Do Italians cheat on their taxes more than Swedes?

This study examines cultural differences in ordinary dishonesty between Italy and Sweden, two countries with different reputations for trustworthiness and probity. Exploiting a set of cross-cultural tax compliance experiments, we find that the average level of tax evasion (as a measure of ordinary dishonesty) does not differ significantly between Swedes and Italians. However, we also uncover differences in national “styles” of dishonesty. Specifically, while Swedes are more likely to be either completely honest or completely dishonest in their fiscal declarations, Italians are more prone to fudging (i.e., cheating by a small amount). We discuss the implications of these findings for the evolution and enforcement of honesty norms.

Here is the research, by Andrighetto,, via the excellent Kevin Lewis.


So what is the average level? It sounds like very few Swedes cheat on their taxes but when they do they cheat a lot. But all Italians cheat but usually only a little.

Sort of like their alcohol consumption patterns.

So they seem to be using an evasive definition of average levels of cheating. The ultimate impact on the Treasury may be the same but the experience of dealing with an individual is likely to be very different. As the Italians will always cheat you, at least a little. While virtually all Swedes will deal with your honestly until, of course, that one in a thousand turns out to be a serial killer who kidnaps you and slowly tortures you to death in his basement.

"Italians cheat more on their taxes" is such a "no kidding" result that they had to switch it up.

There are only two strong results in social sciences - IQ tests have a racial disparity and a random prejudice is likely to be statistically significant.

Or as I would put it - everything my Grandmother told me turns out to be true; everything my professors told me was a lie.

The next round of ground breaking science is bound to be that men don't buy the cow if they can get the milk for free.

"everything my Grandmother told me turns out to be true; everything my professors told me was a lie"

I DO have respect for the potential of folk wisdom to hold some pearls. But can you not see this as a refusal to consider any new evidence?

I think there's an even strong result in social sciences. People who grow up in a place with no schools, no qualified teachers, and no role models in the sense of modern success, tend to have low scores on standardized tests. But what's the point in discussing anything if you have determined that your grandmother is the source of all wisdom and knowledge and anything which refutes any of that is certain to be a lie?

Entertainment, mostly. You thought we were accomplishing something useful here?

"Or as I would put it – everything my Grandmother told me turns out to be true; everything my professors told me was a lie."

The more honest version for you would appear to be "everything my professors told me was a lie except for when it confirmed some racial or ethnic prejudice I had from before." The problem with anti-intellectualism is that you can't very well quote research that happens to agree with you without being hypocritical. The whole point of doing research is that some of what we think we know is likely to be true or close to true while other things we think we know are wrong and we need a way of knowing which is which. Hayek's appreciation for the very limited set of information that most people have about the world is relevant.

As if any of my professors ever had a single original or transgressive thought on racial or ethnic prejudices at all. And you mistake me. I am a Born Again not a Cradle Prejudice Child. Reality hit despite college.

The whole point of research is to push a certain political ideology. Nothing else. It is not anti-intellectualism to point that out. Just as dissent is patriotic, so too criticizing sub-par research is key to restoring and sustaining genuine intellectual activity. Something that no longer takes place in Social Science faculties across the world.

That is why despite all the evidence points to racial disparities in intelligence tests, no one will ever say so if they want to have or keep tenure.

Most people with tenure are unlikely to express an opinion that something is CERTAINLY the cause of a baseline observation, when there are dozens of other available explanations, all of which virtually certain to have at least some effect in expected direction.

It's just not the kind of thing that the kind of people who get tenure are likely to do. It probably has more to do with academic integrity than fear of any sort of thought police.

Thank God there are MR comments, where unchecked racial biases can live on forever.

SMFS, how do you define race? What are the criteria for each group?

Please Jan, don't feed the troll. This has been done and done.

But as I say, like many false things, the idea can live on in the comments section forever.

Anti-science progressives gonna anti-science. Why read the literature when you can read the Guardian and not have to worry about being challenged? God forbid you would have to update your priors, it might be emotionally difficult. And then you might actually support policies that could make a difference instead of pointless campus protests! Much better to be on the PC side of the Guardian than on the side of rationalists like this guy

Cliff, you are obviously another troll, but for the less afflicted, the game modern racists play is to ignore the direct statements of geneticists, and skim them instead, to pull out partial facts and partial truths.

