The IMF's balance of payments financing is supplementary to what is available from the private sector (as well as from other official lenders).¹³ In the conception of the international environment that inspired Bretton Woods, private capital flows were not expected to be a major factor; indeed, unlike current account convertibility, capital account convertibility was not expected to be the rule. In such a context, the financing role of the IMF was clearly intended to be central to the international monetary system. In practice, however, there were few drawings on IMF resources in the early postwar years, and, in recent years, the vast expansion of private capital flows has reduced the need for IMF financing for those countries with secure access to international capital markets--in practice, most of the industrial and some of the developing countries.

The question therefore arises whether the IMF's role as a supplier of balance of payments financing can continue to be justified. As in other areas of government intervention, the general rationale for the IMF's role needs to be grounded in the existence of market imperfections or in the inadequate provision of public goods, both of which IMF lending can help to remedy. At this level of generality, the rationale for the IMF's financing activities remains much the same as it was at the time of the Bretton Woods conference: unimpaired access to private financing of payments imbalances is generally available to some important countries in the world economy, but not to all countries. Accordingly, in the absence of IMF financing (accessible under appropriate conditionality), many countries facing balance of payments difficulties might need to adopt policies that would be unnecessarily destructive of national and international prosperity. By helping to avoid such outcomes, IMF financing benefits both those countries experiencing balance of payments difficulties and their partners in international commerce and finance.

Market Imperfections

Market discipline generally speeds up the recognition of unsustainable policies and thereby brings about needed policy adjustments. However, the market does not always act on the basis of appropriate judgments or reflect socially optimal assessments. At times, markets continue to provide financing for unsustainable policies, delaying needed adjustment. When a belated recognition of problems occurs, the change in market sentiment can lead to a violent reversal of the capital inflows, causing a balance of payments crisis. Without official financing, the crisis would force on the authorities disruptive adjustment, with potentially high costs in terms of output and employment losses. For other countries whose economic fundamentals justify access to private capital, access might be available only at a high cost, or not available at all. Lack of access to international capital markets would exclude a country from benefiting fully from the efficient allocation of capital.

Four important factors affecting private lenders can cause these market imperfections to occur.

• Imperfect information about a country's policies and the economic circumstances facing it can lead to shifts in sentiment that are not justified by fundamentals. A recent example is the large flows to most emerging markets in the early 1990s and their sharp interruption or reversal after the Mexican crisis (see Folkerts-Landau and others (1995)).

• Coordination problems among lenders can lead to bandwagon effects or "free rider" behavior. The Latin American debt crisis of the 1980s illustrates both aspects of the problem. Lack of concern for potential risks and competition among banks to recycle the revenues of oil producing countries produced an initial period of overlending in the late 1970s. The onset of the debt crisis in 1982 led to a protracted period of negotiation of debt relief, during which creditors attempted to position themselves favorably with respect to both the borrowing country and other creditors.

• Problems in enforcing loan contracts on sovereign borrowers can produce a reluctance to lend. The inability of borrowers to credibly commit to repay loans, together with the absence of clear recourse of creditors in the event of default, may lead to suboptimal lending (see Eaton and Fernández (1995)).

• Multiple equilibria can exist. For instance, there may be one equilibrium in which lenders believe the authorities intend to do what they say and in which money is available on reasonable terms that allow policy commitments to be fulfilled, and another equilibrium in which lenders do not believe the authorities and in which investors demand a high risk premium that, in turn, provokes a balance of payments crisis. There is no assurance that the market, left to itself, will choose the better of the two equilibria. Although empirical work in this area is still in its infancy,¹⁴ an examination of speculative attacks on ERM currencies suggests that these episodes are not distinguished from non-attack periods by different economic fundamentals. ¹⁵

The IMF can alleviate these market failures through both its surveillance and its lending activities. Through stronger surveillance, including encouragement of countries to make more timely data publicly available, the IMF would help to provide financial markets with the information that they need to form judgments concerning a country's creditworthiness, alleviating the first problem described above. The IMF has also assisted members in their relations with commercial banks and official creditors through strengthened surveillance, thereby creating confidence in the countries' adjustment policies and facilitating debt-restructuring agreements, and thus helping to correct the second market imperfection (see, for instance, Chapter III of Watson and others (1986)).

