This paper describes CBO’s method for measuring the distributional effects of a tax on carbon emissions and the agency’s rationale for choosing that method, while also comparing it with CBO’s prior method and methods used by other researchers.
By Dorian Carloni and Terry Dinan
Putting a price on emissions of carbon dioxide, either by taxing them or by establishing a cap- and-trade program, is one policy that lawmakers could consider to address climate change. Although such a policy could encourage cost-effective reductions in emissions throughout the economy, lawmakers have expressed concern about whether it would disproportionately affect lower-income households.
Determining the distributional effects—that is, the effects on households at different income levels—of a policy that would price carbon emissions (referred to in this paper as a carbon tax) is challenging, and the results would vary substantially depending on how the effects were measured. Using a method that allocates the burden to households on the basis of their income rather than their consumption, the Congressional Budget Office estimates that the burden on households in the lowest income quintile, measured as a percentage of income before transfers and taxes, would be roughly twice as large as that imposed on households in the highest income quintile. The burden on households appears less regressive if measured as a percentage of income after transfers and taxes, largely because of the progressivity of the existing federal transfer and tax system.
This paper describes CBO’s current method for measuring distributional effects and its rationale for choosing that method, while also comparing it to CBO’s prior method and methods used by other researchers. Compared with CBO’s prior method, the agency’s updated method better reflects the tax burden on annual income and accounts for differences between consumption and income that are due to life-cycle patterns of households’ spending, and it better facilitates a comparison between the burden of a carbon tax and the burden of other existing federal taxes.
We also describe the limitations of CBO’s updated method relative to CBO’s prior method and the methods used by some other researchers. Compared with CBO’s updated method, the method the agency used in 2012 better captured the burden that the tax would impose on households consuming out of accumulated wealth and better aligned the revenues raised in a given year with the burden that households incur in that year.
In addition, CBO’s updated method and prior method share limitations that are common to all methods that measure the tax burden using data on consumption and income from a single year. Specifically, they do not account for the effect of the tax on households at different points in their life cycle, some of the interactions between the carbon tax and the existing tax system, the additional tax burden on saving if the tax rate increases over time, the macroeconomic effects of the tax, and differential behavioral responses to the tax across income groups.