More direct statements here

Cliff - I followed your source to the non-peer review source of the SNP genetic claims. They isolate a handful of genes which appear likely to show heritability, but there is no effort to understand whether these are particularly relevant at a genome level.

Moreover, there is no effort to evaluate the potential for coincidence. An obvious counterargument is that, say, a specific gene is expressed more in a partiuclar racial group, and the racial group may enjoy easier access to economic opportunity - if so, the coincidentally present genetic difference would give a false positive of genetic effect.

Given the database they're working with, it would not have been difficult to perform a basic check for this by controlling for genes related to melanin production or markers for ethnic groups. If the effect of the related SNPs was found across different race groups, that would be a much stronger sign of something that's relevant to improved cognitive function, knowledge which could be used to improve abilities across diverse groups if it was deemed socially desirable upon more careful verification. The controls they did are very theoretically sound, but are not performed in a manner that makes it possible to evaluate the extent to which racism is relevant and/or whether this association can be demonstrated across different groups.

"Or as I would put it – everything my Grandmother told me turns out to be true; everything my professors told me was a lie."


There has been a booming business in the social sciences selling claptrap for decades. I would like to know the origin of these snake oil salesman and how they took root. It is a mystery to me. Possibly a reaction to WW2?

Except what they found did *not* confirm the "no kidding" result.

How doesn't the result exactly confirm every prejudice people have about Italians? And Swedes for that matter?

It says that on average, they cheat on their taxes about equally. Presumably the prejudice was that Italians cheat more.

The point that Swedes are more likely to go all or nothing, while Italians "fudge" slightly more does accord with my preconceptions, but I would not have thought of it before I read this post.

And I bet your prejudice did not predict that *most* subjects, whether Swedish or Italian, would fudge. It is just that among Swedes, there was a detectable minority that did other things.

No they do not cheat about equally. The final amounts are about equal.

The Italians pretty much cheat all the time. The Swedes were noticeably more honest. I would have thought fewer Swedes would fudge, I agree. But that just makes me suspect the definition of the study and their methods.

In the end, this is just what my Grandmother would have told me but she would have done that for free. No massive grant needed.

I've had a lot of dealings with Italians in relation to work-related stuff and in other situations as well.

I can highlight that two of them were very fair dealing, and presumably due to the reputation of Italians, they went to great lengths to ensure that any other conclusion was essentially impossible to arrive at. However, they were very much the exception to the rule, in my experience. One was a farmer that I worked for, another was a restauranteur - both in Australia.

In other countries, I've met a highly disproportionate number who are endlessly scheming ways to milk you for various types of gains, ranging from small to large, whatever they think they might get away with. I could cite several other nationalities which I've reliably had worse experiences with, but Italians are definitely the worst West of the Eastern Bloc. In my experience.

I'm always willing to give them the benefit of the doubt, but given my experience, I exit on the least hint of shenanigans.

@So Much For Subtlety

But the Italians have as high or higher IQ than Swedes. Low IQ folks can and do make great US citizens.

I'm curious how the Race of Swedes vs the Race of Italians differ in IQ once in the USA. Also, which groups should we preferentially admit to our country, the high-IQ and tricky Italians or the lower IQ and higher median honesty Swedes? What about people who have both Southern European blood and a Northern European ancestry? And what shall we expect of the half-Lebanese, half-Mexican who grew up in Mexico City--dumb, smart or terrorist? Just asking questions.

Why are so many MR commenters obsessed with the IQ of groups? If such a thing can even be measured. It's beginning to appear that those regarded as losers in the intelligence lottery have taken the place of racial minorities and some sexual deviants in the bigotry smorgasbord, since those minorities are generally accepted now. Or is "low IQ" a replacement word for persons of various shades of brown? In what sense, exactly, are the low IQs inferior on a moral, if not social, basis to those that consider themselves to be the best and the brightest? Maybe universal IQ testing could determine what humans would be allowed to reproduce, a scenario that's been explored in lots of sci-fi novels. Selective breeding has worked quite well with dairy cattle, thoroughbred race horses and coon hounds, why not humans?

