INTRODUCTION

The economic costs of recent financial crises in emerging markets have been enormous. The predictable response has been calls to reform the system, and there is no shortage of official and academic suggestions as to how architecture of the international monetary system could be improved. The problem with evaluating these proposals is that they are based on very different views about why crises have become so frequent and severe in recent years.

We do not think this very basic question can be answered by examining economic developments in the months or days just before and after a crisis occurs. The unfortunate fact is that regardless of the cause of a crisis, its effects on asset values and economic activity are likely to be observationally equivalent. Fischer (1999) makes the point that a poorly managed liquidity crisis will generate permanent declines in the asset values that are identical to losses that are generated by distorted investment decisions. In our opinion, this seems to be correct and is a reasonable basis for government intervention once a crisis has occurred. But in considering how to improve the international monetary system's performance, it is crucial to know whether recent crises were caused by distorted private credit markets or by runs on otherwise healthy markets. If distortions were the fundamental problem, it would be prudent to reduce the scale of IMF assistance packages and focus on capital controls and prudential regulation in developing countries.