Testing the Hollowing-Out Thesis

Abstract

In recent years, many economists have argued that governments are discarding pegged exchange rates in favor of alternative exchange rate regimes such as monetary unions and currency boards on the one hand, or floating exchange rates on the other. Capital mobility, such economists argue, has made pegged exchange rates costly to maintain for long periods, and thus the pegging option is being “hollowed-out.” Few, however, have tried to present evidence that capital mobility has a direct effect on exchange rate regime choices. I present two sets of tests using different measures of capital mobility that provide qualified confirmation that developing countries peg less as capital mobility rises. These tests indicate that direct measures of capital mobility have some correlation with de facto exchange rate regimes but not with de jure exchange rate regimes. Capital flows, a consequence of capital mobility, may have a direct effect on the choices of both de jure and de facto regimes. Governments do not make changes in their declared exchange rate policies in rational anticipation of the growing costs of pegging associated with increasing capital mobility. Rather, governments normally adapt their declared exchange rate policies after capital flows have increased and actual exchange rates have become more difficult to manage. The tests also indicate that hollowing-out has not only been the result of systemic factors such as increasing capital mobility and capital flows, but also due to domestic factors such as growing public sector indebtedness and the spread of democracy.