Asymmetric Effects of Exogenous Tax Changes

We study whether output responds symmetrically to tax increases and decreases in postwar US data, using the identification strategy in Romer and Romer (2010). We find evidence of important asymmetries: the output response to a tax increase is statistically insignificant, but output shows a significantly positive and permanent increase following a tax decrease. We show that this asymmetry appears to be driven by individual-income tax changes, and is transmitted to the economy through asymmetric response in aggregate consumption to tax increases and tax decreases. We also present a simple model that rationalizes our empirical findings, and illustrates how asymmetric output and consumption responses to sign-based tax changes can be generated by plausible consumption-adjustment costs.