a proportion of GDP. Other variables are the same as in the full selectionmodel.

Selection is ‘‘exogenous’’ if it is not driven by the expected consequences:hence, the predetermined variables in the growth equation do not enter into thereduced form of the selection equation. If the expected consequences do enter intothe structural form of the model, selection is ‘‘endogenous,’’ and the right-handside variables of the growth equation enter into the reduced form of the selectionequation. Obviously, some variables must appear in the selection but not in thegrowth equation for the model to be identified.

Samples differ according to the availability of data and periods. No dataconcerning the basic selection variables are available for the pre-1970 period andmore than one half observations are missing for the post-1970 period. Hence, the‘‘Full’’ selection model is estimated for the smallest, only post-1970, number of

Ž .observations Ns1024 . To test whether our results depend on period and dataavailability, a ‘‘Stripped’’ selection model combined with the ‘‘Barebones’’


Ž .growth model is estimated for the same period Ns2607 and for the entireŽ .1951–1990 period Ns3991 . This is particularly important because our sample

is strongly biased by the availability of data: the 466 observations of countriesunder agreements which enter into our ‘‘Full’’ selection sample had an averagerate of growth of 2.05, while the 614 observations of countries under for whichother data are not available had an average rate of 4.11.

Ž 2 .Statistical tests both F-test and x indicate for all samples and modelspecifications that the ‘‘not under’’ sample is not heterogeneous but that the‘‘under’’ sample exhibits fixed, country, effects. In the ‘‘Panel’’ line of the table,we follow these tests, that is, we use OLS for the ‘‘not under’’ observations andone-way fixed effects panel for the ‘‘under.’’ If we were to calculate the averageeffect if both groups were estimated by panels, the result would be y3.29 for the1024 sample.

To test the robustness of the specification of the growth model, we experi-mented by adding to the basic model the balance of payments as a proportion ofGDP, reserves, deficit, debt service, terms of trade, and the rate of growth ofeducation of an average member of the labor force, per capita income, the averagerate of growth in the world during the year, the percentage difference in output perworker between a given country and the maximum product per worker during thisyear, total number of years a country was under agreements in the past and, forcountries not currently under, the duration of their last stint under IMF. None ofthese variables is significant when the appropriate estimators are used.

Finally, note that the procedure in which we match observations for the valuesof the predetermined variables assumes that they are exogenous with regard to thestates under which these observations are made. If participation affects the valuesof these variables, then the counterfactual ‘‘had the countries been observed underthe same conditions’’ is no longer valid. Having repeated the same procedure withregard to the predetermined variables, we learned that the rate of growth of capital

˙stock, KrK , is lower when countries are under an agreement. We cannot tell,however, whether this is an effect of programs or of the sample for which otherrelevant variables are available. Again, our sample is strongly biased against

˙countries with higher KrK : the mean rate of growth of capital stock for all the‘‘under’’ observations in the large, 1951–1990, sample is 7.15 for the subset forwhich other data are missing and only 2.01 for the observations for which otherinformation is available. Hence, we do not know if the rate of growth of capital

˙stock can be treated as exogenous. The lines ‘‘with endogenous KrK ’’ are basedon the assumption that participation affects this variable and the expected rates of

˙growth are calculated by taking KrK at its mean for each participation status.When the rate of growth of capital stock is considered to be an endogenous effectof IMF programs, the effect of these programs appears much larger.

Hence, if anything, the results reported in the text are conservative. Indeed,what is striking is that IMF programs have a negative effect on growth even if weassume investment to be exogenous with regard to participation.


Appendix D. Country-years in sample

D.1. 4126 ObserÕation-sample: 135 countries 1951–1990

Country Years in sample Spell of agreements

Start End

Algeria 1962 1990 1989 1990Angola 1975 1989 Never underBenin 1960 1990 1989 1990Botswana 1966 1989 Never underBurkina Faso 1960 1990 Never underBurundi 1962 1990 1965 1971

1976 19771986 1989

Cameroon 1961 1990 1988 1990Cape Verde Island 1975 1990 Never underCentral African Republic 1961 1990 1980 1981

1983 1990Chad 1961 1990 1987 1990Comoro Island 1975 1990 Never underCongo 1961 1990 1967 1968

