Conservatives Are Lying on Immigrant Crime

One of the things that has always driven me crazy about the left is how much they lie and twist the facts on the issue of crime. Liberals in my mind have always been the people who whine about blacks being stopped and arrested more than whites, while ignoring crime disparities between the races. They hate racial profiling despite the fact that all police work is profiling, and race is a major predictor of criminality. Leftists in the media and academia spread falsehoods about police shooting unarmed black men as they ignore the orders of magnitude more violence committed by blacks against people of all races. I once heard some online rightists talk about “blood libel” in this context, and although I’m allergic to most forms of rightoid hyperbole, this one seems to fit. Blaming a group of people for a problem that it is not responsible for is something that should be shamed.

That said, as I’ve been watching the Trump campaign and the conservative movement mobilize around the idea of immigrant crime as a serious threat, I’ve started to feel the same kind of revulsion towards ugly lies that I’ve traditionally felt towards analogous leftist narratives. In the case of lies about racist law enforcement, not only is this perspective wrong, but it creates an inverse reality, since it is poor communities that most need police protection. Similarly, conservatives are not simply mistaken on immigrant crime. They are spreading a narrative that makes society dumber and actually obscures the positive impact of new arrivals in this one area of life.

Consider that the central issues of Trump’s campaign are crime and immigration, which he keeps telling us are intimately related. A recent rally in Georgia is representative, when he warned that gangs of Venezuelans were going to take over the entire state of Colorado and then said this,

Our country is dying. Our country is dying. We’re taking all the criminals from all over the world… The crime rates for countries all over the world is way down because the citizens of their countries have demanded it. They’re taking the criminals out and they’re moving them and dumping them into the United States of America. And their crime rates are cut in half and less than that. And their jails, which were packed in many cases are close to empty, and very soon, will be totally empty because they’ve taken their prisoners and they’ve dumped them into the United States of America. We’re like a dumping ground and we’re not going to allow it. We’re bringing them back. We’re bringing them back.

Trump’s language on the topic appears to be getting more graphic, with him telling an audience that migrants “are people at the highest level of killing that cut your throat and won’t even think about it the next morning,” and also the types to “grab young girls and slice them up right in front of their parents.” His campaign of course does not provide evidence for any of this.

Now there’s always a tendency to try and sanewash Trump. When he’s saying things that are obviously false and crazy, someone comes along and informs us that he’s actually getting at some truth in a deeper sense. In this situation, the argument might go that he’s not talking about all immigrants, but specific groups of migrants in particular locations.

The first thing to say is that this style of debate is poisonous for public discourse. I’ve always hated this game where people or the politicians they support say objectively false things and then retreat to an idea that is more defensible when they get called out on it.

Again, there are a lot of parallels here with what liberals do on crime. I remember during the unrest in Ferguson, after the media had to admit the shooting of Michael Brown was justified, they changed the subject to things like a lack of diversity in the local police force and blacks in the city being disproportionately stopped. The Department of Justice under Obama cleared Officer Darren Wilson but also sued the city of Ferguson for supposed civil rights violations. One can debate the question of whether the city was governed in a racist way, but those who push the narrative of white cops shooting unarmed black men should be denounced and shamed for saying untrue things regardless. Lies like this increase racial tensions and make it more difficult to have any serious discussion on topics like race and crime. If you hate when your opponents do this and can wax eloquently on why their transparent lies are bad, you shouldn’t want people on your own side to do the same.

Like those who talk about unarmed blacks being hunted down, on immigration and crime Trump is saying things that are plainly untrue. Conservatives got upset at the ABC moderators fact checking Trump on his claim that the crime rate was up, but the FBI statistics were just released, and the media was right while Trump was wrong. And it’s not surprising that an increase in migration has not led to an increase in crime, as it’s been known for decades that legal migrants commit much less crime than natives, and the differences are quite large. Illegal immigrants are about equal to natives or slightly higher, but since legal immigrants compose the majority of all arrivals, the net impact of immigration on crime is to lower it.

