Tax Policy To Combat Global Warming : On Designing a Carbon Tax

This paper develops several points concerning the design and implementation of a carbon tax. First, if implemented without any offsetting changes in transfer programs, the carbon tax would be regressive. This regressivity could be offset with changes in either the direct tax system or transfers. Second, the production and consumption distortions associated with small carbon taxes, on the order of $5/ton of carbon, are relatively small: less than $1 billion per year for the United States. Stabilizing carbon dioxide emissions at their 1988 levels by the year 2000, however, would require a carbon tax ten to twenty times this size. It would more than triple the producer price of coal and nearly double the producer prices of petroleum and natural gas, would have much more significant private efficiency effects. Third, a central issue of carbon tax design is harmonization with other fiscal instruments designed to reduce greenhouse warming. Ensuring comparability between taxes rates on chlorofluorocarbons and fossil fuels is particularly important to avoid unnecessary distortions in production or consumption decisions. I am grateful to the Istituto san Paolo di Torino, NSF, the MIT Center for Energy Policy Research, and the John M. Olin Foundation for research support. Hilary Sigman provided outstanding research assistance. I am grateful to Peter Diamond, Paul Joskow, Lester Lave, Nancy Rose, Lawrence Summers, John Whalley, and especially William Nordhaus for helpful discussions. Mounting scientific evidence suggests that carbon dioxide emitted in fossil fuel combustion contributes to global warming. This has prompted discussion of carbon taxes, taxes levied in proportion to the carbon dioxide emissions which result from burning different fuels, in virtually all developed nations. Because each individual nation's contribution to global CO2 emissions is relatively small, most advocates of the carbon tax call for coordinated multinational action. The long-run prospects for coordinated action appear dim, however. A tax large enough to significantly slow carbon dioxide emissions would collect revenues equal to several percent of world GDP, and it seems unlikely that national governments would cede control over such a pool of resources to any international body. Most countries cannot noticably slow the rate of global greenhouse gas emissions. Nevertheless, Finland, Sweden and the Netherlands have taken unilateral action in adopting carbon taxes. Other nations may follow their lead in adopting national carbon taxes, with revenues accruing to the domestic treasury. Most previous discussions of the carbon tax, however, have been concerned only with plans for multilateral action. This paper addresses a number of tax design issues which are likely to emerge if the current trend toward unilateral carbon tax adoption continues. The paper is divided into five sections. The first describes the basic structure of the carbon tax, focusing on the policies already in place in Europe as well as proposed taxes for the United States. Section two considers the distributional burden of carbon taxes across income groups. Household data for the United States suggest that the carbon tax falls most heavily on low-income groups. This regressivity could be ameliorated, however, in various ways. The third section examines the production and consumption