1977 19771979 19801986 19881990 1990

Djibouti 1977 1987 Never underEgypt 1951 1990 1977 1981

1987 1988Ethiopia 1951 1986 1981 1982Gabon 1961 1990 1978 1982

1986 1990Gambia 1965 1990 1977 1980

1982 1990Ghana 1957 1990 1966 1970

1979 19801983 19851987 1990

Guinea 1960 1990 1982 19831986 1990

Guinea-Bisau 1974 1990 1987 1990Ivory Coast 1961 1990 1981 1990Kenya 1963 1990 1975 1986

1988 1990


Lesotho 1966 1990 1988 1990Liberia 1961 1986 1963 1977

1979 1986Madagascar 1961 1990 1977 1978

1980 1990Malawi 1964 1990 1979 1986

1988 1990Mali 1961 1990 1964 1965

1967 19721982 1990

Mauritania 1961 1990 1977 19781980 19821985 1990

Mauritius 1968 1990 1979 1986Morocco 1956 1990 1959 1960

1965 19721980 1990

Mozambique 1975 1990 1987 1990Niger 1961 1989 1983 1989Nigeria 1960 1990 1987 1990Rwanda 1962 1990 1966 1970

1979 1980Senegal 1961 1990 1979 1990Seychelle 1976 1990 Never underSierra Leone 1962 1990 1966 1967

1969 19701977 19821984 1989

Somalia 1961 1989 1964 19711980 1989

South Africa 1951 1990 1958 19591961 19621976 19771982 1983

Sudan 1971 1990 1972 19751979 1985

Swaziland 1968 1989 Never underTanzania 1961 1988 1975 1976

1980 19821986 1988

Togo 1961 1990 1979 1990Tunisea 1961 1990 1964 1970

1986 1990


Uganda 1962 1990 1971 19721980 19841987 1990

Zaire 1960 1989 1976 1989Zambia 1964 1990 1973 1974

1976 1987Zimbabwe 1965 1990 1981 1984Bahamas 1978 1987 Never underBarbados 1966 1989 1982 1984Belize 1981 1990 1984 1986Canada 1951 1990 Never underCosta Rica 1951 1990 1980 1990Dominican Republic 1951 1990 1964 1965

1983 1986El Salvador 1951 1990 1958 1973

1980 19831990 1990

Grenada 1985 1990 Never underGuatemala 1951 1990 1960 1962

1966 19731981 19841988 1990

Haiti 1961 1989 1961 19671970 1989

Honduras 1951 1990 1957 19661968 19731979 19831990 1990

Jamaica 1962 1990 1963 19641973 19741977 1990

Mexico 1951 1990 1954 19551959 19591961 19621977 19791983 1990

Nicaragua 1951 1990 1956 19611963 19651968 19731979 1979

Panama 1951 1990 1965 19661968 1987

Trinidad and Tobago 1962 1990 1989 1990


USA 1951 1990 1963 1965Argentina 1951 1990 1958 1963

1967 19691976 19771983 1990

Bolivia 1951 1990 1956 19701973 19741980 19811986 1990

Brazil 1951 1990 1958 19591961 19621965 19731983 19861988 1990

Chile 1951 1990 1956 19591961 19701974 19761983 1990

Columbia 1951 1990 1957 1974Ecuador 1951 1990 1961 1967

1969 19731983 1990

Guyana 1966 1990 1967 19821990 1990

Paraguay 1951 1990 1957 1969Peru 1951 1990 1954 1971

1977 19801982 1985

Suriname 1975 1989 Never underUruguay 1951 1990 1961 1963

1966 19731975 19871990 1990

Venezuela 1951 1990 1960 19611989 1990

Bangladesh 1971 1990 1974 19761979 19831985 1990

China 1961 1990 1981 19811986 1987

India 1951 1990 1957 19581962 19661981 1984


Indonesia 1961 1990 1961 19641968 1974

Iran 1956 1990 1956 19561960 1962

Iraq 1954 1987 Never underIsrael 1954 1990 1974 1977Japan 1952 1990 1962 1965Jordan 1955 1990 1989 1990South Korea 1954 1990 1965 1977

1980 1987Laos 1985 1990 1989 1990Malaysia 1957 1990 Never underMongolia 1985 1990 Never underMyanmar 1951 1989 1969 1970

1973 19751977 19791981 1982

Nepal 1961 1986 1976 19771985 1986

Pakistan 1951 1990 1958 19591965 19661968 19691972 19751977 19781980 19831988 1990