However, it is through financing and the accompanying exercise of conditionality and provision of policy advice that the IMF can have the most impact. The commitment of resources in support of a program signals the IMF's confidence in a member's policies. Such a signal is more likely to convince private lenders of the creditworthiness of a borrower than mere announcement of support. In this way, the IMF can catalyze other sources of financing and help extract commitments from other lenders.

The IMF must also safeguard its resources and take account of the repayment of resources to the IMF in its conditionality. The exercise of conditionality makes the third problem cited above less severe. Thus, the IMF may be willing to provide support in some circumstances when private lenders are not, as the latter have less influence on the policies of borrowing countries. Moreover, the exercise of conditionality would be expected to have externalities on other lenders, who would also reap the benefit of sounder policies in borrowing countries and their increased ability to repay. In this way, as well as by signaling confidence in a member's policies, the IMF may help avoid bad equilibria--the fourth type of market imperfection.

International Public Goods

A second major area where the IMF needs to supplement private capital flows is the provision of international public goods. In particular, the IMF, by being prepared to make resources available in times of adverse shocks, can encourage countries to greater openness to trade and capital flows. Because openness is a public good (gains from trade accrue to all parties), it is in the global interest to encourage it (see Frenkel, Goldstein, and Masson (1988)); indeed, the growth of trade and the elimination of foreign exchange restrictions are explicitly listed as purposes of the IMF in Article I. The IMF can encourage provision of these public goods by supplying a form of insurance against adverse shocks for countries (for example, Mexico, Argentina, the Czech Republic, and Poland, among many others in recent years) that are liberalizing or have liberalized their economies. This insurance involves access to IMF credit on reasonable terms should the balance of payments deteriorate. Private financial markets are unlikely to provide such insurance in adequate amounts; if they do, it is at a risk premium that is too high because lenders cannot fully appropriate the benefits to global welfare. The availability of IMF resources provides the assurance necessary for countries to engage in desirable policies that have favorable externalities and thus enhances the ability of other organizations (for example, the World Trade Organization) to fulfill their mandates in this area.

It is true that, just as in other insurance activities, such as fire insurance, there is a potential moral hazard problem: the insured party may be induced to take on more exposure to adverse shocks than is desirable. If the IMF is likely to bail out countries when they run into trouble, won't they undertake inadvisable policies, in particular, unnecessarily risky ones?

In examining this moral hazard problem, three things have to be kept in mind. First, as in the case of fire insurance, some risk taking is desirable;¹⁶ the risk-minimizing solution, which might prevail in the absence of a safety net, would be to close the economy to foreign trade and capital. Second, excessive risk taking is limited by coinsurance, so that the insured party can expect to suffer a significant part of any loss. In particular, a country that chooses inappropriate policies that would lead to balance of payments difficulties typically pays an important cost in the subsequent adjustment process. In addition, IMF financing does not consist of grants, and the recipient of such financing must pay interest and promptly repay the principal.¹⁷ Finally, as an extension of the analogy with fire insurance, the IMF can be said to play the role of fire inspector through its exercise of surveillance and conditionality, which reduces the scope for excessive risk taking.

Another international public good is an exchange rate system that facilitates smooth adjustment and promotes stability; as Article I(iii) indicates, its promotion is one of the purposes of the IMF. Of course, the context of infrequently adjusted pegged rates has changed radically with the replacement of the Bretton Woods system by generalized floating. In large part, this change reflects the infeasibility of defending fixed exchange rates at a global level in the context of expanded capital flows (as discussed above) and the preference of the largest industrial countries to devote their monetary policies to promoting domestic stability objectives (as described in Mussa and others (1994)). As these countries have been generally successful in recent years in delivering low inflation, it can be argued that it would be a mistake to distract their attention from keeping their own domestic houses in order (see Frenkel, Goldstein, and Masson (1991)). Nevertheless, the IMF retains an important responsibility to help ensure through its surveillance activities the smooth functioning of the flexible exchange rate system with a minimum of volatility and misalignments.