Evaluation of currency regimes: The unique role of sudden stops

This paper tackles two established puzzles in international macroeconomics literature. The first is the lack of systematic differences in the macroeconomic performance across exchange rate regimes. The second is the absence of a clear empirical relationship between macroeconomic performance and capital-account liberalization. We suggest that these negative findings may be due to empirical methods that fail to account for a latent economic 'crisis state', influenced by exchange-rate and capital account regimes, and to allow that latent variable to influence the growth effects of policy regimes. In practice, we model and estimate the latent state of the economy as a crisis probability. Our proposed framework of analysis allows exchange rate and capital-market liberalization regimes to have both a direct effect on short-term growth, and an indirect effect on growth that is channelled through their effects on the crisis probability. The empirical decomposition of the total effect into two conflicting effects helps resolve the puzzles.