Chuck - yes, I think basically anyone on this forum who refers to "low IQ people" is basically talking about various shades of brown. I think they've deluded themselves into believing something like "I'm not racist, I just KNOW those other guys are inferior". Which, by the way, is basically the definition of racism.

Haven't you heard? It's all the rage these days, called "race realism". Blending psychology, social science, and IQ studies to establish a pecking order in the races. A new dog whistle way of talking about white supremacy (grudgingly modified to give respect to Jews, East Asians) and eugenics.

Yep, the white supremacy that "ranks" Asians and Jews above Europeans, that makes sense


We've been down this road several times before and it is always apparent that you are the one who "just knows" things without any evidence to support them at all (e.g. "one professor mine said a thing 20 years ago!"). The people who disagree with you endlessly cite to science which you studiously ignore. It has nothing to do with "superiority" except in the sense that African Americans are "superior" statistically at basketball or sprinting.

Cliff - a) I never claim to "just know" things. Unlike you, I actually explain my reasoning and openly acknowledge the areas of uncertainty. b) Those professors, well, that was in courses in microbiology and genetics at a premiere world research university. Much more recently than 20 years ago.

I worry about how much to trust this kind of experiment. Paying pretend taxes in a game morally quites different from paying real ones.

If I knew I was a subject in such an experiment then even though I wouldn't know what the experiment was really about, I would still be in a trifling, game like, moral universe quite different from real life.

A couple years ago I dabbled in options trading. While learning the ins and outs of various strategies, and practicing with fake portfolios, the one piece of advice that I heard from all corners was "playing with fake money is useful for learning the basics, but you will never know what it's really like until you put real money at risk."

For most people the psychology of trading easily overwhelms the math. Likewise, the psychology of cheating on taxes is probably heavily tied to the likelihood of getting caught.

Search for the words "Economist laboratory experiment" and when found discard results. This should be a central axiom of life.

Yes, I hope that TC was purposely hiding the fact that this is based on a laboratory experiment which abstracts from any possibly interesting difference in the populations. Otherwise I would say it is quite dishonest. The study does NOTHING to answer the question in the blog entry title. The stereotype is that Swedes pay taxes willingly because they think they are getting value for their money and it is part of their Swedish Kantian bargain. There is no way to measure that with an experiment. Also the sample size is way too tiny and unrepresentative to expect it to tell us anything about a phenomenon of outliers.

In short - "Search for the words “Economist laboratory experiment” and when found discard results. This should be a central axiom of life."

Lots of interesting findings, but this one stood out to me: "we see that compliance responds positively to the efficiency of redistribution: individuals were willing to declare more when they knew that tax revenues produced more public goods". In the U.S., the bulk of federal spending is mandatory spending (primarily social security and Medicare), while the bulk of discretionary spending is for the military. Public goods (besides relatively small sums for "transportation" there are small sums for "housing", "environment", "science", etc.) make up a tiny fraction of total federal spending. Given how the U.S. federal government spends taxes, it's a wonder anyone complies.

Of course, the political implication of this is clear: if the goal is small government (and low taxes), don't spend tax revenues on public goods: the public might decide they like government. Most communities have visible public goods built by the federal government dating back to the Great Depression era, in mine the balustrade that runs the length of scenic Bayshore Blvd. Ironically, pictures of it always figure prominently in ads by the local chamber of commerce promoting the city.

I'm wondering how you define "public good."

If you must look back to 1938 for your "public goods" . . .

In May 1961 JFK promised to put man on the Moon before the close of the decade. In May 2016, Obama promised to put man in women's toilets.

It is amazing the number of bright shiny objects people prefer to the reality of US tax cheating, and the explicit choice not to pursue it.

A few transgenders in bathrooms are more of an issue than billions in lost revenue, the 4:1 ROI on enforcement.

What a country.

I know a man that lives in Italy and pays tax in Switzerland m. To hear him tell it, he cheats large and often.

Confirmed: Italians exaggerate.

"by Andrighetto,": I like the way the punchline was left until the end.

" the evolution and enforcement of honesty norms" Maybe it's different in Italy and Sweden but in the US there seems to have been a devolution of honesty norms, not only in tax compliance but in every other social relationship. Corruption is endemic in government, religion, law enforcement, sports, education, business, the entire spectrum of existence.