If your argument is that Trump is just talking about Venezuelans in Colorado and Haitians in Springfield, these are tiny populations, so this is a dumb thing to base a presidential campaign on. For Trump’s entire message of American carnage to make sense — and this is a message he’s consistently pushed since first running for president in 2015 — it has to rely on the idea that immigration in the aggregate makes the country more violent. This is true whether you take what the man says literally or what is implied by the narrative that he is running on. Trumpism and its perspective on crime simply have no validity or coherence in a world where immigration makes the country less dangerous in the aggregate, even if the same cannot be said when analyzing every population individually.

In 2007, Butcher and Piehl used Census data that surveyed 5% of households in 1980, 1990, and 2000 to determine the institutionalization rate for native-born Americans and those who were born abroad. This includes all people in prison, but also those in mental institutions, hospitals, drug treatment centers, and long-term care facilities. The data did not differentiate prisoners from the other categories after 1980, but analyzing that year showed that 70% of the institutionalized population was incarcerated, and using incarcerated numbers alone didn’t change the results.

Overall, they found that the foreign born were much less likely to be institutionalized in each decade.

This is true despite migrants being less educated, younger, and more Hispanic than the population at large. If you control for these factors, the foreign-born population looks even better.

I asked Jason Richwine of the Center for Immigration Studies about this topic, and he pointed me to the section “Problems with Census Data” in this report as a rebuttal to Butcher and Piehl. It argues that people in prisons might not be willing to honestly out themselves as noncitizens because it might open them up to deportation. Also, in 2000, 53% of the time the question of whether a prisoner was native or foreign born was imputed or assigned, in part because prisoners’ records were often filled out with administrative data.

But a closer look at the underlying government report (p. 298) addressing the response rate issue shows that there is nothing unique here about the citizenship or foreign born questions.

While responses for citizenship and place of birth were imputed or assigned at rates of 53% and 54% for prisoners in 2000, the numbers for questions like residence five years ago, mobility disability, and veteran status are even higher. Moreover, imputed or assigned data seems to have been less of a problem in 1990, though respondents giving false information still might have been.

We can’t even be sure that the Census Bureau and prisons do a poor job of guessing whether a person was born abroad. False positive and false negatives might be equally likely, balancing things out. One could imagine the Census data even overestimating the foreign-born population in prison, as recorders assume that everyone who looks or talks like someone outside the mainstream falls into that category, say a native Hispanic who grew up in a barrio and speaks heavily-accented English. It’s more difficult for me to imagine many foreigners being mistaken for someone who is native born, since everyone who arrives in a new country after a young critical period tends to spend their whole life speaking with an accent.

Alternatively, administrators or the Census Bureau might just classify prisoners as native-born citizens as a default when they don’t have solid information. All of this is just speculation, and the CIS critique of the Census data in 2000 is strong enough that it indicates that perhaps we shouldn’t count Butcher and Piehl as conclusive. Luckily, a more recent paper from Abramitzky et al. helps settle the issue. It finds that immigrant men, meaning all those born abroad, have been more law abiding than natives for 150 years. This paper uses full-count Census data from 1870 to 1940, and then Census/American Communities Survey samples for 1950-2020.

The following chart shows incarceration rates for men among native-born Americans and first-generation immigrants from different regions of the world, all the way from 1870 to 2019.

There are two reasons that I think the results of this paper can be trusted, despite arguably having many of the same problems one might point to in Butcher and Piehl. First of all, the longer time period makes it unlikely that the data has always been wrong in the exact same way every year. In recent years, you would need to be underestimating the number of foreign-born prisoners by a factor of two to three to make immigrants even equally as criminal as the native born, and you would need to be doing this consistently all the time. The American Communities Survey might have the same problem as the Census, but it may not, and they both have to be wrong in a similar way for immigrant crime to be a major issue.

Another piece of evidence presented by Abramitzky et al. that cuts strongly in the direction that immigrants are a low crime population is the fact that they have much higher rates of employment, marriage, and good health than natives of similar levels of educational attainment.

For US-born males who are high school dropouts, for example, the labor force participation rate is slightly over 60%, compared to about 90% for the foreign born. Crime is mostly committed by the poor and uneducated, and we have a great deal of evidence showing that among that population, immigrants are simply better across the board.

