Evidence supporting the positive effects of capital account liberalization on growth is mixed at best. Even after conditioning on the quality of domestic financial institutions, a significant number of studies still find no effect. One possible explanation is reverse causation. If low growth countries liberalize in order to spur growth, the observed correlation between growth and liberalization will underestimate the impact of capital account openness. To eliminate this bias, I instrument capital account liberalization with the average level of openness of other countries to capture the "fad" element in financial liberalization. IV estimates indicate a significant positive effect of liberalization on growth, confirming the predictions of economic theory. (C) 2008 Elsevier Inc. All rights reserved