Using arbitrary precision arithmetic to sharpen identification analysis for DSGE models

Abstract

Summary We introduce arbitrary precision arithmetic to resolve practical difficulties arising in the identification analysis of dynamic stochastic general equilibrium (DSGE) models. A three‐step procedure is proposed to address local and global identification and the empirical distance between models. The method is applied to monetary and fiscal policy interaction models, revealing exact observational equivalence in a small‐scale model between an indeterminate passive monetary and fiscal policy regime and determinate regimes, and near observational equivalence in a medium‐scale model. Additionally, the method yields new insights for a model with news shocks, demonstrating that wage markup shocks can be replaced by unanticipated moving average shocks, resulting in near observational equivalence.