Everything we know about the way the world works suggests that a group that is more likely to be married, healthy, and working is probably going to have a low propensity towards crime.

Just for fun, here are some charts from the Abramitzky et al. appendix showing incarceration rates for the top twenty sending countries compared to natives in the years 1980, 1990, 2010, and 2019.

The data is about what you would expect, with Asian migrants the least criminal, then Europeans, and then black and Hispanic newcomers. All of this passes the sanity check.

What about illegal immigration? This is somewhat harder to study. The papers above take the entire foreign-born population, which includes legal immigrants, illegal immigrants, naturalized citizens, and even those on temporary visas. The undocumented are one subset of this larger population, and we can decide to focus on them specifically to be able to have a response to the people who say they love legal immigration, it’s just people coming illegally that they have a problem with.

Texas is the only state that checks the legal status of prisoners. Earlier this year, Richwine used data from that state to argue that illegal migrants had relatively high rates of conviction for murder and sexual assault. Yet the murder difference is quite small. Between 2012 and 2019, taking the average of all years, illegal immigrants have a 13% higher murder conviction rate than the population as a whole. Now, it must be noted that there is no control for age and sex in the data. Given that illegal immigrants skew younger and male, even Richwine’s numbers indicate that they kill people at similar or lower rates than the native born when controlling for relevant factors. The gap in sexual assaults is larger, yet again without controls it’s difficult to say how much we should worry about that.

Alex Nowrasteh has argued that Richwine’s analysis is flawed because he supposedly double counted illegal immigrants in prison due to how the data from Texas was made available. Richwine has claimed he didn’t do this. I’ve talked to both Richwine and Nowrasteh about this dispute, and each has told me that he has had more direct contact with the Texas Department of Public Safety than the other and that his numbers are clearly the correct ones.

I started to look into the question and finally decided that it’s too much trouble to dig any further because not that much is at stake given the differences between them. For the sake of argument, I’m going to accept Richwine’s data saying that illegal immigrants have a 13% higher murder rate in Texas on average over a ten year period, rather than Nowrasteh’s most recent data saying that the number is 36% lower.

I asked Richwine about the lack of controls for sex and age, and he replied that if young males commit more crime, then this is a relevant factor to consider regarding the question of how upset we should be about illegal immigration. I dislike this argument a lot, as it implies we should try to become a nation of old ladies because then we would finally feel safe. Nobody ever uses this logic for any issue other than immigration.

If we learn the fertility rate has gone up, it’s generally celebrated as good news, and we never hear anyone say that this is something that we should worry about because it indicates we’ll see more crime in twenty years. That might be true, but the benefits of having a more youthful society are so overwhelming that the influence on the crime rate is never thought to be worth mentioning. This idea that making your population more male and younger increases crime, and this is something worth thinking about for public policy, independent of how crime prone any group is relative to the population, only exists in the context of fights over immigration. In other words, using the standards we use in every other debate, of course we should control for age and sex when talking about immigrant crime.

The Richwine/Nowrasteh debate comes down to whether illegal immigrants are slightly more or significantly less murderous than the general population in Texas. Yet controls for age and sex would bring both of their numbers down, so that even if we take Richwine’s data as correct, illegal immigrants are about equally violent as native-born Americans or even slightly less violent. Worrying about “immigrant crime” as a general matter is stupid, but one can’t make one’s concerns look much more rational simply by focusing on the undocumented alone.

In summary, all of this is to say that we have overwhelming data that legal migrants commit a lot less crime, and illegal immigrants are perhaps comparable to natives, for a net effect of immigrants lowering the crime rate. By blaming migrants as a group for increasing crime, Trump and other Republicans are simply lying. Crime isn’t even going up, and when it does, it is practically always the result of changes in criminality disproportionately among those already highly represented in the criminal justice system. There’s no moral distinction to be made here between right-wing lies about immigration and those of leftists who talk about the dangers of white supremacist violence.

