Americans Changed How They Drive, Yet Gas Tax Regressivity Remained (Mostly) Stable: A Study on How Evolving Relationships of Mileage and MPG with Income Influenced Gasoline Tax Regressivity in America between 1977 and 2017

Abstract

Taxation of gasoline is an integral part in modern-day American life, and its regressivity has important distributional consequences for millions of Americans. By using two different methods to impute household gasoline consumption, this paper presents new findings on how the regressivity of gasoline tax in America changed between 1977 and 2017. This analysis identifies a misleading description of the relationship between 1977 vehicle fuel economy and household income presented in an official FHWA report and cited in mainstream economics literature. Specifically, I find that gasoline tax regressivity (as measured by the slope of gasoline tax burden on household income) decreased from 1977 to 2001 and then increased back to 1977 levels in 2009 and 2017. I show that gasoline tax regressivity was stable between 1977 and 2017 because changes to mileage-income and MPG-income profiles offset each other. Since 1977, the positive relationship between annual mileage and household income has weakened, causing gasoline tax to become more regressive. This effect is counterbalanced by a similarly weakening positive relationship between vehicle fuel economy and household income over the decades. The paper briefly discusses several social and behavioral factors that may have contributed to these changes in household vehicle choice and driving behavior, such as the advent of online shopping and delivery services, the filtering effect of vehicle repurchase, fluctuations in gasoline price, and the adoption of hybrid vehicles.