سیاستهای صندوق بین المللی پول چگونه می تواند مخاطرات اخلاقی کشور و حساب را برای اثرات جانبی از بین ببرد؟

Costly crisis prevention has positive external effects, which leads to free-riding of governments on each other's efforts. “Ordinary” IMF loans aggravate existing externalities, reinforcing the under-investment problem. We consider the reform proposals of the “Meltzer commission” in both loan and insurance models and show how the IMF can eliminate country moral. The efficiency-ensuring loan policy accounts for given externalities and involves effort-contingent discounts on interests or the extension of credit volume. Similar results hold for the insurance framework. Ex ante participation requires that smaller countries be “subsidized” by large ones, or that IMF policy consider distributional aspects in addition to efficiency.

Calls for fundamental reform of the International Monetary Fund (IMF) and the World Bank are common, yet what type of measures need to be taken appears to be a very controversial issue. Nobel prize winner and former chief economist of the World Bank Joseph Stiglitz focused his criticism on the “market ideology” of the Fund and claimed that the “Washington Consensus” was not considered as a means to the ultimate objective of reducing unemployment and poverty, but was regarded as the Holy Grail itself (Stiglitz, 2002). He accuses the IMF of having deepened, if not caused financial crises by forcing austerity on countries in recession, and by implementing liberalization and privatization before the necessary institutions, regulation, and supervision had a chance to build up and accompany the process. On the other hand, influential economists and critics of the IMF such as Jeffrey Sachs and Allan Meltzer have been proposing for long reforms in a different direction. Back in 1998, as part of a $18 billion funding package for the IMF, the US Congress established the International Financial Institution Advisory Commission, chaired by Allan Meltzer and therefore better known as the “Meltzer commission”. It concluded its work in March 2000, sharply criticizing the two Bretton Woods institutions and offering detailed proposals for far-reaching structural changes.1 Meltzer and the majority of the commission members argue that the programs often overlap, are inefficient and tend to distract resources from their primary targets. According to Meltzer, the inefficiency of the current system is partly due to the delays when a crisis occurs: the IMF is called for financial aid, but since it wants to make sure the money is not wasted, it conditions help on detailed structural reforms. This is usually inconvenient to the government concerned, so time-consuming negotiation takes place, while international investors withdraw their capital, reinforcing the crisis. A second criticism concerns the observation that IMF lending often ends up bound on a long-term basis, thereby allowing countries to delay reforms, whereas the Fund's statutory mission is to provide short-term liquidity when markets threaten to close, acting as a quasi lender of last resort. Therefore, Meltzer concludes, the IMF's main tasks could best be achieved if countries had to qualify in advance in order to receive aid without delay in the event of a crisis.2 In addition, the country moral hazard problem3 (anticipated bailouts dampen incentives to prevent crisis4) could be mitigated this way. Pre-qualification would serve as a seal of approval to private investors. One obvious problem with Meltzer's proposals is the credibility of not supporting “unqualified” countries struck by a crisis.5 As Feldstein (2002) points out, there is the further problem that a pre-qualified country would presumably have to be re-approved from time to time. If the country failed to be re-approved, that would be a very serious signal to the markets, indicating that financial aid is no longer available and that the IMF had concluded that the country's policies were such that it no longer met the standard. The approach proposed in this model partly overcomes these drawbacks. We show what the optimal IMF policy would look like if a pre-qualification commitment was possible. Departing from the ideal world, we argue that the chances for (partial) implementation of the derived policy are not as bad as one might think. The present model in way updates the traditional Keynes–White “coordination” model of the IMF (Keynes, 1942). In today's world of contagious emerging market crises, negative spill-overs come from real distortions rather than competitive devaluations. While simple balance of payments lending on identical terms could stop a devaluation race among healthy economies, such lending to crisis countries is not optimal. The key assumptions are that countries cannot borrow in times of crisis from the world capital market, that crises (or, in a positive sense, the deletion of crises) involve spill-overs across countries, and that effort (crisis prevention measures, i.e. sound macroeconomic policy and suitable regulation) is observable. It turns out that the optimal policy – which provides the proper incentives for sovereign member countries to apply the socially optimal level of crisis prevention measures in a globalized world – requires pre-qualification to attain efficiency and avoid the country moral hazard effect. This confirms in a formal way the Meltzer proposals and other contributions, such as Goldstein (1999) and Jeanne and Zettelmeyer (2001b).6 In the