- Joint Program Report
A primary reason for implementing a carbon or greenhouse gas tax is to reduce emissions, but in recent years there has been increased interest in a carbon tax’s revenue potential. This revenue could be used for federal deficit reduction, to help finance tax reform, support new spending priorities such as infrastructure spending, offset the burden of the tax on households, or other purposes. With an environmental goal to reduce emissions to very low levels, programs that become dependent on the revenue may come up short when and if carbon revenue begins to decline. To date, the revenue potential of a carbon tax has not been studied in detail. This study focuses on how much carbon tax revenue can be collected and whether there is a carbon “Laffer Curve” relationship, with a point where revenue begins to decline. We employ the MIT U.S. Regional Energy Policy (USREP) model, a dynamic computable general equilibrium model for the U.S. economy, for the numerical investigation of this question. We consider scenarios with different carbon prices and emissions reductions goals to explore how they may affect whether and at what tax rate revenues peak. We find that a sufficiently high tax rate would induce a revenue peak between now and 2050. For the scenarios we study, however, we find that carbon tax revenue is a dependable source of revenue to finance federal fiscal initiatives over a thirty-year period at the minimum. We also explore how the cost of low-carbon technology and existing energy policies interact with tax rates and revenues. Our results indicate that lower costs of abatement technology make emissions more responsive to the tax rate, and removing regulations on renewables and personal transportation results in more carbon tax revenues. Our results also show that either lowering technology costs or removing existing policies would reduce the welfare cost of a carbon policy with specific reduction goals, with a larger offsetting gain from eliminating distortions associated with existing policies.