The Criminal, Social, and Economic Cost of Immigrants in the EU

There’s been a huge backlash against immigration in the West recently. In the UK, anti-immigration protests have taken place:

Image from a recent UK demonstration against immigration

The UK Prime Minister is actively trying to display an image of toughness against immigration—and failing:

Across Europe, people support policies to stop and revert immigration.

The Danish socialist prime minister implemented a zero asylum policy and is toughening deportation laws. A Dutch party in the ruling coalition is pushing for sweeping anti-immigrant rules. But most governments are not acting as fast as public opinion is asking, so right-wing parties around Europe are gaining traction and are becoming front-runners

Source: Tomas Pueyo for Uncharted Territories, data gathered by Claude.

But the EU has had plenty of immigration for decades, since Schengen opened border movement, and despite some initial fear, it turned out quite well. What happened? A lot of the immigration came from within the EU, but now the vast majority is from outside. Of them, the biggest group is Muslims, at 40%.¹ This is causing friction:

• Austria has banned headscarves for girls below 14.

• German Chancellor Merz vowed to expel 80% of the 1M Syrians currently living in Germany within three years.

• Even in the US, which doesn’t receive a particularly high amount of Muslim immigrants, President Trump called Somalian immigrants garbage and signed an expanded travel ban, applicable to many from the Middle East and Africa.

When you read the media, social media, and talk on the street, the main complaints you hear are the social and economic costs: Immigrants cause more crime and more terrorism, they destroy social cohesion, work less, and consume much more welfare. Are these claims true? For which immigrants?

1. The Social Costs

As we discussed in the previous article, the social costs of immigration can be crime, terrorism, native flight, and the loss of social cohesion, so let’s look at each.

Crime

The most typical complaint I see people make is this:

Source: Tomas Pueyo for Uncharted Territories, with data from Eurostat

Indeed, while the EU population increased by less than 2% between 2014 and 2024, sexual assaults increased by 72%, rapes by 150%, sexual exploitation by 89%, attempted intentional homicides by 47%, and serious assaults by 45%. Violent crime has decreased, though: robberies by 33%, burglaries 15%, and theft by 22%.

So what’s going on?

This chart combines data for five European countries: Austria, Denmark, Finland, Germany, and Italy.

Clearly, immigrants from Muslim countries are overrepresented across most of these host countries. Let’s look at each one in turn.

Germany

These crime statistics were just released for Germany in 2025. This is how much more likely foreigners are to commit crime vs locals:

I took total crime rates per country of origin from the German Police here (table T62, there are many, it’s confusing). I translated with Claude, but checked the numbers and they work. Note that people with double citizenship German-something else are counted as German, so this likely undercounts the crime rate of foreigners. I took the nationality of foreigners from the German stats organization DeStatis here (also confusing, the right one is 12521-0002, and then you need to toggle to citizenship). Both are for 2025, so this should account for recent changes. I took first the per capita rate of crime for 4 different measures of crime: overall, murders, sexual crimes, and violent crimes. Note that overall includes the others, and that violent crimes include murders and some sexual crimes. I then took the ratio of each of these rates of crime to Germany’s rate. Then I took an average of the four. So this is basically taking overall crime per capita relative to Germans, but overweighting violent crime, sex crimes, and murder.

Algerians in Germany are 26x more prone to crime than the average German. 13 other Muslim-majority nationalities are at least 5x more criminal than the average German. Only one non-Muslim majority meets that requirement (Georgia).

If you break the numbers down by pickpocketing, sex crimes, murder, and overall crime rate, this is what you get:

An interesting fact is that Sweden appears in the Top 20 murdering nationalities in Germany. I dug into this and found 6 murders by Swedes in Germany in 2025. I tried to find out more details but couldn’t. What I did find, however, is this report from Sweden about Swedes attempting murder in Germany. They were of immigrant background and planned to do it in the name of the Islamic State. We’ll look at Swedish crime later, but essentially, this suggests that these are immigrants with Swedish background.

As you can see, green (Muslim) is at the top of all the rankings, with special mention for the winner, Algeria, with the brutal record of one crime for every three Algerians every year. Algerians are ~1500x times more likely to be a pickpocket than Germans, to murder at 35x the rate of Germans, to commit 23x more crime overall, and are 8x more likely to commit sex crimes.

Your first reaction might be: Yeah but that’s not because they’re Muslim; it’s because they’re poor. That’s a good thought to have:

Both of these axes are logarithmic.

Indeed immigrants from poor countries tend to commit more crime, but there’s plenty of confounders: human capital, culture, education, age, sex, labor-market access… Some poor countries don’t commit any crime (eg Madagascar), and other countries with the same GDP per capita commit dramatically more crime (eg, Central African Republic).

