Even though they can last for decades, fixed exchange rate regimes are increasingly seen as temporary arrangements, (Eichengreen, 1994). The hollowing out hypothesis, see, e.g. Fischer (2001), holds that fixed but adjustable exchange rate regimes of various kinds are scheduled to disappear in favor of either rigidly fixed rates (dollarization or currency boards) or free floats. The evidence so far is mixed but suggests that pegged regimes are not indeed fit for survival. For instance, in their comprehensive analysis Husain et al. (2004) report that, taken together, exchange rate regimes have an average duration of about 11 years in the post-Bretton Woods period. They also find that fixed regimes last longer than free floats for all countries taken together, but with sharp differences across country groupings. In particular, freely floating regimes last considerably longer amongthe group of developed countries while pegged regimes exhibit low duration in the group of emerging market countries, about 8 years.1 Extrapolating their observed transition probabilities, they predict a decline in pegged regimes. Furthermore, if countries graduate from the developing to the emerging markets category and from the latter to advanced, their transition probabilities predict a widespread move to floating.