The smarter rightoids tend to be racialists, and they have a few standard responses to the data on crime and immigration. First, a lot of them say that the native born population includes blacks, and they commit so many crimes that it distorts the data. To which I say, what’s your point? If you are part of a movement that wants to argue that immigration increases crime, you don’t get to subtract a group just because it helps you make your case. Blacks are part of the American population. So when migrants move into a city that is majority black, they reduce the crime rate. It’s a very weird objection, and it makes me wonder what is going on when people make this argument. I think that their reasoning goes something like this: “Richard Hanania made an anti-racist point. But I’ll make a racist point by saying native-born blacks are more criminal. My racist point cancels out the anti-racist point so I win.” Or something like that. Many on the right seem more committed to maintaining racist vibes than making logical arguments.

Perhaps it’s just muscle memory. Immigration restrictionists are used to arguing with liberals, who don’t like to admit that blacks have a high crime rate. You can therefore “own the libs” by pointing to black crime as a response to the fact that the immigrant crime rate is low. But you can’t own me because I admit the truth about black crime!

The thing is though that even if you use just native white males as a comparison, immigrant men of all races are currently about 30% less likely to be incarcerated according to Abramitzky et al. When all native men are included, the number goes up to 60%. There’s no good reason to arbitrarily exclude blacks, but even if you do, nativists are still wrong on immigrant crime.

Another argument they make is that to determine the impact of immigration on crime rates you need to look at the second generation and beyond. Largely, what this means is that we have to take into account the ultimate effect of becoming a more Hispanic country. As we can see, Hispanics born in the US have an incarceration rate higher than that of the nation as a whole.

In addressing this argument, we again have to start by acknowledging that Republicans’ entire messaging here is based on lies. Trump is running on the idea that immigration today is making America more dangerous. Not having a problem with him lying because it gets at some larger truth about demographics makes one no different than leftists who say it’s fine to exaggerate and mislead people about police shootings because it opens up a larger conversation about racism. Lying doesn’t only tend to distract from issues, but it makes discourse difficult because each side can justifiably look at the other and say these people say things that are blatantly false, so I can dismiss their arguments completely.

Regarding the substance of the racialist argument, we can say that, unlike Trump’s messaging, it is based on some actual facts. What it in effect says is let’s forgo the lower crime rate and economic benefits that immigration brings today based on what we think will happen 15 or 20 years down the line. Yet this perspective takes an unusually long-term view, and if you are really that worried about crime, trying to engineer the perfect demographic composition of the population a generation from now is probably not the best thing you could be doing with your time.

All around the world, we see interventions that make the crime rate fall precipitously on much shorter timescales. El Salvador under Bukele is the most famous recent example, but there are other interventions that may be undertaken in a democratic context that can also have a major impact. A 2021 paper by Jennifer Doleac and two co-authors investigated the effect of a 2005 law in Denmark that increased the likelihood of criminal offenders being included in a DNA database from 4% to almost 40%. They concluded that DNA registration reduces the probability of a new conviction by 42% within the first year, and about 31% after three years. Doleac found a similar impact across American states based on when they expanded DNA databases. She argues that this kind of surveillance is much more cost-effective than most other law enforcement tools.

If one simple reform can reduce recidivism by 30-40%, then if you’re worried about crime you should talk a lot about DNA databases. Consider the table above showing that the incarceration rate difference between native-born Hispanics and all Americans is about 13%. If you stop immigration today, maybe you reduce the Hispanic population by say 5% in 20 years, and you bring down the overall crime rate in 2044 by (0.13 * .05) * 100= 0.65%. Congratulations, you saved America. This is even setting aside the fact that most people in Washington who oppose Hispanic immigration also don’t want Asians around either, and will find themselves excluding whites too even if that is not their intention. Nativists getting their way means that the crime rate goes up, today and likely in the future, and this must be taken into account when deciding whether to cheer them on. But even if they limit themselves to excluding Hispanics, you’re still not going to impact the crime rate all that much, and even if you get the effects you want it won’t be for a long time.

Instead of playing these games, and likely shooting yourself in the foot, you can simply listen to Jennifer Doleac, do the stuff that she wants, and completely obliterate urban crime. And you can start today, instead of waiting twenty years to see the results of your work.

