Empirical similarity for revealing the US interest rate policy: modeling case-based decisions of the FOMC

Abstract We model the Fed’s decisions about interest rate adjustments via the flexible nonlinear empirical similarity (ES) concept which relies on ideas of case-based reasoning particularly suitable for decision making under uncertainty. We postulate that the Fed’s adjustment decision in a given situation should be close to those in similar economic situations. We evaluate the empirical fit of the ES concept in comparison with the linear reaction function related to the Taylor rule for the period from 1987 till 2008. We identify and analyze critical time points particularly for those Fed decisions which were conducted in rather uncertain economic environments.