Economists love carbon taxes. According to polls, so do roughly two-thirds of Americans. Yet voters in one of the bluest states in the nation, Washington State, twice rejected a carbon levy in recent years.

What’s going on? Why is a carbon tax, an effective tool to combat climate change that appears popular in public opinion surveys, such a political loser?

In a new paper published in July by the National Bureau of Economic Research (NBER), three researchers attempt to answer the question. And they have drawn some important lessons for those interested in the political viability of carbon pricing.  

The authors, economists Soren Anderson at Michigan State University and Ioana Marinescu at the University of Pennsylvania and NBER, and political scientist Boris Shor at the University of Houston found that:

• Political ideology is the biggest driver of voter attitudes towards a carbon tax. Resistance to personally paying more taxes matters too, but not as much.
• Voters give little weight to the benefits of a carbon tax, including promised tax rebates.
• Well-organized, well-funded opposition to a carbon tax can sharply erode support for the levy.
• Based on projections from the Washington State experience, no voters currently would approve any state-based carbon tax initiative, though some contests would be close.

The researchers looked at Washington’s 2016 and 2018 voter initiatives. They studied precinct-level voting data and matched those results with a national public opinion survey on attitudes towards a carbon tax.

Both Washington initiatives would have imposed a tax starting at $15 per ton of CO2 and gradually rising. The 2016 proposal would have rebated the revenue by lowering the sales tax rate and increasing the state Earned Income Tax Credit. The 2018 version would have spent the revenue on renewable energy and other “green” initiatives. Both were rejected—the 2016 version got 40.8 percent of the vote while the 2018 version received 43.4 percent.

The study found that perceived personal tax burdens do matter. People who faced higher taxes because they commuted to work by car or lived in a big house were less likely to support the carbon levy than those who got to work other ways or lived in a smaller home.

But ideology was by far the biggest factor in explaining how people voted on the carbon tax. For instance, voters who favored unrelated liberal causes such as gun control were more likely to back a carbon tax, a result that seems consistent with patterns throughout the US electorate these days.

Did the design of the tax matter? Economists have worked hard to structure a carbon tax that is both effective and politically palatable. And the question of whether, and how, to rebate the tax revenue is important to them. But this study found that voters didn’t care much at all. The green spending version did a bit better than the rebate model, but neither passed.  One reason may have been that voters simply did not believe political promises of how the money would be spent.

Perhaps the authors’ most powerful result was the effect of the campaigns themselves. Opponents of the 2018 version outspent supporters by 2-1 and successfully reframed the measure from the “fee” used in the ballot language to a tax. In the end, Washington voters were 20 percentage points less likely to support a carbon tax than survey respondents in other states—even though there was little difference in their support for 11 other environmental initiatives.

Why? Because Washington voters were exposed to a real campaign for an actual carbon tax. They confronted immediate costs in return for uncertain future benefits. While those in other states were asked merely about a theoretical plan, Washington voters heard powerful criticism of a specific idea. They also may also have been affected by a common bias against change.

The authors not only found that a carbon tax was a tough sell in Washington State, they concluded it would fail in every other state with a ballot initiative. In some states 45 percent or more of voters would back the green version. A Washington-like carbon tax initiative would win a bare public majority in Vermont—except the Green Mountain State does not have voter initiatives.

Some carbon tax experts fear that analysts may have “overlearned” from the Washington experience. My Tax Policy Center colleague Adele Morris says that a national carbon tax might fare better than a state version.  Among the reasons: Since a federal tax would replace clean air regulations, it more likely would win business support; in addition, a major national initiative would be a more credible weapon against climate change than a tax in a single, relatively small state.  

Still, this careful study is further evidence of the reluctance of voters to tax themselves to reduce carbon emissions. Carbon taxes have run into similar problems in France and Canada and selling them here will not be easy. If supporters of what seems to be an effective solution to an existential problem are going to turn their ideas into policy, they are going to have to tell a much better story.