(Michael Nagle/Bloomberg)

Just how many politicians, athletes and bankers parked their money in tax-free accounts in Panama?

Whether the amounts are big or small, nobody likes to pay taxes. While comparatively few of us set up an offshore shell company, some might be tempted to ask for unrecorded cash for services or to hide a bit of income to avoid a tax hit. When enough people or businesses start cheating just a little, societies are left with lower tax revenues and poorer public services.

Why do some countries cheat more on taxes than others?

Social scientists have evidence that differences in culture and social rules help explain why some people are more willing to cheat. Consider Sweden and Italy, two European countries that arguably lie at opposite ends of the spectrum on paying taxes. Swedes tend to think of themselves as honest people who generally pay their taxes. Other Europeans agree: In an European Union survey, 28 percent picked Sweden as the most honest E.U. country.

In contrast, most other Europeans don’t consider Italians particularly honest. Even Italians think their compatriots are untrustworthy. Italian journalist and writer Giuseppe Prezzolini once described Italy as the “country of cunning people” (“paese dei furbi”) who “worship cunning so much that they even go so far as to admire those who use it against them.” Italians are notoriously likely to evade taxes. Alessandro Santoro’s 2010 book, “L’Evasione Fiscale,” estimates that Italy’s underground economy amounts to some 250 billion euros, all of it untaxed.

So do cultural differences make Swedes diligent taxpayers and Italians tax cheats? Or do the differences in rules, enforcement and incentives lead to differences in behavior?

To find out, we tested how Swedes and Italians behaved.

We set out to test these questions in a large cross-national study funded by the E.U.’s European Research Council, under a project titled “Willing to Pay? Testing Institutionalist Theories With Experiments.” Our laboratory experiments in five countries (the United States, Britain, Italy, Sweden and Romania) have now tested more than 2,000 people on tax behaviors.

This research project has one basic question: How would people in different societies act, given the same set of incentives, choices and institutions? Scholars trying to compare countries face a big problem: The vast differences among different national systems make it nearly impossible to figure out which differences explain what. If properly done, behavioral experiments allow us to focus on specific factors, such as institutions, and see how they influence people from different societies.

Our experiments were very simple. First, we allowed participants to earn real money for performing a simple clerical task. Then we asked them to declare their income for tax purposes. They were free to declare whatever they wanted, but there was a 5 percent chance that they would be audited. If caught cheating, they would have to pay a fine equal to twice the taxes due. In seven of the nine rounds, participants received a part of their tax contribution back as a form of redistribution.

The experiment ran for nine rounds (we describe only eight here, omitting a final “charity” round). In each round, we made a few changes. For example, in the first three rounds we set a 30 percent tax rate, but we varied the amount of money that was redistributed back to the participant. (We’ve described a similar experiment in greater detail in the Monkey Cage and The Plot.)

The results are surprising.

We detailed the findings from our recent experiment in a Frontiers in Psychology article, “Are some countries more honest than others? Evidence from a tax compliance experiment in Sweden and Italy.” Do people from different countries have different levels of “ordinary honesty,” as measured by their tendency to underreport income for tax purposes? We tested about 300 participants in both Sweden and Italy — and we were surprised.

We found no significant difference between the propensity of Swedes and Italians to cheat on taxes.

Average reported income by round and nationality
Average reported income by round and nationality

However, when we dug down a little, we did find important differences in what kind of dishonesty is seen across the two countries. People might be completely honest (paying all their taxes), completely dishonest (paying none of them), or they might be partly dishonest, paying more than nothing but less than they owe.

The average result was the same, but the behavior was very different.

Our findings suggest that Swedes are more likely to be either completely honest or completely dishonest, whereas Italians are more likely to partake in what Dan Ariely describes as “fudging,” or moderately dishonest behavior. Participants who fudged also were more likely to be dishonest with themselves about their behavior during the experiment. Of the fudgers in our survey, 18 percent indicated that they reported their full income. On the other hand, none of those who were completely dishonest said that they reported their full income.

This last finding fits nicely with evidence from Ariely (2012) and Shahar Ayal and Francesca Gino’s social psychological research (2011) showing that individuals choose fudging strategies to maintain a positive moral reputation and self-image. For example, one might be able to maintain a positive reputation and self-image while neglecting to ask for a receipt, because in certain societies norms might actually encourage this type of behavior. The absence of a receipt means a company can opt not to report the sale or service and would then not pay tax on the unreported income

Compare this, however, to the reputational consequences of parking your money in Panama. Maybe we should ask the prime minister of Iceland.

So Italians and Swedes have about the same average level of compliance with taxes. However, when we look beyond the averages, we find differences by country. Italians are more inclined to fudge their taxes. This might seem harmless on an individual level, but this type of moderately dishonest behavior is particularly hard to control. And it adds up.

Moderate dishonesty can be harder to police than outright fraud — but hurts society just as much.

Moral ambiguity gives individuals ample room to cheat a bit — as well as to tolerate moderate wrongdoing. For example, a landscaper might offer a significantly lower price if the customer is willing to forgo a receipt. This creates a gray area, and some people would be attracted by the lower price. If people opt for lower prices without receipts — well, that’s tax evasion.

So fudging can make it difficult for people to do the right thing. And the gray areas make it particularly difficult for governments to crack down on tax evasion. These relatively small acts of dishonesty may tend to spread much more rapidly — and cause more aggregate harm — than the sensationalized dishonest behavior that we hear about in the news.

John D’Attoma is a postdoctoral research fellow at the European University Institute.

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].