The Penn Effect revisited: new evidence from Latin America

In this paper, we examine the role of relative productivity growth in real misalignment of exchange rates in Latin American countries. Specifically, we verify the validity of the Penn Effect for selected countries in this region. Our sample consists of 15 countries for the period 1951 to 2010. We employ both short- and long-panel data techniques, which allow us to experiment with estimators suitable for short and long time dimensions of panel data. The Penn Effect is found to be supported for the entire sample, and for subsamples. Relative productivity growth is dominant in the real exchange rate movement during periods of mild or weak speculative attacks, as compared with periods of severe speculative attacks. To correct for real misalignment of currencies in Latin America under speculative attacks, relative productivity growth must be sizeable.