The political economy of motor-fuel taxation.

INTRODUCTION

Given the important role that public infrastructure plays in economic growth (Aschauer, 1990), whether or not our current highway financing schemes yield economically efficient outcomes is an important public policy question. Yet the available empirical evidence, viewed as a whole, is inconclusive in this matter (Gramlich, 1994). The paper will contribute to this debate by investigating state motor-fuel tax policy within the context of a general-fund model of tax systems. Elected legislators in democratic economies are likely to consider both economic and political factors in designing tax policy. Our paper incorporates both political and economic influences into a theoretical model of tax structures and provides a unique application to the U.S. gas tax policy.

The focus of this study is twofold. First, we develop a theoretical model building on the work of Hettich and Winer (1988), hereafter H-W. Specifically, H-W's analysis is extended to determine how a vote-maximizing politician might respond to changing pretax prices of taxable activities by altering the tax structure. Second, using a large state-level data set for the years 1960-1994, the politico-economic determinants of U.S. gas tax policy are examined. Over this period there have been dramatic fluctuations in the real (pretax) price of gasoline. Have state legislatures responded to these fluctuations by adjusting motor-fuel tax policy in a manner consistent with the predictions of a vote-maximizing model of legislatures? In this model gas taxes are viewed simply as one means of general government finance. This view is in contrast to the often-held view that such taxes are an example of "benefits-received" tax policy, an approach to government finance that comes close to yielding outcomes that satisfy norms of efficient resource allocation (Fisher, 1996; Lee and Wagner, 1991; Lindahl, 1919).(1)

The findings are likely to have important public policy implications. In particular, support for the general fund model of revenue systems would call into question the standard textbook "benefits received" characterization of gasoline taxes. Our findings will also provide evidence on the extent to which U.S. motor-fuel tax policy is driven by the fluctuations in energy prices. Finally, the analysis is potentially applicable to a variety of other settings where the pretax prices of taxable activities change, including similar studies for other countries.