So is it more the money, or the human capital from the country? A recent Finnish study gave some random families $600/month for a couple of years and measured the impact on crime. It was 0%. This paper showed that the link between inequality and crime is basically zero. More inequality doesn’t mean more crime.

The trendlines on the graph suggest that immigrants from Muslim countries commit on average +50% crime vs those of other countries, but that might be because of more conflict in the origin countries (eg, Syria, Afghanistan, Somalia), more asylum seekers vs economic workers, the fact that a majority of them are men… Conversely, Indonesian and Malay immigrants tend to be Muslim too, but their crime rate is extremely low.

So both income and religion probably are part of the picture, but they’re definitely not the entire picture.

Denmark

Apparently, Kuwait and Lebanon here really mean Palestine, but since it’s not a recognized country in Denmark, immigrants from there can’t be labeled Palestinian. Source.

Here, they corrected for age and sex of the immigrants.:

Source

As you can see, this does reduce the crime rate of immigrants vs the baseline, but doesn’t change the overall picture. This tells you that receiving young male immigrants, especially from crime-prone countries, is a way to dramatically increase the crime rate.

This Danish report showed that a few hundred Palestinians were accepted into Denmark in 1992 under a special law. Of them, about 64% had criminal records: 21% ended up in prison, and 43% got at least one large fine.² About 59% were on welfare. Of their children, 34% have been convicted of a crime, 13% received a prison sentence.

Here’s a breakdown per type of crime:

The Danish Parliament looked at the share of men born in 1987 who had been penally convicted of a crime by the time they were 30, and this is what it found:

Sweden

Sweden doesn’t have much data, but I was able to find some from the official Swedish National Council for Crime Prevention.

In Sweden, the share of foreign-born people has grown from 0.7% in 1990 to 20% today, of which about a third are from Muslim countries. Source for the relevant table.

Spain

Spain is interesting because, unlike most other European countries, it attracts a much more diverse mix of immigrants.

However, in Spain, four of the top five most crime-prone countries of origin are Muslim.³

Source: Tomas Pueyo for Uncharted Territories, based on data from the Informe General 2024 de Estadística General Clasificada de Población Penitenciaria de la Secretaría General de Instituciones Penitenciarias, via this

Finland

Here’s what I found for Finland.

Via this, which has a paywall so I couldn’t view it, but it was shared here

Italy

And for Italy:

Source

Norway

In Norway:

UK

The UK doesn’t have much information, but a freedom of information request to the London Metropolitan Police delivered this data:

Another Freedom of Information request to English and Welsh police provided this data:

I’m only showing the top 40 countries. 25 of them are Muslim. 9 of the top 10 are, too.

Netherlands

The Netherlands publishes crime data but not per country. There, non-EU immigrants are 4.5x more crime-prone than the Dutch:⁴ They are 15% of the population but commit ~40% of the crime.

Note that this problem does not seem to exist in the US. There, the share of Muslims is quite small, and most criminals are American (more likely to be criminal than the average immigrant). As you can see, wherever we look in Europe, there’s an overrepresentation of Muslim countries at the top of the crime lists. Controlling for age, sex, and income reduces this effect, but it remains strong. Immigrants in general commit much more crime than locals. Within immigrants, non-EU immigrants are much more crime prone. Within those, Muslim immigrants cause the most crime.

Terrorism

Terrorism data is harder to get because there’s less of it. But we do have some. Around the world, about 75% of terrorism deaths are caused by Islamism.

I asked Claude to look into the Global Terrorism Database and give me the topline results of Islamist terrorism. It told me that 75% of global terrorism deaths in the world in 2023 were caused by Islamism, which has killed 250,000 people in 163 countries, with just four Islamist terrorist organizations accounting for the majority: Islamic State, Hamas, JNIM, and Al-Shabaab.

Here are terrorists in the US for the last 30 years, by type of terrorism.

Data from here, transcribed with Claude.

Here are security threat prosecutions in Germany in 2025.

In the UK, Islamism accounts for 67% of attacks since 2018, 75% of terrorism investigation cases, and 90% of convictions.

So I think we can all agree that the biggest threat in worldwide and European terrorism is Islamism, and it’s so stark that the 2^nd most significant threat doesn’t even matter.

Social Cohesion

Several papers have found that ethnic diversity reduces neighborhood trust and social capital, including this meta-analysis, although the linchpin might not be diversity alone, but also segregation: If a certain ethnicity is concentrated in a small area, neighborhood trust and social capital will shrink.

