[Part I; Part II; Part III; Part IV; Part V; Appendix; Single page]

by David Severa

• OPENO: Let’s move from welfare to crime then.
• RESTRICTES: So far this discussion has been quite bloodless, focused on statistics and not emotion. That’s appropriate for abstract economics questions, but crime has an ineradicable emotional component that you have to understand, just like terrorism can’t be understood merely by measuring lives lost. Ultimately we have to decide public policy rationally, but that includes rationally understanding people’s emotional, sometimes irrational reactions. So let me open with one example, Rotherham, to show why ordinary people are so worried. In Rotherham, England, over a period of 16 years gangs of ethnic Pakistani men sexually abused over 1,400 children. Girls were “doused in petrol and threatened with being set alight, threatened with guns, made to witness brutally violent rapes and threatened they would be next if they told anyone.” All of this was covered up for years by officials for fear of offending immigrant communities. This wasn’t some paranoid conservative fiction, it happened. And I’m not even getting into the horrific details, I just want you to understand in some way the visceral reaction people have. People read about this and they think of their own kids. Crime isn’t just statistics to people, it’s locked doors and empty playgrounds and an inescapable miasma of mistrust. It’s quite difficult to quantify these additional costs, but as we discuss numbers, please remember that the penumbra of crime covers all of society, not just individual victims.
• OPENO: I agree that not everything can be fruitfully reduced to mere numbers, but-
• RESTRICTES: Here’s a number for you: 10,000 soldiers in France guarding Jewish sites earlier this year. Imagine having to send your children to a school in that environment.
• OPENO: -But numbers can relieve unfounded worries. Both of your examples were from Europe, which isn’t surprising. In fact, research in America has mostly found that immigrants have a lower crime rate than natives. In Europe and elsewhere, the opposite is true.
• RESTRICTES: I think that “However America is doing, Europe is doing worse” is becoming a theme. But doesn’t a lower crime seem implausible? After all, Latin America is one of the most violent regions in the world. I could accept similar or slightly higher…
• OPENO: It depends in part on how you control for demographics.
• RESTRICTES: I hope you don’t mean controlling for education, income, or employment. You can’t treat those as some exogenous variables! If we want to look at how much crime is actually going to be committed, you can’t just assume away that immigrants differ from the rest of the population.
• OPENO: True, though that does have its uses in answering other questions. But in this case I meant that immigrants are disproportionately young men, the most criminal group. 1.6% of immigrant men age 18-39 are imprisoned, vs. 3.3% of natives, but there are more immigrants in that group. Since their children will not have similarly skewed demographics, it does make sense to adjust in that way, depending on the question you’re answering. And our discussion is about long-term immigration effects, not short-term ones. (Though even without making those adjustments, it isn’t clear that immigration increases crime. Remember that the immigrant population has surged while crime rates have dropped by half since the early 1990s.)
• RESTRICTES: And does this demographically adjusted low crime rate carry over to future generations?
• OPENO: Not exactly. Second generation immigrants have higher crime rates than their parents, but not hugely so. Apparently they acculturate to the surrounding environment and commit crimes at about the same rate as natives. It isn’t clear to me whether they’re adjusting for socioeconomic factors. I don’t think so, but I think they are lumping high and low skilled immigrants together which is effectively the same thing. So if a group of immigrants is poorer/less educated and the poor/less educated are more criminal, then the group will be more criminal, but not relative to other comparable groups. In that case the long term prospects for the crime rate are tied to future economic assimilation.
• RESTRICTES: So it depends on whether that 80% convergence holds.
• OPENO: Correct. Hispanics, most of whom are descended from relatively recent immigrants, are incarcerated at 1.8 times the rate of non-Hispanic whites, but less than a third the rate of blacks. That puts them very near the overall national average. Even if economic convergence doesn’t improve then the crime rate will not be much affected by immigration.
• RESTRICTES: It still seems very implausible to me, given how crime-ridden Latin America is. I mean, tens of thousands of people have been killed in Mexico’s drug war.
• OPENO: Well, the drug war is an aberration, I think, due to the proximity of a big rich country and a big middle income country. It’s exacerbating what was an already high murder rate, but isn’t itself some deep part of Mexican culture. I mean, it used to be Colombia, now it’s Mexico… Also immigrants aren’t a representative sample of their home country, so we shouldn’t be surprised that there are differences.
• RESTRICTES: Perhaps. I will say again, even if this is all true America could be doing better by only taking in educated high skilled immigrants. We don’t have to settle for keeping the crime rate steady when we could be lowering it. That immigrants aren’t always as harmful as I’d expect doesn’t mean our current policies are optimal. Let’s avoid false dichotomies. And I’m not going to let you get out of talking about Europe.
