Determinants of Tax Pass-Through Rates: A Study of the U.S. Beer Industry

In 1990, the U.S. Congress approved an increase in the federal excise tax on beer from $9 to $18 per barrel. This tax was required to be paid by all brewers and importers, on all produced units as of January 1991. The hike, which was equivalent to an additional 65 cents in federal taxes per 288 ounces (a 24 pack), represented the largest federal tax increase for beer in U.S history. Interestingly, retail prices increased by an average of $1.40 per pack; that is, the tax pass-through was “over-shifted” by approximately 115% (i.e. 75 cents above the 65 cent increase). Economic theory raises questions about the standard assumption that the pass-through rate of alcohol taxes to consumer prices is equal to 100%, but does not provide exact predictions. This study analyzes the determinants of tax over-shifting observed as a result of the 1991 federal tax increase on U.S. beer production. This thesis reports cross-sectional OLS regressions where several variables at the market, firm and brand level are used to explain the change in the nominal price of beer between the last quarter of 1990 and the first quarter of 1991. After controlling for as many factors as the available dataset permits, a robust result across specifications is that non–price vertical restraints (exclusive territory and exclusive dealing contract between beer manufacturers and beer distributors), advertising expenditures (a proxy for product differentiation) and the number of brands are important determinants of pass-through rate. While the three determinants appear to be statistically significant, it is the first two that seem to be of greater economic importance. The fact that vertical restraints are associated with a smaller pass-through rate is consistent with the idea that vertical restraints can serve to mitigate the double marginalization problem; the reason for this interpretation is that theoretical work suggests that a more severe double-marginalization problem can magnify the pass-through rate. Conversely, the effect of advertising is consistent with the theoretical notion that less price-elastic (i.e. more heavily advertised) products will experience a higher pass-through rate. Together, advertising expenditures (evaluated at the mean of the data) and the absence of vertical restraints, can account for $1.075 out of the $1.40 price increase (i.e. 76.8% of the pass-through rate)