Looking specifically at Muslims, this paper found that 2^nd generation Muslim immigrants had lower trust than their non-Muslim counterparts, and that this was due to cultural transmission rather than discrimination or exclusion.

This graph looked not at the crime, but the feeling of safety in different communes in Denmark. It then compared it to the share of people of MENAPT (Middle East, North Africa, Pakistan, Turkey) origin.

This reflects the idea of lower trust in more diverse areas. Via this. Note the Y axis though: It’s not like trust goes to zero, but the drop is substantial.

There’s a real cost in social cohesion that is paid by immigration, whether that is due to Muslim immigration or that of others. In Europe, however, it’s more acute with Muslims.

Alright, let’s move to economics.

2. The Economics

Employment

Some Anglosphere countries manage to get immigrants to work more than natives, but that’s not the norm. In most countries, their unemployment is higher—in Sweden, foreign unemployment is an astonishing 2.5x the native level!

The bottom of this list represents the countries where unemployment is lowest for immigrants. They’re all part of the Anglosphere (US, Australia, Ireland, Canada, UK).

This immigration has many different origins, but when we zoom in to Muslims, the data holds (here, employment rate rather than unemployment rate).

Those who are not immigrants are mostly 2nd generation immigrants, so these numbers capture the long-term unemployment of immigrants.

One reason for these differences is regulatory. The US makes it easy for foreigners to start working immediately—they even penalize them if they don’t. Others make it very hard, and might even support not working by giving welfare—in some cases, at higher levels than to locals. This is important in the short term, but also the very long term, because the first months after arrival are crucial. If you establish yourself in a system where you have to work and you must meet and socialize with locals to do so, you’re more likely to be integrated and continue working. But if you can’t work, you’ll stay more with family and other compatriots, and you’ll get used to a lifestyle where you don’t have to work.

Another reason might be education. US Muslims are more educated than the average American.

Source

In part, this is because the US attracts high-education Muslim immigrants.

So that’s the destination effect. But the origin effect also matters.

We see similar data whenever we zoom in to specific countries.

Norway

Source

We can see a breakdown of the problem from some data from Spain.

Spain

In Spain, where the vast majority of African immigrants are Muslim, their employment rate is substantially lower than that of locals or Latin American immigrants.

Source

This is driven basically by the fact that Muslim women generally stay at home

This is the activity rate, which counts those with a job or looking for one. What if we look at actual work rates?

Both men and women immigrants work less than natives. Muslim men work more than non-Muslim foreigners. But only 30% of Muslim women are employed!

The type of work also matters. The education rate of MENA (Middle East and North Africa) immigrants in Spain is substantially lower than that of locals or other immigrants.

UK

In the UK, Muslims are 2.2x less employed than the average of the population.

And unemployed actually hides the true impact, because many who don’t work don’t appear in these statistics, if they’re not actively looking for a job.

Source

Muslims also tend to have lower-paying jobs:

Source

Among other reasons, because they have less education:

Source

We see the same impact of lower female participation:

Source

And this data shows this is not an effect of age. In the UK, Arabs and Pakistanis are substantially less employed than any other ethnicity, and the gap for women is astonishing.

At their peak, only 50% of Arabs in the UK are working!

Germany

Germany is interesting because the data is slightly different. We again see female Muslims with a very low employment rate. But men are slightly more likely to work than immigrants of other nationalities.

Source

On balance, it looks like the two big contributors to lower work participation rate are immigration and, within that, Islam.

France

In France, the unemployment rate of African immigrants is more than twice that of locals, and it gets worse in the 2^nd generation.

~80% of African immigrants to France are Muslim

The shares of social housing (so, welfare) remain extremely high for the children of immigrants, and poverty rates remain quite high, too.

US

In the US, Muslim immigrants are more educated than the average American. However, they tend to be somewhat underemployed compared to Americans. The difference is not massive though.

Unemployed means no job but looking for one.

OK so what we’re seeing for Muslim immigrants is that those in Europe tend to have lower education than natives, and they also work less. About 50-70% of the men of working age work, and the number for women is usually about half of that. For the US, I don’t have more granular data, but it doesn’t look like these hold.⁵

Now, let’s look at state contributions: net costs vs benefits in terms of taxes contributed vs spent.

Fiscal Cost-Benefit

Denmark

Denmark has some of the best data on this topic, so let’s look at it. I shared this before:

Somalia, Syria, Lebanon, Iraq, Afghanistan, ex-Yugoslavia (some Muslims), Morocco, Turkey, Pakistan, Bosnia (half Muslim), Iran and Sri Lanka immigrants are all net negative tax contributors in Denmark.