• OPENO: I wouldn’t dream of trying to escape.
• RESTRICTES: Like you said, most studies in Europe have found a link between immigration in crime. Unfortunately, much of this research is out of date or it lumps all immigrants (sometimes including intra-EU migrants!) together. So let’s look at incarceration rates to get a better idea of how things are actually going. In France, Muslims make up 12% of the population, but between 60 and 70% of all prisoners. In Britain it’s 3% vs. 11% in prison. And so on, throughout European countries with significant Muslim immigration. (Note that there aren’t good official statistics, unlike in the US. Hmm…) By comparison, in America blacks are 13% of the population, but 38% of all state prisoners. Hispanics are 17% vs. 21%. Europe has rapidly recreated – exceeded even – America’s huge racial disparity in crime. (And don’t forget that prisons are a breeding ground for extremism.)
• OPENO: That’s disturbing, but remember that America locks up way more people, so those numbers aren’t necessarily directly comparable. For instance, France’s incarceration rate is less than a seventh than that of America’s, so even if the disparity is real, the levels are still much lower.
• RESTRICTES: Here we run into problems with the data. I’d like to give you varying homicide rates (generally the best for cross-country comparisons) for different groups, but France and Germany willfully don’t collect those sorts of statistics, and the UK uselessly lumps all Asians together. (It’s easy enough to find information on xenophobic hate crimes though!) In Germany, what data there is suggests that more recent refugees have higher crime rates than earlier migrants. It’s possible that what’s happening is that, as in the US, first generation immigrants have lower crime rates than their kids. However, the first generation has a similar crime rate to natives, rather than lower, so their children are dragging the crime rate up.
• OPENO: But from a lower base rate than in America.
• RESTRICTES: Right. It isn’t clear what’s going to happen to crime rates, or how high they might go if the immigrant-descended population continues to grow. I wouldn’t guarantee that rates will stay below America, but the data just isn’t there to make confident predictions. Though the mere fact that the data is so sparse is to me very suggestive that there is a serious problem being covered up. Not to suggest a conspiracy or anything like that, but the difference relative to the US is quite striking. However, the prison population is proof enough that immigration has increased crime significantly. And even if a high crime rate like America’s isn’t high enough to tear apart society, why should Europe give up the peaceful societies it has taken centuries to create? Remember all the secondary effects of higher crime I mentioned.
• OPENO: Perhaps…
• RESTRICTES: I’d also like to point out that while immigration in general doesn’t increase corruption, immigration from corrupt countries does. And, of course, Middle Eastern and Latin American countries are much more corrupt than those in the West.  Government corruption is a more important sort of crime than is generally recognized, since government can work its tendrils into practically every aspect of life. Anecdotally, think of the Bell, California scandal (a city that’s 93% Hispanic). I think this just confirms people’s intuitions.
• OPENO: Well, one study found no correlation between officials convicted of corruption and the Hispanic population*, but I’d like to make a broader point. Everyone knows that successive waves of immigration from Europe in the late 19^th/early 20^th centuries fueled the rise of corrupt machine politics in America. The machines have gone, but the corruption often lingers. (Illinois governors are literally more likely to go to jail than Illinois murderers.) Gangs and organized crime were obviously concentrated among immigrant groups. Catholics and Eastern Europeans were never going to be able to adapt to democracy, or to learn English. The poor would depress native wages. Etc., etc., etc. For a long time, many of these concerns seemed well-founded. And yet, almost nobody regrets these past waves of immigration. Eventually something changed. Assimilation happened. As we’ve discussed, it’s never been absolutely total, but it’s definitely been enough. When you look at past waves of immigration on a large enough time scale, you see success. (And a century or more is the appropriate length.) So we can look at all the numbers we have, and try to peer into the dim future, but trends change and the past is an equally important – if only intermittently quantifiable – guide. Maybe this time is different, but maybe not.

* It did find that increasing ethnic diversity led to greater corruption, but I saw other studies that found no effect.

• RESTRICTES: The fact that there are such differences in outcomes between high and low skilled immigrants, and between low skilled immigrants to the US and Europe suggests that we have no guarantee that past examples of migration will map well onto current ones. And most past large migrations have been mass folk wanderings defined by violence and displacement. The experience of Europeans migrating to European colonies may be very much an exception. There aren’t enough examples to confidently analyze the causes of successful immigration and assimilation. It’s only theorizing and speculating. Even high skilled immigration probably has risks; there’s just so much we don’t know.

Next: Cultural assimilation

(Source for header image)