A large empirical literature has found positive effects of economic freedom on economic outcomes, such as output and per capita growth. However, several variables in the index are very likely to decline in conjunction with recessions. The purpose of this paper is to determine whether, in the absence of these variable, whether the positive relationship between economic freedom and economic output remains.
This paper makes use of a dynamic panel to compare the performance of economic freedom with and without variables endogenous to business cycles, which pertain to levels of government spending, rates of inflation, government borrowing and interest rates.
Two specifications fall in their statistical significance from the 1 to the 10 per cent level when variables relating to inflation are omitted. The worst case considered finds one specification size of the effect is still 66.3 per cent of the effect size of the standard measure of economic freedom.
These findings are consistent with this kind of endogeneity being a minor problem with the data set when imperfect identification strategies are used, but the issue should be strongly considered when business cycles are pertinent to a research question that makes use of economic freedom data.
This paper contributes to the small literature focused on the robustness of the effect of economic freedom on output, while raising a specific concern that has not yet been explicitly addressed.
