Modeling the Effects of Carbon Taxes on U.S. Household Incomes Using RTI-ADAGE™

Guiding key policy discussions on greenhouse gas mitigation with advanced economic modeling and analysis

Client
RTI-funded

As political leaders in the United States and around the world tackle the looming issue of climate change—and with greater urgency since the 2015 United Nations summit in Paris—carbon taxes represent one possible policy solution to help reduce greenhouse gas emissions. But the term “carbon tax” is often tossed around as a buzzword, while the details and potential effects of such a system remain vague to the public and even some inside government.

Leading economists, including a group from RTI, sought to shed some light on carbon taxes at the Stanford University-led Energy Modeling Forum. This recurring event brings together experts on energy and the economy to study issues that society will face in the near future. At the 2014 Energy Modeling Forum, which focused on policies for controlling greenhouse gases, we applied an RTI-designed model of the economy to estimate the effects of carbon taxes on the income of households across the U.S.

Projecting Impacts on Regional Economies of Multiple Revenue Distribution Scenarios

At first glance, carbon taxes might appear to affect industry alone, but they could actually make a difference in the bottom line for millions of households. Families could end up paying more for gas, electricity, and many other products as companies pass along the added cost to their customers. But depending on how the government decided to distribute revenues from such a tax, Americans could find some of the money coming back to them in the form of payroll tax cuts, capital tax cuts, or direct payments. Our research considered all of the above scenarios, as well as two potential levels of carbon taxes.

To carry out this complex analysis, we used our own “computed general equilibrium” (CGE) model, RTI-ADAGE™. The model, whose name stands for Applied Dynamic Analysis of the Global Economy, allows us to take multiple industrial and policy factors into consideration when analyzing energy and environmental issues several decades into the future.

We found that a carbon tax could affect different regions of the country in significantly different ways. Households in areas with more established green economies, such as New England, the mid-Atlantic states, and the Pacific coast, tended to do relatively better economically regardless of which tax distribution method we used. The oil-producing West South Central states saw the opposite effect, with households tending to see larger impacts to income.

In terms of how tax revenue was returned to households, we found that direct, lump-sum payments were the most progressive, meaning that low-income households bore a less than proportionate share of the economic burden relative to higher-income households. Lump-sum payments led to a net gain, albeit sometimes a small one, for households at the bottom of the income scale. Capital tax rebates were the most regressive, resulting in gains for higher-income households, while labor tax rebates produced mixed results depending on region.

Helping Policymakers Understand the Economic Impact of Carbon Taxes

A carbon tax would have ripple effects throughout the U.S. economy. Our work at the Economic Modeling Forum contributes to the policy community’s understanding of many of those effects, leading to better-informed planning. Regional variations we found could indicate where a carbon tax is likely to find support or opposition. Our analysis of different methods of distributing carbon-tax revenue could help government leaders better understand the broad economic impacts of carbon policy.

The results our team contributed to the Energy Modeling Forum inform important policy discussions, helping to shape the way policy makers understand and act on this issue. Further modeling and study will provide invaluable information for academics, industry leaders, and policymakers comparing options for dealing with greenhouse gas emissions.