Galton's fallacy and economic convergence

The term 'Galton's fallacy' has more than one meaning. Usages of the term starting with Francis Galton are reviewed. Recently the term has been used to denote problems encountered when the neoclassical convergence model is tested in a cross-section of country GNP histories (Barrow-Baumol regressions). M. Friedman and D. Quah independently identify problems which they separately call Galton's fallacy. Friedman and Quah mean different things by the term. Once the nature of various Galton fallacies have been clarified, it is possible to elucidate some issues of econometric estimation that may be encountered in economic convergence regression estimation. Copyright 1999 by Royal Economic Society.