The empirical literature on the determinants of economic growth, which relies on cross-country growth regressions, has struggled with the statistical problem of bias from two sources: omitted variables and simultaneity. 1 The first problem arises because of country characteristics that affect growth but are not observed by the econometrician. The second is because of the fact that the determinants of growth of GDP, such as investment in physical capital, may also be affected by this growth; evidence of this is provided by M. Blomstrom, RE Lipsey, and M. Zejan. 2 In recent years, the problem of omitted country characteristics has been addressed via fixed effects. 3 The problem of simultaneity, however, needs more attention. As is well known, the standard solution to the problem of simultaneity bias is instrumental variable (IV) estimation. For instruments we need variables (1) that are correlated with our endogenous explanatory variable,(2) have no direct impact on the dependent variable, and (3) are not affected by changes in the dependent variable. Since the dependent variable in this literature is either the level or growth rate of gross domestic product (GDP), finding instruments that satisfy criterion (3) is, as NG Mankiw, D. Romer, and D. Weil put it,“a formidable task.” 4 Because the effects of changes in GDP are pervasive, it is hard to conceive of truly exogenous variables in the growth context, a fact that has led F. Caselli, G. Esquivel, and R. Lefort to state,“We wonder whether the very notion of exogenous variables is at all useful in a growth framework (the only exception is perhaps the morphological structure of a country’s geography).” 5