What we increasingly see is that technology makes living with crime a choice. In addition to DNA databases, facial recognition technology and shot spotter have emerged as effective tools of law enforcement, yet they face challenges in being implemented for the usual reasons related to charges of racism. And no, it’s not Hispanics or other recent immigrants who stand in the way of these kinds of technologies being adopted more widely, but native-born leftist activists and black political leaders.

Some restrictionists argue that we should only care about the raw number of crimes that immigration causes, not crime rates. This is obviously ridiculous. One thing that needs to be mentioned in these debates is that immigrants are not only sometimes the perpetrators of crime, but also victims. And the ratio between the numbers in each category has to matter for our analysis if we are concerned about the safety of natives. From that perspective, this probably makes the pro-immigration case stronger than what one would come up with just from looking at immigrant crime rates alone.

To see why, let’s say that you have a population of 100 people, and 5 of them are criminals. Each has a single victim, so there are 5 crime victims. Then 20 immigrants enter the country. They have the same crime rate, so 1 immigrant, representing 5% of the population, is a criminal. If this individual victimizes a member of his own community, then you’ve added one perpetrator and one victim through immigration. You had a 5% victimization rate for natives before migration and nothing has changed for them after 20 new arrivals joined the population.

Now imagine the exact same scenario, but one of the native criminals victimizes a migrant instead of a fellow native. The native population you started out with now has 5 criminals but only 4 victims. For them, the chance of being robbed or killed falls from 5% to 4%. By the same logic, it’s possible that a group has an even higher crime rate than natives but still has a null or even positive effect on their safety. A group might just victimize itself, while also becoming the victims of native criminals, thus in effect shielding native-born Americans from the impact of crime.

Do we have reason to suspect that immigrants are more likely to be victims than perpetrators? I believe so, based on the fact that they are highly concentrated in major cities, around high crime populations while they themselves have low crime rates.

In Chicago, for example, immigrants make up about 21% of the city but 36% of the entrepreneurs. I bet if you looked at store owners in the most crime-ridden neighborhoods, the numbers would be even more lopsided. I used to drive between the south suburbs and the University of Chicago, and sometimes would need to stop at a gas station in one of the more dilapidated parts of town. Every single time the business owner would be Arab or South Asian. Like in other Midwestern cities, it is a normal occurrence for migrants working in these areas to be robbed or killed by natives. I’ve never heard of an immigrant robbing and killing a native-born American working behind the counter at an inner-city gas station, but it’s probably happened at some point.

There is regional variation in terms of which kinds of immigrants tend to start businesses in our major ghettos and therefore end up as victims of native crime, though the story is largely the same across the country. Tensions between Korean shop owners and black residents of Los Angeles came to national attention during the 1992 Rodney King riots. The year before, Ice Cube released “Black Korea,” in which he complained about “Oriental one-penny countin’ motherf*ckers” following him down the aisle when he’s trying to buy a drink, threatening that if they don’t show enough respect to local blacks “we’ll burn your store, right down to a crisp.”

In recent years, a great deal of the urban unrest we’ve seen has been set off by conflicts between black youths and immigrant store owners. Michael Brown was shot by a white police officer after robbing an Indian-American owned liquor store. In the moments before he died, George Floyd ended up in a confrontation with police because he tried to pass off a fake $20 bill to a Palestinian-American business. Even outside of the context of convenience stores and gas stations, immigrants living disproportionately in urban areas means that they are often more likely to fall victim to crime than other Americans.

I can’t find data specifically addressing this question, but the idea that immigrants are more likely to be victims than perpetrators lines up with everything we know about differential crime rates between groups and where various populations tend to live. All of this means that immigrants therefore not only reduce the crime rate in the US by simply having a low crime rate, but they do even more than the numbers would suggest to drop the victimization rate for natives, in effect shielding native-born Americans from crime by absorbing much of the dysfunction of the inner cities. If Arabs, Indians, and other new arrivals didn’t own convenience stores, grocery stores, and gas stations in violent urban areas, there probably just would be fewer such businesses in the first place, but the ones that did exist would be staffed by native-born Americans who would themselves suffer at the hands of shoplifters and armed robbers. And just by walking the streets of cities like Baltimore and Chicago, immigrants add to the population of potential mugging victims much more than they do to the population of muggers. The same applies regarding second-generation immigrants who might stay in these communities and provide a buffer between the violent urban cores of our cities and the rest of the population.