present model, we do not define inefficient investment in effort in itself as moral hazard but rather the aggravation of externalities that result from IMF intervention. This view of country moral hazard does not rely on subsidy elements in IMF transfers or political under-representation of the receiving country's taxpayers, as e.g. in Calomiris, 2000 and Mussa et al., 2000 and Mussa (2002), among others. It is also quite distinct from other major contributions which emphasize coordination failures in capital markets and thus the Fund' potential role as an International Lender of Last Resort ( Sachs, 1995, Jeanne and Wyplosz, 2001, Jeanne and Zettelmeyer, 2002 and Corsetti and Roubini, 2003), asymmetric information and incomplete contracts ( Marchesi and Thomas, 1999 and Tirole, 2002), or domestic policy failures, reflecting conflicts between the local government and special interest groups ( Drazen, 2002 and Mayer and Mourmouras, 2002). In their view of moral hazard, Morris and Shin (2003) find that the “conventional debtor moral hazard” effects (slackening debtor effort in anticipation of IMF assistance) may or may not predominate the potential “catalytic” effect of IMF intervention on the debtor country (which would otherwise default) and/or private investors (who would otherwise withdraw). In order to develop a purely externality-based framework, we chose to introduce welfare functions that are linear in utility (i.e. risk aversion and consumption smoothing are not part of the model). The public good character of IMF intervention on behalf of externalities has also been addressed in Masson and Mussa (1996). Clark and Haizhou (2001) elaborate on how the provision of liquidity by the Fund can help to reduce the cost that arises from financial contagion. As opposed to the mechanism in the present model, the authors introduce exogenous shocks to output and focus on negative spill-overs accruing from an increased (uniform) world interest rate in the aftermath of country defaults. Found “conditionality” (defined as political pressure leading to increased country output and “bought” with an interest rate below the market rate) then has a role in limiting the number of defaults. In another recent model of IMF conditionality and private sector involvement, Federico (2001) addresses pre-qualification using a principal-agent framework. He finds that the IMF can only effectively improve matters if the crisis is neither “too small” nor “too large”. In the first case, the IMF does not have sufficient leverage on the recipient country to impose reforms. In the latter, the Fund must compensate the country for not defaulting on third-party debt, which also prevents the imposition of conditionality. Federico (2001) models, however, a principal (the IMF) with fixed preferences: maximization of country effort and minimization of financial aid. In contrast, our model sees the Fund essentially acting in the interest of its member countries (maximizing global welfare). Our main result is that, in a loan contract, the optimal interest rate (or, alternatively, the optimal credit volume) is determined by the scope of externalities and the rate of time preference specific to the borrowing country. In order to make the country accept the credit terms, effort must be rewarded by reductions in the interest rate (or increases in the credit volume). The optimal loan policy differs much from what is common IMF practise. In the absence of sufficient sanctions that force loan repayments of sovereign debtors, loan contracts may become infeasible, and insurance contracts may remain the only instrument available. In fact, the case of Argentina in 2001/2002 could be interpreted as a case where the country ran into a crisis which was so severe that ex post solvency was unsure, even with respect to IMF lending. This is the reason why in some newer models of financial architecture, the First Best can only be achieved through an insurance fund rather than a lending-based IMF, see Jeanne (2002). We show that the optimal insurance policy is governed by the same economic mechanisms: efficiency can be ensured by deducting external damages from crisis payments. Again, effort must be rewarded. Ex ante participation, however, is yet another issue: the model suggests that in the optimal insurance system, small countries must be subsidized by large ones or via a refined IMF policy in order to make them participate. The remainder of this paper is organized as follows. Section 2 introduces a simple model with two states of nature (good and bad; the latter representing financial crisis), which allows us to illustrate each country's under-investment in crisis prevention if no extra incentives are given. In the third section, we show that the IMF can cause country moral hazard by magnifying existing externalities with “ordinary lending”. We proceed by asking how an (ideal) international organization like the IMF could mitigate or even eliminate this inefficiency. First, we consider a loan contract and derive the optimal loan policy functions that both eliminate moral hazard and take external effects into account. In a second step, we analyze the corresponding insurance contract that leads to the same efficiency results but evokes more subtle participation issues. These will be subject of a subsequent section, in which the case of small countries is addressed, leaving large and intermediate countries to the Appendix B. Section 5 summarizes and discusses the results.

