There exists a widespread impression, deliberately fed by some conservative media and sometimes by President Trump himself, to the effect that illegal immigrants are especially prone to crime, and to violent crime. By extension, by illogical implication, immigrants in general are tarred with the same brush.
Immigrant crime agitates for different reasons that are not always disentangled from one another. First, there is the general social disorganization that any crime causes and the worsening of the quality of life it entails. Second, mostly youthful immigrants could have the power to reverse the general decline in crime that accompanies the aging of American society. They might be the agent of a step backward for American society. Third, capturing, trying, and incarcerating anyone is very expensive, more expensive than schooling, for example. Immigrant crime in general is especially apt to disturb emotionally because it seems to involve a kind of crass ungratefulness: I let you into my living room, or you enter while I am not paying attention, and you show your appreciation by stealing the silverware, and worse.
Belief in the criminality of immigrants as a group is not necessarily the result of a kind of emotional prejudice. Immigrants are predominantly young, ill-educated, and poor, all known ingredients of criminal propensity in any context. (To my surprise, current immigrants are not predominantly male, although maleness is a strong factor of criminality.) So, if immigrants to the US are like just about every population of the same age, income, and education ever studied, they should exhibit a higher crime rate than the native-born population that is, on the average, better educated, more prosperous, and, especially, considerably older.
It’s difficult to figure out the basic truths about immigration and crime because gross miscounting by partisan journalists is common. For instance, in his 2018 article in Reason on a report based on 2009 figures, Alex Nowrasteh shows how easy it is to make horrendous but simple mistakes of enumeration: Counting single events of incarceration as if they were individual immigrants, for example, as if illegal immigrants could not be repeat offenders. (“Restrictionists Are Misleading You About Immigrant Crime Rates.” Reason, Feb. 1, 2018.)
I explicitly do not accuse the authors cited below of such miscounting.
In his 2006 article in Liberty, Cox showed that about 2.6 % of inmates in federal prison and an astonishing 12% of people incarcerated in local jails and prisons were illegal immigrants in 2002. That’s 14.6% of all persons then incarcerated in the US. The highest estimate of the number of illegal aliens I could locate for any year is 15 million. With that estimate, conservatively, incarcerated illegal aliens would be about 5% of the then US population of 293 million. Roughly, illegal aliens, according to Cox, were thus incarcerated in 2002 at almost three times the rate of their occurrence in the general population. This rate did not include incarcerations for merely being illegally present in the US (which must have been a small number since that was only a misdemeanor). Cox’s study is based on figures for only one year for the whole country.
At any rate, Cox’s incarceration figures concern mostly illegal aliens. It’s not clear whether illegal immigrants’ propensity to commit crimes is similar to the corresponding propensity of legal immigrants, nor if their crimes are similar. Legal and illegals may come from different countries and regions. Even if they come from the same places, they may issue from different classes in their societies of origin. Even if from the same places and same classes, they may constitute different samples of the populations of origin from the standpoint of motivation and thus, of personal psychology. It takes different virtues to arrange for legal immigration via whatever path, on the one hand, and to swim the Rio Grande, or coolly to overstay one’s visa, on the other. These different virtues could easily be associated with different levels of different criminal tendencies. Finally, legal and illegal immigrants have different incentives to break the law or not, the latter being in a good position to not draw attention to themselves. That’s at least until the sanctuary movement.
A study published by the libertarian Cato Institute in February 2015 examined criminal conviction data provided by the Texas Department of Public Safety. It found that native-born residents were much more likely to be convicted of a crime than immigrants in the country either legally or illegally. For all crimes together, the legal immigrants’ score was less than one third that of the native-born. The difference in the likelihood of being convicted of homicide, specifically, was very large between legal immigrants and the native-born. The former were 15 times less likely to be convicted of homicide. Even illegal immigrants were only 70% as likely as the native born to be convicted of homicide. (“Two charts demolish the notion that immigrants here illegally commit more crime,” Christopher Ingraham, Washington Post on-line accessed circa June 28 2018.)
Note that the Cato Institute study is for one state only and for only one year. Cox’s figures cited above are also for one year only but for the whole country.
It’s not obvious how one should relate these contradictory sets of findings to one another. (There are many more such studies. I chose two researchers on my side of the political fence that seemed to me to have worked with seriousness.) First, figures about comparatively rare events such as homicide are notoriously unstable. The corresponding homicide figures for 2004 or for 2014 might be very different. Moreover, both sets of figures, Cox’s and Cato Institute’s use the heterogeneous categories “legal immigrants” and “illegal aliens.” To generalize from their findings requires making the silent assumption that the composition of both is stable from year to year. This assumption is unwarranted. Nothing regulates the composition of illegal immigration and little insures that the composition of legal immigration will be similar from year to year. The varying numbers of refugees alone could sway the legal figures one way or the other.
Here is a realistic scenario: For a period of a few years, both immigration flows consist mostly of rural, mountain Mexicans from rural areas where crime is scarce. In a subsequent and contiguous period, a large flood is added, through both refugee legal immigration and through illegal immigration, of urban Central Americans (thus, of people from some of the highest crime areas of the world). Both the frequency and the nature of immigrant crime may change swiftly as a result of this sudden (and realistic) change in the composition of immigration, legal or illegal, or both. The composition of legal immigration may change drastically in a single year because of the influx of refugees from a single location hitherto unrepresented in the US. The composition of illegal immigration may also change suddenly because of a disaster in a region that the American federal government does not recognize as a legitimate source of refugee status. It’s hazardous to extrapolate from one period to any other period. Hence both Cox’s and Cato Institute’s findings may be correct but, at the extreme, each of them for one year only.
It’s also risky to extrapolate from one domestic location, for example, Texas, to another, for example, the whole United States. Here is one reason among others why it is so: The (innocent) rural mountain Mexicans I mention above are likely to move to the Central Valley of California and to similarly agricultural areas in Florida. Crime-prone Central Americans, on the other hand, may seek their fortune in the more familiar big cities anywhere, including, Los Angeles, Chicago, and Houston. In this imaginary scenario, immigrant crime in Texas (Houston and Dallas, per chance) may be grossly unrepresentative of immigrant crime anywhere else in the US.
Finally, as Cox pointed out to me in a personal communication, the comparison category, “native-born” is itself heterogeneous with respect to crime. The rates for African-American – most of whom are native-born – are several times as high as others’. Perhaps, the native-born would far better for incidence of crime if blacks were excluded from the comparison. I suspect this is true but I don’t know according to what theoretical principle, this exclusion should be made.
[Editor’s note: In case you missed it, here is Part 5]