Our proposal anticipates greater reliance on fluctuating exchange
rates. But even in countries with fixed exchange rates, market
processes, incentives, diversification and competition will improve
safety and soundness and reduce the risk of financial collapse
accompanying exchange rate devaluations. Better information and
supervision help these processes to work, but they are complements
to, not substitutes for, market discipline.

By eliminating IMF discretion over the circumstances under which
lending occurs, and over the conditions and terms of that lending,
our proposal would end discretionary interventions by the IMF to
distribute emergency foreign aid to insolvent governments and
financial institutions as part of a bailout plan.

It is easy to construct examples of potentially beneficial emergency foreign aid using complicated economic models to show that, in some circumstances, the bene fits exceed the costs. The IMF often justifies its actions as a means of preventing crises from spreading to other countries. This argument has some merit. A crisis in one country calls attention to unwise policies and weak financial systems elsewhere. Global lenders suffer losses, so they may restrict credit to solvent borrowers in other countries that they previously financed.

But the solution does not lie in rescuing foreign lenders. That encourages continued imprudent behavior. A better solution is to give countries incentives to reform their financial structures and improve their policies to make them less vulnerable to contagion. Markets may err for a time, unable promptly to distinguish the solvent from the insolvent borrowers and the more risky from the more secure loans. These errors do not persist for long.

History teaches that misaligned incentives, not inherent financial fragility, are the primary source of insolvency crises in the world today. Government safety nets and IMF bailouts are a major part of those incentive problems. It is possible to correct these core incentive problems by constructing a world financial system subject to market discipline, with fewer and smaller liquidity crises.

Although our proposal would prevent the IMF from giving ad hoc foreign assistance to insolvent financial systems, other mechanisms would be a useful supplement. Bankruptcy laws that delineate loss-sharing rules are the answer to insolvency problems in financially developed economies. Similar rules would have evolved much faster in underdeveloped economies if IMF-orchestrated bailouts had been absent over the past twenty years.

In April the IMF announced a new facility to lend to qualified countries in advance of a crisis. Although the proposal moves the IMF in a direction we recommend, the change is insufficient and incomplete. First, the program is an addition to, not a substitute for, current IMF programs. Lending to countries with insolvent banks could continue. Second, lending criteria are vague and subjective. Unless objective conditions are clear in advance, response to crises will continue to be delayed. Third, there are no collateral requirements to maintain lending standards and reduce risk. The new program, like previous programs, invites discretionary judgment and political influence.

We envision a new IMF, one providing elastic and immediate liquidity to member countries that share a commitment to sound financial practices, and one engendering a system based on rules that increase incentives for prudent behavior by lenders and borrowers, the private sector, the governments and the IMF.

What the IMF would lose is its power to demand policy changes as a condition of its loans. It could not act as the agent of any government; its staff would gather information and make interventions under predetermined rules; it could not exercise discretionary lending that could be prone to corruption. The rules would be the same for all borrowers, no different for those favored by the U.S. Treasury Department than for others. Countries could at their own risk choose other policies by opting out of membership in the IMF.

Such reforms as we propose would greatly reduce the frequency and size of financial crises to the benefit of all nations, and would convert the IMF into a less costly, more stabilizing institution that could offer incentives for prudent policies and market-based solutions.

Charles W. Calomiris is professor of finance and economics at Columbia University. Allan H. Meltzer is professor of political economy at Carnegie Mellon University. Both are visiting scholars at the American Enterprise Institute.