Results are quantile regression coefficients indicating the effect of final goods tariffs on productivity and firm size. All specifications control for input tariffs and include year and
industry dummy variables. Standard errors, in parentheses, are calculated using a block bootstrap that re-samples over state-industry clusters. In order to reduce computational
time, I follow Abrevaya (2001) and sample (with replacement) m out of n ffiffiffiffiffiffiffiffiffiffiffi clusters, where m=0.1n. Standard errors are based on 500 bootstrap estimates, and are corrected by
m= n
p
. *, **, and *** represent significance at the 10%, 5% and 1% levels, respectively.