A CRITICAL ASSESSMENT OF MEASURES OF CENTRAL BANK INDEPENDENCE

Abstract

The methodology of measuring central bank independence suggested by Bade and Parkin three decades ago and inverse correlations between the measures and inflation have been widely accepted. The measurement literature has made important contributions identifying elements of institutional design that establish the relationship between the central bank and the government and the measurements from an ordinal perspective provide some useful insights about the major institutional redesigns of central banks during the past several decades. At the same time, the measurement literature is problematic. First, a dummy variable bifurcating central banks into groups of independent and less independent central banks can duplicate the standard correlations suggesting the information content of the specific measures is limited; second, the correlations between inflation and measures of independence are not as stable as claimed; and most important, the measurements are subject to classification problems that have not been fully appreciated. Review of the institutional design and history of the Bank of Korea, Bank of Japan, and the Federal Reserve provide evidence the classification problems are important. The measures of independence are more appropriately regarded as ordinal rankings of central bank independence rather than considered cardinal measures suitable for econometric modeling. (JEL E50, E58, E65)