U.S. Department of the Treasury Internal Revenue Service (IRS) 1040 Individual Income Tax forms for the 2014 tax year are arranged for a photograph in Tiskilwa, Ill., U.S., on Monday, March 16, 2015. (Daniel Acker/Bloomberg)

Funding government is difficult. To do it, Americans tend to want the wealthy and corporations to pay more taxes. For example, in a recent Pew Research Center poll, 75 percent of Democrats and 52 percent of Republicans said that they were bothered “a lot” by “the feeling that corporations don’t pay their fair share.  Similar fractions were bothered by the same feeling about the wealthy.

But rather than raise taxes on the wealthy, governments, especially at the state and local levels, have preferred to keep taxes low, or even cut them. To fill budget gaps, they increasing rely on fees. Why? David Segal writes that “Politicians tend to regard fees as more palatable than taxes, and more focused too.”

But fees can backfire. In what has turned into a nightmare for St. Louis County politicians and public administrators, fines and charges targeting low-income people and African Americans have become a major grievance in ongoing disputes between community members and elected officials.

And Missouri is not alone. Gary Wagner, an economist at Old Dominion University, found a correlation between a decrease in government revenue and a rise in tickets by police in North Carolina. These trends are common around the country.

Our research has turned up a surprising finding, however: fees are regarded more negatively than taxes. In fact, our experiments suggest that tax compliance is actually lower when the word fee is used than when the word tax is used.

Our “Willing to Pay” Project has conducted tax compliance experiments in the United States, United Kingdom, Sweden, Italy and Romania.  In these experiments, participants are asked to perform a series of clerical tasks in which they copied rows of text from a spreadsheet onto a computer. For each row copied correctly, individuals earned “money” denominated in an artificial currency that we created for these experiments.  Each row copied correctly was equal to about one penny in American currency.

In each round of the experiment, participants were told their income and asked to report what they earned. They then paid taxes, and some of the money collected in taxes was redistributed back to the students or a real public institution.

Here’s the catch: the students were free to report what they pleased, with a 5 percent chance of being audited. If they cheated and were caught, they had to pay a fine equal to twice the amount of taxes due. In the end, the compliance rate was measured as total income reported divided by total income earned.

We conducted several experiments at Royal Holloway University in London. On the first day we read the word fee to our participants. On the second day in which we read the word tax. The rationale for using the word fee was that we thought the word would be more neutral, as do many politicians. We couldn’t have been more wrong.

steinmo1
In fact, in our experiments when participants were read the word fee instead of tax, they paid, on average, 10 percent less of the taxes they owed.

When these individuals were put in a situation in which their taxes were redistributed based on earnings, those in the “fee” session complied 19 percent less than respondents in the “tax” sessions.

steinmo2

So not only are “fees” more negatively construed than “taxes,” but when participants are given the chance to pay a fee or a tax into a fund which progressively redistributes the money, they would overwhelmingly rather pay the tax.

Although we don’t have enough data at this point to make strong conclusions, we believe that these results matter. Other social scientists can easily replicate our experiments and test responses to taxes and fees.

Practically, politicians should take note: the word tax may not be as politically charged as we think. Instead, the word fee might be the more negative of the two.

John D’Attoma is a research associate at the European University Institute and Ph.D. candidate at the University of Missouri-Saint Louis and Sven Steinmo is a Research Professor at the European University Institute. The research leading to these results has received funding from the European Research Council under the European Union’s Seventh Framework Programme (FP7/2007-2013)/ERC Grant Agreement n. [295675].