摘要：The present paper investigates whether popular measures of economic freedom used in regression analyses by development economists and others are one dimensional. Using the indices provided by the Fraser Institute, Heritage/Wall Street Journal, and Heritage and Freedom House, a principal component analysis indicates that all of the indices above perform about as well as the statistically best single index, and in every case the percentage of total variance explained could be improved by using several principal components. Because the components of the indices are orthogonal, this could be done without multicollinearity problems in regression equations. In sum, the results indicate that economic freedom is not one dimensional and that efforts to squeeze so much into a single index results in lost information and a mis -ranking of the economic freedom of many developing countries.