In this paper, we looked at coordination problems that arise between sovereign countries when there are mutual spill-overs of costly domestic policy (such as crisis prevention measures), creating incentives to free-ride on each other. We adopted the view that policy itself is almost costlessly observable. We analyzed the proposals of the Meltzer commission which suggested that IMF financial aid should be conditioned on the past economic policies of a country. The paper introduced a loan and an insurance model with similar features. It showed that ordinary loan contracts aggravate country moral hazard, i.e. induce IMF member countries to relax crisis prevention. But institutional coordination can do much better with appropriate policy; it can even guarantee a socially First Best of crisis prevention. Pre-qualification is crucial to attaining efficient investment in crisis prevention. However, it does not seem to be necessary to exclude bad-policy countries from IMF lending as Meltzer proposed; it suffices to impose less favorable conditions. Moreover, the model made clear that looking at the absolute level of effort is not enough; the extent of externalities must be taken into consideration as well. Reductions in interest rates or an increase in crisis payments contingent on observed effort are necessary: if such rewards on crisis prevention are not granted, then the member countries have no reason to seek the IMF's assistance. A further reduction in the optimal interest rate is possible by rationing the credit volume. Interest rate and credit volume are therefore complementary tools to push the countries' crisis prevention into the desired direction. While for the loan model, ex ante participation is not a concern because each contract stands for its own, it not immediately clear why small countries should sign the insurance contract and pay up-front premiums. They could free-ride larger countries' crisis prevention measures and save their own expenses. Indeed, the model confirms that economically unimportant countries take the system as given and demand subsidies as a pre-condition to participation, while large countries recognize that their exit would leave the world without coordination, which may make them willing even to subsidize the smaller countries, provided that their economy benefits sufficiently from overall crisis prevention. An immediate policy implication of this finding is that large countries should not insist on a completely “fair” funding of the IMF system; they should accept their leading role which is associated with a larger financial burden. Furthermore, any form of protectionism on the large countries' part would be detrimental to the system because then the external effects of small and intermediate countries on the large ones would not be sufficient to support their participation. Intermediate countries act like small countries in that they also demand subsidies. But their decision may influence the feasibility of the system if their non-participation leaves the large countries worse off. This leads to the paradoxical result that participation in a coordinating mechanism needs to be coordinated if inefficiencies are to be avoided. In fact, an appropriate refinement of the derived IMF policy – given the necessary information on external effects and crisis prevention cost – can always ensure participation. Unfortunately, the fact that basically all countries of the world are currently members of the IMF cannot be taken as a sign that the optimal system is feasible: apparently, the derived IMF policy has little to do with its current setup. We recognize that motivates other than the internalization of externalities and the elimination of moral hazard play a role, such as consumption smoothing (which implies risk aversion), or political considerations. Yet if more crisis payment is desired, e.g. to smooth consumption, then we suggest that the derived policy be installed in addition to an ordinary insurance system in order to ensure the optimal effort levels in the first place. The solution presented here guarantees ex ante efficiency by construction. Critics might argue that, in the real world, such a rule is not credible ex post because the presence of external effects will drive the international community to provide much more financial aid than initially promised. There are two answers to this. The first and easy one is that in the present model, policy makers have no incentive to provide more funding simply because this would not increase their welfare. However, the model does not account for political pressure, the fact that countries meet in repeated game etc. So the second answer is that in the real world, the Fund would need a commitment technology that assures policy makers it is always going to stick to transparent rules governed by a few verifiable parameters, such as the extent of an economy's openness (external effects), its rate of time preference, and potential damages in the event of a crisis (which could be approximated by the size of the economy). The positive experience with central banks in the industrialized countries suggests that it may be possible for an independent IMF to build up a conservative reputation. This would imply that the Fund's executive directors must remain independent from instructions of their home nations.17 An interesting formal extension of the present model could be to have the IMF pursue the pre-qualification rule with probability smaller than 1, e.g. the probability of IMF rule-based intervention becomes a function of a country's importance to the global economy. Another extension could tackle the private sector which has been left out of the present model. For now, we can conclude that Meltzer's reform proposals are indeed helpful and should be implemented as far as possible.