Pii: S0305-750x(00)00006-1

The world is just emerging from the worst ®nancial and economic crisis since the Great Depression. While ®nancial variables, such as exchange rates, have stabilized in East Asia, unemployment remains far higher than before the crisis, and real wages far lower. The World Bank estimates that the 1999 output of the ®ve crisis countries of East Asia will still be 17% below what it would have been had the growth trend of the 10 years before the crisis continued. In addition, large parts of the world remain in a precarious economic positionÐwith deep recession or depression facing several countries in Latin America, and output in many of the economies in transition still markedly below what it was a decade ago. It has become increasingly clear that ®nancial and capital market liberalizationÐdone hurriedly, without ®rst putting into place an e€ective regulatory frameworkÐwas at the core of the problem. It is no accident that the two large developing countries that survived the crisisÐand continued with remarkably strong growth in spite of a dicult global economic environmentÐwere India and China, both countries with strong controls on these capital ̄ows. The crisis in East Asia was not the only crisis of recent years. Indeed, there have been, by one reckoning, 80±100 crises over the past quarter century (Lindgren et al., 1996, p. 20). Crises have become more frequent and more severe, suggesting a fundamental weakness in global economic arrangements. As I put it in one lecture, when there is a single accident on a highway, one suspects that the driverÕs attention may have lapsed. But when there are dozens of accidents at the same bend in the same highway, one needs to re-examine the design of the road. I suggested that one might compare capital account liberalization to putting a race car engine into an old car and setting o€ without checking the tires or training the driver. Perhaps with appropriate tires and training, the car might perform better; but without such equipment and training, it is almost inevitable that an accident will occur. One might actually have done far better with the older, more reliable engine: performance would have been slower, but there would have been less potential for an accident. Similarly, the international economic architecture must be designed to ``work'' not just in the presence of perfect economic management, but with the kind of fallible governments and public ocials that in fact occur in democratic societies. As the crisis spread from East Asia to Russia, and then to Latin America, it became clear that even countries with good economic policies and relatively sound ®nancial institutions (at least as conventionally de®ned) were adversely affected, and seriously so. (Indeed, this was consistent with earlier research that had shown that changes in capital ̄ows, and even crises, were predominantly precipitated by events outside the country, such as changes in interest rates in the more developed countries (Calvo et al., 1993; see also Fern andez-Arias, 1995)). Thus, the rhetoric with which the crisis was begunÐthat globalization, liberalization, and the market economy delivered its fruits to World Development Vol. 28, No. 6, pp. 1075±1086, 2000 Ó 2000 Elsevier Science Ltd. All rights reserved Printed in Great Britain 0305-750X/00/$ see front matter PII: S0305-750X(00)00006-1 www.elsevier.com/locate/worlddev