We estimate the effect of exogenous changes in taxes on the US unemployment rate and on several other labor market variables. Our estimates are based on a revised version of the Romer and Romer (2010) narrative record of exogenous tax innovations, with the additional benefit of distinguishing between capital income and labor income taxes. We first show that accounting for the difference between automatic and discretionary tax changes is crucial in order to obtain an unbiased measure of the tax multipliers. An increase in tax receipts of one percent of GDP has a sizeable positive impact on the unemployment rate, and a negative impact on hours worked, labor market tightness and job finding probability. The effect on GDP is also sizeable, but somewhat in the mid range of other values found in the literature, due to the fact that we account for the difference between discretionary and automatic changes in tax revenues. The effect on the unemployment rate of variations in business taxes is larger than that of personal income taxes. We suggest that the latter result poses interesting challenges for future research.