Abstract In a widely cited study on the distribution of corporate tax burdens Zimmerman 1983 concludes that, "The roughly fifty largest U.S. ... firms ... have significantly higher worldwide tax rates than other firms." However, Porcano 1986 provides conflicting evidence, finding an inverse relationship between firm size and U.S. effective tax rate. In this paper we use the 1989 COMPUSTAT data base to replicate and reconcile the results obtained in Zimmerman and Porcano. We demonstrate that the disparate results reported in these studies can be attributed largely to the differences in their empirical procedures. In particular, we show and explain how alternate effective tax rate definitions, sample selection procedures, firm size proxies, and data aggregation methods affect the direction and degree of the relationship between firm size and effective tax rate. Our results suggest that the theories which purport to explain cross-firm differences in effective tax rates, and the empirical representations of the conceptual variables, need to be further refined before consistent results can be obtained.
Wilkie et al. (Thu,) studied this question.