....As we continue to a lower trust society.

Yep, beat me to it.

Seriously? If there is evidence of this, I would love to see it. I think corruption was much more common in the past.

The increasing complexity of institutions and technological development create more and more opportunities for corruption, which creative humans are quick to exploit. Mortgage-backed securities, performance-enhancing drugs, Byzantine legal structures, arcane testing devices, etc. are all opportunities for dishonesty that have only come into existence recently.

But there are more, and more active, checks on corruption now, too.

They don't seem to be working. Witness the number of NY state politicians behind bars. They probably occupy their own cell block. Same thing in Massachusetts and Illinois.

What if they are getting caught now, and not so much before? I think that is a very real possibility.

Easily predictable response. If they're being caught, they're proving corruption.

But that says nothing about the relative levels of corruption in previous generations vs now. We have very good incentives and models of identifying corruption now. This both deters it to a much greater extent than in the past, while ensuring that few actually get away with it. (I'm talking about the US here.)

Three words: Mrs. Bill Clinton

So your evidence is one person, who has not even been confirmed to be "corrupt" by a court, despite having something like half the country profess hate her, and who are therefore motivated to suss out and prove any actual corruption? Do you have real evidence?

Also, it looks like a good deal of the data on Italians came from Northern Italy, which may be something of a confounding factor.

Actually, what the paper shows is the effect of norms.

Every Italian cheats, and cheats a little.

Some Swedes cheat, but others do not.

Variability is the story; not the average.

The summary quoted here is compatible with that, but the data are more difficult to interpret.

In the experiments, most Swedes also cheated a little bit, just like the Italians. It's just that among Swedes there was a noticable minority who cheated completely, and another that was completely honest. Maybe honesty was the "norm" in some abstract ethical sense. But the _normal thing_ was to behave like the Italians.

Re norms and from the paper:

" Swedes cheat less frequently, but when they cheat, they are likely to do so completely. By contrast, Italians cheat more habitually, but the intensity of their cheating is more restrained: they hold back from “cheating all they way.”

And live to cheat another day.

First if it is true that the sample is largely from Northern Italy this majorly understates the results. Apparently socio-economic differences between S-Italy and N-Italy are larger than US Black-White differences.
Secondly of course Sweden is no longer a homogeneous society, exactly fitting in what Those Who Can See would expect.

'As of 2011, Statistics Sweden reported, around 1,858,000 or 19.6% inhabitants of Sweden were from a foreign background: that is, each such person either had been born abroad or had been born in Sweden to two parents who themselves had both been born abroad'
Source: wikipedia

(Additionally let me remark that this does not count third generation immigrants or mixed parentage)

Pfft, those Who Can See human nature writ large?

As it happens I am all northern european genetically. I hate cheaters. I cheat. This makes me totally typical of the species, I think.

On a related note, religiousness doesn't increase honesty, but increases the tendency to see oneself as more honest than others.

This is NOT contradictory.

Of course Swedes cheat, Italians cheat, Somalis cheat. The question is to what degree and what percentage.

I fail to see how your religiousness link have to do with the question at hand, since I never mentioned religion? Let me just note that religiousness is probably strongest under Somalis, then Italians, then last of all Swedes.

I can only conclude that your comment is less a factual argument or response than a tribal signal. Please prove me wrong.

You should probably spend more time with Dan Ariely books. You would be more comfortable with human nature as a universal, mostly positive, but never perfect thing.

You might grow to understand that looking for differences might be useful, but might also cater to the worst part of ourselves.

"Leave the guns, take the tax returns...I don't want an audit."

Not a big worry, our revealed preference is for cheating.

The Treasury estimates, for example, that every additional $1 invested in IRS tax enforcement beyond current levels would yield $4 in increased revenue. As Commissioner Koskinen summarized, “Essentially, the government is losing billions to achieve budget savings of a few hundred million dollars.”

From my link above.

thought experiement:

put an italian in a room with swedish food and a swedish food in a different room with italian food. i call this the great food switch experiement.

Swedish food is everywhere because of IKEA cafes. Italian food is everywhere because it is generally decent.

Comments for this post are closed