On the Political Feasibility of Fuel Taxes

Abstract

This paper proposes a sufficient statistics approach to the political feasibility of fuel tax reforms and applies it empirically to the last attempted French reform. I show that redistributing tax revenue would have turned the reform progressive and made households from the first three income deciles, on average, net beneficiaries. Despite a total share of winners greater than 56%, I find that more than 20% of households from the first three deciles are losing from the reform, as they are strong fuel consumers and therefore the most negatively affected by the reform. Using different social welfare functions translating different views of the social planner, I show that there is a tension between vertical and horizontal concerns, as the largest fuel consumers are more numerous in high-income decile. Finally, I incorporate into my framework a recently documented cognitive bias when it comes to tax & dividend policy: Neglect of Distributed Effects (NoDE). I show that it can have a large impact on the ex-ante evaluation of the political feasibility of a fuel tax reform.