Least-squares estimates of the response of gasoline consumption to a change in the gasoline price are biased toward zero, given the endogeneity of gasoline prices. A seemingly natural solution to this problem is to instrument for gasoline prices using gasoline taxes, but this approach tends to yield implausibly large price elasticities. We demonstrate that anticipatory behavior provides an important explanation for this result. Gasoline buyers increase purchases before tax increases and delay purchases before tax decreases, rendering the tax instrument endogenous. Including suitable leads and lags in the regression restores the validity of the IV estimator, resulting in much lower elasticity estimates.Note: This table reports coefficient estimates and standard errors from four separate least squares regressions. The dependent variable in all specifications is the month-to-month change in log gasoline purchases, and the coefficients of interest correspond to indicator variables for the month before, the month during, and the month after state gasoline tax changes. All regressions are estimated using the complete panel of 50 states plus the District of Columbia, each observed for 230 months, for a total of 11,730 observations. Standard errors are clustered by state. Asterisks denote statistical significance at the ***1%, **5%, and *10% level.J. COGLIANESE ET AL.