Philippines 1951 1990 1962 19651973 19811983 1990

Singapore 1965 1990 Never underSri Lanka 1951 1990 1965 1972

1974 19751977 19811983 19841988 1990

Syria 1961 1990 1962 19621964 1964

Taiwan 1952 1990 Never underThailand 1951 1990 1978 1979

1981 19831985 1986

Yemen Arab Republic 1970 1989 Never underAustria 1951 1990 Never under


Belgium 1951 1990 1952 1957Bulgaria 1981 1990 Never underCzechoslovakia 1961 1990 Never underDenmark 1951 1990 Never underFinland 1951 1990 1953 1953

1967 19681975 1976

France 1951 1990 1956 19591969 1970

Germany 1951 1990 Never underEast Germany 1971 1988 Never underGreece 1951 1990 Never underHungary 1971 1990 1982 1985

1988 1990Iceland 1951 1990 1960 1963Ireland 1951 1990 Never underItaly 1951 1990 1974 1975

1977 1978Luxembourg 1951 1990 Never underMalta 1964 1989 Never underNetherlands 1951 1990 1957 1958Norway 1951 1990 Never underPoland 1971 1990 1990 1990Portugal 1951 1990 1977 1979

1983 1985Romania 1961 1989 1975 1978

1981 1984Spain 1951 1990 1959 1961

1978 1979Sweden 1951 1990 Never underSwitzerland 1951 1990 Never underTurkey 1951 1990 1961 1971

1978 1985UK 1951 1990 1956 1959

1961 19651967 19701975 1979

USSR 1961 1989 Never underYugoslavia 1961 1990 1961 1961

1965 19671971 19861988 1990


Australia 1951 1990 1961 1961Fiji 1970 1990 1974 1975New Zealand 1951 1990 1967 1968Papua New Guinea 1975 1990 1990 1990Solomon Islands 1981 1988 1981 1984Vanuatu 1984 1990 Never underWestern Samoa 1980 1990 1980 1980

1983 1985

D.2. 1024 ObserÕation-sample: 79 countries 1971–1990

Country Years in sample Agreements

Start End

Benin 1978 1989 1989 1989Botswana 1977 1987 Never underBurkina Faso 1974 1989 Never underBurundi 1986 1989 1986 1989Cameroon 1976 1981 1988 1990

1986 1990Cape Verde 1979 1989 Never underCentral African Republic 1981 1988 1981 1981

1983 1988Chad 1973 1977 1987 1990

1984 1990Congo 1972 1972 1986 1988

1981 1989Egypt, Arab Republic 1976 1990 1977 1981

1987 1988Ethiopia 1984 1986 Never underGabon 1974 1977 1980 1982

1980 1988 1986 19901990 1990

Gambia, The 1974 1990 1977 19801982 1990

Ghana 1973 1983 1979 19801983 1983

Guinea-Bissau 1987 1990 1987 1990Cote d’Ivoire 1980 1990 1981 1990Kenya 1973 1990 1975 1986

1988 1990


Lesotho 1983 1990 1988 1990Liberia 1976 1986 1976 1977

1979 1986Madagascar 1973 1975 1981 1990

1981 1990Malawi 1972 1989 1979 1986

1988 1989Mali 1978 1989 1982 1989Mauritania 1976 1980 1977 1978

1983 1989 1980 19801985 1989

Mauritius 1977 1990 1979 1986Mozambique 1985 1989 1987 1989Niger 1977 1987 1983 1987Nigeria 1973 1979 1987 1990

1983 1990Rwanda 1974 1990 1979 1980Senegal 1971 1974 1979 1980

1976 19761978 1980

Sierra Leone 1975 1990 1977 19821984 1989

Somalia 1973 1989 1980 1989Sudan 1973 1990 1973 1975

1979 1985Swaziland 1975 1989 Never underTanzania 1984 1988 1986 1988Togo 1978 1989 1979 1989Tunisia 1973 1990 1986 1990Uganda 1984 1989 1984 1984

1987 1989Zaire 1972 1989 1976 1989Zambia 1973 1990 1973 1974

1976 1987Zimbabwe 1978 1990 1981 1984Barbados 1973 1989 1982 1984Costa Rica 1973 1988 1980 1988Dominican Republic 1973 1990 1983 1986El Salvador 1971 1982 1971 1973

