Handling Economic Freedom in Growth Regressions: A Reply to Cole and Lawson

COLE AND LAWSON (2007) STATE THAT “EQUATION (1) IS Lawson’s preferred specification, while de Haan et al. favor Equation (4).†That is not an appropriate summary of our position, however. We do not have a preference for Equation (4). In our papers on the relationship between economic freedom and economic performance we have always estimated Equations (3) and (4), using the Extreme Bounds Analysis to test whether (the level or the change in) the Fraser index is robustly related to economic growth. Our results are that the level of economic freedom is not robustly related to growth, in contrast to the change of the economic freedom. In our reply (De Haan and Sturm 2006) to Lawson (2006), we explain that the main reason that we do not consider Equation (1) a proper specification is that Equation (1) is equivalent to Equation (2). All sides in the debate seem to agree that Equation (2) is definitely not a good model as there is a serious problem of endogeneity of one of the right-hand side variables (i.e. EF1).