Because of this, every year Denmark spends 1.3% of its GDP in supporting MENAPT immigrants—about half the public budget spent on education. This is just to support these immigrants fiscally. This graph from The Economist shows the breakdown per age, compared to other types of Danish people:

Before adolescence, natives and immigrants in Denmark cost the state approximately the same amount. Then they start diverging. Between roughly 25 and 68 years old, Danes and Western immigrants contribute positively to the state through taxes. After retirement, the contribution is increasingly negative. The hope is that the earning years generate more money than the cost years (so the surface above the “0” line in the middle years is higher than the surface below it as children and old people). Non-Western immigrants don’t contribute as much, but the most expensive in Denmark are those from the Middle East, North Africa, Pakistan and Turkey, as well as their descendants. Source.

At no age do MENAPT immigrants contribute positively to the state. If you break that down per country of origin:

MENAPT are not the only immigrants to cost the state, but they cost the most. Every Somalian⁶ costs $27k per year!

Netherlands

This is the most massive study I’ve seen for any country on earth on the fiscal impact of immigrants. In it, the Netherlands shows fiscal contribution per country of origin (like in Denmark), but over their lifetime instead of per year, and for 1st and 2nd generation immigrants. Here it is, shown on a map of the world:

Lifetime net fiscal contribution of 1st and 2nd generation immigrants to the Netherlands. Note that many eventually repatriate to their origin countries, which reduces their tax burden. For immigrants who remain, the cost across both generations is approximately double. Source.

And here’s a detail per country/small region:

An average native-born Dutch family pays five thousand euros in taxes per year that end up being spent on immigrants.

A surprising fact: For the 2^nd generation of immigrants, only those of Danish, Swedish, Finnish, and Chinese background were net positive. Every other origin drained state coffers rather than contributing to them.

Why? Here’s the share of benefit recipients vs the likelihood of them leaving within 10 years of arriving:

Source

The most likely to ask for welfare support are the least likely to leave.

Finland

This is for Finland:

Here’s a breakdown per country:

Portugal

The data that we have for Denmark, Finland, and the Netherlands isn’t as readily available for other countries. Portugal has data per person per country, but only tax contributions, not welfare costs.

Here, we can’t see Muslims, because only 0.4% of the Portuguese population is Muslim. They don’t tend to immigrate to Portugal.

So if we summarize the economic side, Muslim immigrants to Europe, compared to locals, tend to:

• Work less

• Have lower levels of skill

• Consume much more welfare

Takeaways: What Should the EU Do?

The takeaway so far is that in Europe, immigrants:

• Work less than the average native.

• After locals, the next most productive are EU citizens. The least productive tend to be Muslims

  • This is in a big part because of women not working, but not entirely.

• Immigrants are nearly always fiscally expensive.

  • Muslim immigrants cost in general more money to the state.

  • This cost isn’t erased at the 2^nd generation.

• Immigrants commit substantially more crime than any other group.

  • Across destination countries, Muslims commit the most crime.

• Islamists are behind most of the terrorism.

• Immigration reduces local trust and social capital, probably more due to segregation than other factors, but also because of cultural transmission.

  • Again, it seems like this is more true of Muslim immigrants than others.

Does that mean that massive waves of deportations should take place in Europe? No:

1. Europe needs immigrants.

2. Many countries are quite good at taking in immigrants.

3. Many immigrants take several generations to converge to local performance.

4. The numbers above are not complete. They’re missing some other factors.

5. A large part of the problem is caused not by the immigrants, but by European legislation.

6. Not all immigrants are the same. They should be treated differently.

These are the things we’re going to look at in the next article.

1. The exceptions are Spain and Portugal, which get more from Latin America, and the countries that received a lot of Ukrainians, like Poland. Data from Eurostat.

2. One or more, as repeats were not counted.

3. In the case of Albania, it’s a plurality, not a majority. Spain appears to have a serious problem with Albanians. They account for only 0.1% of foreigners, but 2.5% of criminals in prison.

4. In the Netherlands, the biggest sources of immigrants are Indonesians and Surinamese. I don’t know the baseline crime rate of Surinamese, but for Indonesians it’s very low in Germany, so I assess it’s very low in the Netherlands too. The next two groups are Turks and Moroccans. Given their track record in other countries, it’s reasonable to estimate they have a high crime rate in the Netherlands too.

5. Immigrant Muslims have more education than the average American, and they work just a bit less, which suggests both men and women Muslim immigrants work more than in Europe.

6. Usually included in the definition, even though it belongs to neither North Africa nor the Middle East. But it’s close by and also majority Muslim.