1980 1982Guatemala 1973 1986 1973 1973

1990 1990 1981 19841990 1990


Haiti 1981 1988 1981 1988Honduras 1973 1988 1973 1973

1979 1983Jamaica 1976 1988 1977 1988Mexico 1973 1990 1977 1979

1983 1990Nicaragua 1971 1984 1971 1973

1979 1979Panama 1974 1990 1974 1987Trinidad and Tobago 1977 1990 1989 1990Brazil 1979 1979 1983 1986

1981 1990 1988 1990Chile 1973 1983 1974 1976

1983 1983Colombia 1972 1990 1972 1974Ecuador 1974 1990 1983 1990Guyana 1971 1986 1971 1982Paraguay 1973 1990 Never underPeru 1971 1981 1971 1971

1977 1980Uruguay 1973 1983 1973 1973

1975 1983Venezuela 1971 1990 1989 1990Bangladesh 1974 1990 1974 1976

1979 19831985 1990

India 1975 1990 1981 1984Indonesia 1973 1983 1973 1974Iran, Islamic Republic 1975 1983 Never underJordan 1984 1990 1989 1990Malaysia 1973 1984 Never underNepal 1977 1986 1977 1977

1985 1986Pakistan 1974 1990 1974 1975

1977 19781980 19831988 1990

Philippines 1973 1990 1973 19811983 1990

Sri Lanka 1971 1972 1971 19751974 1990 1977 1981

1983 19841988 1990


Syrian Arab Republic 1986 1989 Never underThailand 1973 1987 1978 1979

1981 19831985 1986

Malta 1973 1979 Never under1981 1987

Turkey 1971 1987 1971 19711978 1985

Fiji 1971 1987 1974 1975Papua New Guinea 1976 1987 Never underSolomon Islands 1982 1987 1982 1984Vanuatu 1985 1987 Never under

References

Abowd, J.M., Farber, H.S., 1982. Job queues and the union status of workers. Industrial and LaborRelations Review 35, 354–367.

Achen, C.H., 1986. The Statistical Analysis of Quasi-Experiments. University of California Press,Berkeley.

Alvarez, M., Cheibub, J.A., Limongi, F., Przeworski, A., 1996. Classifying political regimes for theACLP data set. Studies in International Comparative Development 31, 3–36.

Amemiya, T., 1985. Advanced Econometrics. Harvard University Press, Cambridge, MA.Bird, G., 1996. Borrowing from the IMF: the policy implications of recent empirical research. World

Development 24, 1753–1760.Bjork, J., 1995. The uses of conditionality. East European Quarterly 29, 89–124.Blejer, M.I., Cheasty, A., 1989. Fiscal policy and mobilization of savings for growth. In: Blejer, M.I.,

Ž .Chu, K.-Y. Eds. , Fiscal Policy, Stabilization, and Growth in Developing Countries. IMF,Washington DC.

Camdessus, M., 1990. Statement before the United Nations Economic and Social Council in Geneva,Ž .July 11. IMF Survey 19 15 , 235.

Financial support for member countries complements economic policy changes. In: Cheney, D.M.Ž .Ed. , IMF Survey: Supplement on the IMF. 1993. pp. 13–15.

Connors, T.A., 1979. The Apparent Effects of Recent IMF Stabilization Programs. InternationalFinance Discussion Papers 135. Board of Governors of the Federal Reserve System.

de Vries, M., 1986. The IMF in a changing world: 1945–1985. IMF, Washington, DC.Dixit, A.K., 1996. The Making of Economic Policy: A Transaction-Cost Politics Perspective. MIT

Press, Cambridge.Edwards, M.S., 1999. Things Fall Apart: Why Do IMF Agreements Break Down? Paper Prepared for

the Duke University Center for International Studies Conference International Institutions: GlobalProcessesrDomestic Consequences, Durham April 9–11.

Edwards, S., Santaella, J.A., 1993. Devaluation controversies in the developing countries: lessons fromŽ .the Bretton Woods Era. In: Bordo, M.D., Eichengreen, B. Eds. , A Retrospective on the Bretton

Woods System. University of Chicago, Chicago, pp. 405–455.Fischer, S., 1999. On the Need for an International Lender of Last Resort. Paper prepared for delivery

at the joint luncheon of the American Economic Association and the American Finance Associa-tion, New York, January 3 1999. Available at www.imf.org.

