The Very Effective Carbon Tax and its Political Hurdles
by Ioana Marinescu
Economists conventionally describe carbon taxes as the most efficient policy tool for curbing greenhouse emissions —with the added benefit of raising government revenue. The carbon tax is particularly effective because it raises the cost of carbon pollution, which incentivizes lower carbon emissions, as well as innovation to lower emissions in the longer run. Furthermore, unlike regulation that simply caps emissions for everybody, a carbon tax is more flexible and therefore more cost-effective: those economic agents who can reduce emissions at lower cost reduce them more, while those who face higher costs of reducing emissions can continue their activities with more limited emissions reduction.
While it has long been economists’ pet policy, the carbon tax has also recently gained popularity among Americans: 50% of Americans in a 2016 survey say they support reducing greenhouse gas emissions by taxing carbon-based fuels, while 72% of Americans in a 2019 survey say they support requiring fossil fuel companies to pay a carbon tax .
Yet, despite the support of economists as well as the general public, the United States has yet to adopt a comprehensive carbon tax. In a recent working paper co-authored with Soren Anderson and Boris Shor, we use the first state-level carbon tax initiative in the US to shed light on the political obstacles to the carbon tax and pave the way forward for adoption.
In 2016, Washington State had on the ballot the first state-level carbon tax in the US (I-732). It lost with 40.8% of voters saying yes. In 2018, Washington state had a new ballot initiative (I-1631), similar to the first one but designed to appeal more to liberals. It lost again, albeit with a slightly higher vote share at 43.4%. We draw a number of lessons from analyzing these defeats.
First, ideology is the most important determinant of the carbon tax vote. Democrats are more likely to vote for a carbon tax, and a finer measure of ideology can predict votes even better. In comparison, pocketbook issues do not explain much of the difference between voters: whatever their ideology, everybody similarly dislikes the energy price increases entailed by the carbon tax. Ideology is what makes the difference in voters’ choice: a liberal ideology can more than compensate for energy price increases.
Given that ideology is so important, for a carbon tax to pass, it must be carefully tailored to appeal to the majority of a constituency’s ideology. In particular, spending the revenue of the carbon tax to lower taxes appeals more to conservatives, while spending the revenue on green projects appeals more to liberals.
But ideology is not everything: political campaigns also matter. Washington state’s experience suggests that an effective “no” campaign can torpedo an initially popular carbon tax. In particular, the “no” campaign spent twice as much as the “yes” campaign: $32 million vs. $16 million. One month before the election, an Elway poll of registered voters showed 50% support, but on election day I-1631 lost with only 43.4% “yes” votes. Therefore, for a carbon tax to pass, campaigns must be carefully planned and well financed. The recent creation of the Americans for Carbon Dividend political action committee is one avenue for financing campaigns for the carbon tax.
Finally, age matters: younger people are more supportive of a carbon tax than older people. Therefore, mobilizing young people may facilitate the passage of a carbon tax. In the longer run, as new generations replace older ones, support for the carbon tax may grow, just as support for gay rights has been growing among younger generations.
Overall, the carbon tax is a very effective tool to reduce carbon emissions at lowest cost. However, it faces significant political hurdles, which may be overcome by careful ideological positioning and vigorous campaign spending.
Ioana E. Marinescu is Assistant Professor in the School of Social Policy and Practice and Faculty Research Fellow at the National Bureau of Economic Research.