We reassess the connection between capital account openness and capital flows in an empirical framework that is grounded in theory and makes use of previously unexplored variation in the data. We demonstrate how our theory-consistent regressions may overcome some ubiquitous measurement problems...