Pitfalls in Estimating Asymmetric Effects of Energy Price Shocks

Abstract

A common view in the literature is that the effect of energy price shocks on macroeconomic aggregates is asymmetric in energy price increases and decreases. We show that widely used asymmetric vector autoregressive models of the transmission of energy price shocks are misspecified, resulting in inconsistent parameter estimates, and that the implied impulse responses have been routinely computed incorrectly. As a result, the quantitative importance of unanticipated energy price increases for the U.S. economy has been exaggerated. In response to this problem, we develop alternative regression models and methods of computing responses to energy price shocks that yield consistent estimates regardless of the degree of asymmetry. We also introduce improved tests of the null hypothesis of symmetry in the responses to energy price increases and decreases. An empirical study reveals little evidence against the null hypothesis of symmetry in the responses to energy price shocks. Our analysis also has direct implications for the theoretical literature on the transmission of energy price shocks and for the debate about policy responses to energy price shocks.