International Financial Crises in Long-Term Historical Perspective

National Bureau Of Economic Research Inc, Cambridge MA

Investigators

Abstract

This proposal involves applications and further development of a recently compiled historical data set for the study of international financial crises. The data set covers 66 countries since their time of independence or dating back as far as eight centuries. It includes, among many other things, data on government domestic and foreign debt and defaults, inflation, banking crises, exchange rates, trade, government expenditures and revenues, conflict and trade. While the data set builds on the research of many previous scholars, it is considerably more extensive in many dimensions, encompassing for example, data on Asia including colonial India and pre-communist China, as well as Africa and all of Europe. One novel feature is cross county historical data on domestic debt covering most of the world going back to the early twentieth century. Previously, domestic debt data had been thought relatively small and unimportant for most emerging market countries until the very recent period. The authors intend to make the entire data set and documentation easily available online. The project has three goals. First, the authors wish to explore the extent to which high debt levels are the main fundamental driver of defaults, inflation, and financial crises, as opposed to factors such as the currency composition and maturity structure of debt. The hypothesis is that, in fact, the latter phenomenon are more a manifestation of vulnerability to debt crises rather than proximate drivers (as conjectured by Diamond and Rajan, 2001). The second project looks at the phenomenon of "graduation" from serial default. Why does it seem to take more than a century for most countries, and what are key factors that separate successful graduates from recidivists. The final project explores exchange rate regimes and serial default, building also on related work on exchange rate classification. A central aspect of the proposal is the maintenance, updating and improvement of data sets on financial crises and exchange rates, which are widely used by other researchers. These data sets involve not only collecting and sorting data, but training students to implement (and improve) a wide variety of sorting and classification algorithms (for example, on measuring exchange rate flexibility and developing metrics to gauge whether a country has graduated (from serial default.) Broader Impact The question of how to deal with international financial crises remains one of the central international macroeconomic policy challenges of recent time. The bulk of existing work is based on relatively short or narrow data sets, obscuring important longer term historical patterns. The results of this project should be of great interest to academics, policymakers at central banks and international lending institutions, as well as to private financial market practitioners.