Goldstein, M., Montiel, P.J., 1986. Evaluating fund stabilization programs with multicountry data:some methodological pitfalls. IMF Staff Papers 33, 304–344.


Greene, W.H., 1993. Econometric Analysis. 2nd edn. Prentice Hall, Englewood Cliffs, NJ.Gylfason, T., 1987. Credit Policy and Economic Activity in Developing Countries with IMF Stabiliza-

tion Programs. Studies in International Finance 60, Princeton.Heckman, J.J., 1976. The common structure of statistical models of truncation, sample selection, and

limited dependent variables and a simple estimator for such models. Annals of Economic andSocial Measurement 5, 475–492.

Heckman, J.J., 1979. Sample selection bias as a specification error. Econometrica 47, 153–161.Heckman, J.J., 1988. The microeconomic evaluation of social programs and economic institutions. In:

Chung-Hua Series of Lectures by Invited Eminent Economists, No. 14. The Institute of Economics,Academia Sinica, Taipei.

Khan, M.S., 1990. The macroeconomic effects of fund-supported adjustment programs. IMF StaffŽ .Papers 37 2 , 195–234.

Killick, T., 1995. IMF Programs in Developing Countries: Design and Impact. Routledge, London.Knight, M., Santaella, J.A., 1997. Economic determinants of fund financial arrangements. Journal of

Development Economics 54, 405–436.Ž .Nelson, J. Ed. , Economic Crisis and Policy Choice Princeton University Press, Princeton.

Pastor, M., 1987a. The International Monetary Fund and Latin America: Economic Stabilization andClass Conflict. Westview Press, Boulder.

Pastor, M., 1987b. The effects of IMF programs in the third world: debate and evidence from LatinAmerica. World Development 15, 365–391.

Poirier, D.J., 1980. Partial observability in bivariate probit models. Journal of Econometrics 12,209–217.

Polak, J.J., 1991. The Changing Nature of IMF Conditionality. Princeton University, InternationalFinance Section, Department of Economics, Princeton.

Przeworski, A., Limongi, F., 1996. Selection, Counterfactuals and Comparisons. Working Paper,March 11.

Putnam, R.D., 1988. Diplomacy and domestic politics: the logic of two-level games. InternationalOrganization 42, 427–460.

Reichmann, T.M., Stillson, R.T., 1978. Experience with programs of balance of payments adjustment:stand-by arrangements in the highest tranches, 1963–72. IMF Staff Papers 25, 292–310.

Remmer, K.L., 1986. The politics of economic stabilization, IMF standby programs in Latin America,1954–1984. Comparative Politics 19, 1–24.

Ž .Santaella, J.A., 1996. Stylized facts before IMF-supported adjustment. IMF Staff Papers 43 3 ,502–544.

Schelling, T.C., 1960. The Strategy of Conflict. Harvard University Press, Cambridge.Spaventa, L., 1983. Two letters of intent: external crises and stabilization policy, Italy, 1973–77, In:

Ž .Williamson, J. Ed. , IMF Conditionality Institute for International Economics, Washington, DC.Stein, H., 1992. Economic policy and the IMF in Tanzania: conditionality, conflict, and convergence.

Ž .In: Campbell, H., Stein, H. Eds. , Tanzania and the IMF: The Dynamics of Liberalization.Westview, Boulder, pp. 59–83.

Ž .Tanzi, V., 1989. Fiscal policy, stabilization and growth. In: Blejer, M.I., Chu, K.-Y. Eds. , FiscalPolicy, Stabilization, and Growth in Developing Countries IMF, Washington, DC.

Tanzi, V., Davoodi, H., 1998. Roads to Nowhere: How Corruption in Public Investment Hurts Growth.IMF, Washington, DC.

Ž .Taylor, L. Ed. , The Rocky Road to Reform: Adjustment, Income Distribution, and Growth in theDeveloping World The MIT Press, Cambridge.

Vaubel, R., 1986. A public choice approach to international organization. Public 51.Ž .Vaubel, R., 1991. Problems at the IMF. Swiss Review of World Affairs 40 12 , 20–22.

Vreeland, J., 1999. The IMF: lender of last resort or scapegoat? Prepared for the International StudiesAssociation Conference, Omni Shoreham Hotel, Washington, DC, February 16–21.

The Effect of IMF Programs on Economic

Documents

Transcript of The Effect of IMF Programs on Economic