Evaluating New-Keynesian Models of a Small Open Economy

Abstract

We suggest a strategy to evaluate members of a class of New-Keynesian models of a small open economy. As an example, we estimate a modified version of the model in Svensson [Journal of International Economics (2000) Vol. 50, pp. 155-183] and compare its impulse response and variance decomposition functions with those a structural vector autoregression (VAR) model. The focus is on responses to foreign rather than to domestic shocks, which facilitates identification. Some results are that US shocks account for large shares of the variance of Canadian variables, that little of this influence is due to real exchange rate movements, and that Canadian monetary policy is not adequately described by a Taylor rule.