IMF Conditionality and the Intertemporal Allocation of Resources

• Date: 01 June 2014
• DOI: http://doi.org/10.1111/irfi.12029
• Published date: 01 June 2014
• Author: Hassan Naqvi

IMF Conditionality and the

Intertemporal Allocation

of Resources

HASSAN NAQVI

Graduate School of Business, Sungkyunkwan University, Seoul, Korea

ABSTRACT

This article analyzes the impact of IMF (International Monetary Fund) con-

ditionality on the intertemporal allocation of resources in an emerging

market economy. The study identifies a principal-agent problem between

the government of the emerging market and its citizens and shows that

conditionality has the potential to mitigate the resulting misallocation of

resources. Nevertheless, the analysis indicates that if IMF lending were influ-

enced by geopolitical motives then the suboptimal allocation of resources

would remain notwithstanding IMF conditionality.

I. INTRODUCTION

IMF (International Monetary Fund) loans are usually provided under an

arrangement that stipulates the conditions that the borrowing country must

agree to meet in order to gain access to the loan. These conditions that take

the form of specific economic policies which the borrowing country is

obliged to implement are usually referred to as ‘IMF conditionalities’ or simply

‘conditionalities’.

There has been a widespread debate on whether or not the IMF should

impose conditionalities on the borrowing countries. However, as argued by

Drazen (2002), any discussion of conditionality is meaningless unless some

source of heterogeneity of interests is identified among the different stakehold-

ers. In the absence of any heterogeneity of interests, there would be no basis for

conditionalities and they would not be required. This is because if the interests

of all relevant parties were aligned, then any conditionalities imposed on a

country will not be binding and hence they would not be required.

Thus, if the IMF cares about the welfare of the borrowing country, and further

if there are no frictions between any parties within the country, then the

country itself will implement policies that are in its best interests regardless of

whether or not conditionalities are imposed.

Any discussion of conditionality which does not consider heterogeneity of

interests implicitly assumes that the IMF has some sort of an informational

bs_bs_banner

International Review of Finance, 14:2, 2014: pp. 203–235

DOI: 10.1111/irfi.12029

© 2014 International Review of Finance Ltd. 2014

advantage over the government or the policy makers of the country over which

conditionalities are imposed. Hence, the assumption implies that in the absence

of conditionalities, the policy makers may choose bad policies merely because

they are not aware of any better alternatives. This assumption is rather ad hoc

and if anything we would expect the government to have more information

about its country vis-à-vis the IMF. Thus, it would be implausible to justify

conditionality on the basis of any IMF informational advantage.

We thus believe that any discussion of conditionality is futile without first

identifying the source of heterogeneity of interests. Broadly speaking, there

might exist ‘external heterogeneity of interests’, say between the IMF and the

borrowing country, because of differing objectives, or ‘internal heterogeneity of

interests’ within a country, say between the government and its citizens.1

Finally, there might exist a combination of external and internal heterogeneity

of interests.

So far, as internal heterogeneity of interests is concerned, IMF conditionality

can be beneficial as long as it takes into account the divergence of interests

within a country and imposes the welfare improving policies. However, as far as

external heterogeneity of interests is related, conditionality will not be effective

if the IMF is not necessarily concerned with the welfare of the country.

Drazen (2002) focuses on the heterogeneity of interests between the gov-

ernment and domestic interest groups. He studies the situation where the

government wants to introduce positive reforms but is unable to do so

because of domestic opposition from pressure groups. He shows that condi-

tional lending in such a situation will have a positive outcome as long as the

domestic interest groups prefer IMF lending with reform to no lending with

no reforms.2

In our paper, we study a principal-agent problem between the government

and the citizens of an emerging market economy (EME). In contrast to the

situation that Drazen (2002) studies, the government in our model (the agent)

makes decisions that are in its own self-interests rather than that of the citizens

(the principal). More specifically, the government misallocates resources over

time and gives too much weight to the current period since it is not sure

whether or not it will remain in power in the next period. We show that IMF

conditionality can be useful in such a circumstance as it can help resolve the

conflict of interests between the government and the citizens.

We next extend the analysis to consider the situation where IMF lending is

influenced by geopolitical motives and it internalizes any political benefits that

an influential member like the U.S. may gain from the Fund’s lending to the

1 One example of external heterogeneity of interests is where the borrowers are not willing to

repay a loan even though they are able to do so. Asiedu and Villamil (2002) show that foreign

aid can serve as an enforcement mechanism by increasing the borrowers’ willingness to pay.

2 Mayer and Mourmouras (2008) present a political-economy model where independent special

interest groups oppose reforms by the government. They also show that conditional assistance

can reduce the distortions induced by the presence of special interest groups.

International Review of Finance

204 © 2014 International Review of Finance Ltd. 2014

EME. We argue that IMF lending should be based solely on economic funda-

mentals and should be independent of any political influences. Thus the IMF

incurs a reputational cost if it were publicly known that it took politics into

account when lending. Hence, in order to avoid the reputational cost, the IMF

makes a hidden transfer to the US rather than openly admitting that it took

politics into consideration when lending. To justify its politicized lending to the

EME, the IMF overstates the economic fundamentals of the EME which in turn

affects the potency of the conditionality that it imposes, since conditionalities

are in general a function of economic fundamentals.

As argued by Coate and Morris (2005), the key to understanding hidden

transfers is that the party making such transfers cares about the reputational loss

that it would have suffered had it made such transfers in a transparent way.

Thus, the IMF prefers to make a hidden transfer as it is concerned about its

reputation even though such a transfer is more costly since it hampers the

effectiveness of conditionality. Hence, we show that because of the presence of

external heterogeneity of interests, IMF conditionality is no longer justified and

the intertemporal misallocation of resources in the EME remains.

Our results are consistent with our earlier suggestion that conditionality can

be effective as long as there exists internal heterogeneity of interests, but

this conclusion may no longer be valid once external heterogeneity of interests

sets in.

A number of empirical studies find evidence that IMF forecasts of economic

fundamentals of emerging economies suffer from an upward bias. Beach et al.

(1999) find that IMF forecasts of Gross Domestic Product (GDP) and inflation

for developing regions tend to be overly optimistic. They argue that IMF fore-

casts are adjusted so as to support its lending activity. Similarly, Aldenhoff

(2007) finds an optimistic bias in IMF forecasts and he attributes this bias to

political factors. Timmermann (2007) also finds that World Economic Outlook

forecasts of GDP growth tend to exhibit systematic overprediction while

the inflation forecasts tend to exhibit underprediction. Dreher et al. (2008)

find evidence that the allies of IMF’s major stakeholders (in particular the

US) receive more optimistic forecasts for growth and inflation. Mussa and

Savastano (1999) document that the Fund-supported program for Mexico in

1995–1996 was based on optimistic economic assumptions. These findings are

consistent with our theoretical result that the IMF overstates the economic

fundamentals of the EME in order to avoid the reputational cost associated

with politically influenced lending.

The remainder of the paper is organized as follows. Section II models the EME

and analyzes the principal-agent problem between the EME government and its

citizens. Section III introduces a model of IMF lending and discusses how IMF

conditionality can restore the first-best allocation of resources. Section IV dis-

cusses the political influences in IMF lending and reconsiders the effectiveness

of conditionality when the IMF does not want to make it public that political

influences had a bearing on its lending. Finally, section V discusses and sum-

marizes our results.

IMF Conditionality

205© 2014 International Review of Finance Ltd. 2014