There’s a larger point here about how immigrants have shielded Americans not only from the consequences of crime, but much of the dysfunction that has come out of the American inner cities since the 1960s. They do this in part by being willing to risk their lives taking jobs that are too dangerous for most native-born Americans, and also pushing the politics of our cities to the right by their voting and activism. But the story of how immigrants saved America — in their roles as urban residents, business owners, and voters — by mitigating much of the damage of 1960s liberalism will have to wait for another time.

In the end, I think we all know that most people who oppose immigration don’t really care what the numbers say. They dislike people who are different, and don’t want them around. They can’t be honest about this, even with themselves, so they have to talk about things like crime. If you show they’re wrong on crime, they’ll make some economic argument, and if you knock that down, they’ll move to culture or their arbitrary preference. They could save everyone a lot of time by just starting with the identitarian argument. But the reason that politicians like Trump tell lies about topics like immigration and crime is that nobody is convinced when you just say you don’t want people around because they are different.

Maybe there’s no changing the minds of the most committed nativists, but they have the potential to mislead people who aren’t paying that much attention, which is why it’s useful to go into how and why they are wrong on individual issues.

Out of necessity, the nativist right often ends up making dumb arguments, which is why educated elites tend to reject them. Ann Coulter’s ¡Adios, America! The Left’s Plan to Turn Our Country into a Third World Hellhole (2015) has a chapter called “Immigrants and Crime.” She begins by assuring us that no one knows whether immigrants commit a lot of crime, because a fear of being called racist prevents researchers from looking into the question. Yet, as we’ve seen, this is simply not true. Never mind that we’ve somehow ended up with tons of data on race and crime in the United States despite the left not liking the results, and we also know that migrants to Europe are much more criminal than natives.

She then goes on to do things like point out that the most wanted criminals list of the LAPD doesn’t include any Anglo names, and that the incarceration rate in the US is much higher than it was in 1965, when immigration laws were liberalized. She ignores anything that doesn’t fit her narrative that immigration is bad, like the fact that the sharpest increase in crime occurred before immigration took off and crime has declined since the 1990s as immigration has been higher. This should be obvious to Coulter, since she spends much of her time in California, where the foreign-born population keeps growing and the murder rate is now one-third of what it was in 1993.

One more funny note on Coulter is that in her book she claimed that the official count of around 11 to 12 million illegal immigrants in the US was wrong, and that the real number was about 30 million. As Nowrasteh points out, if this were true, then estimates of illegal immigrant crime have to be significantly adjusted downward. When analysts try to calculate the crime rate for a group, they use a simple formula with the number of arrests, convictions, or incarcerations as the numerator and their total population in the denominator. If Coulter was familiar with the data on immigrant crime, she would perhaps have realized that her claims about the topic contradict what she says about overall population estimates.

Coulter is pretty smart and a good writer, but she’s not a data person, and it’s clear that she simply doesn’t want foreigners around and is going to grasp at whatever she can to make her case. I’m spending a bit of time on taking her arguments apart because she was someone I used to trust on the topic of immigration, before realizing that the facts simply didn’t back up what she was saying.

Leftists want to believe white racism is what causes the problems of black people, so will believe any accusations, studies, or anecdotes that tell that story. In a similar way, most rightists who talk about race are either explicit or implicit white nationalists. They’ll be on solid ground when talking about Europe or black crime, but when it comes to immigrants to the US the data is not on their side, so we come across many of the same tricks we see leftists use when they don’t have good arguments.

Nativists are usually correct when they say that they are tapping into majoritarian sentiments. Surveys show that Americans believe that immigration makes crime worse, despite the evidence, and Trump has found his messaging on this issue effective. But public opinion is often self-contradictory, and while it must be taken into account it can’t alone determine what is good policy. The masses like foreigner bashing, but if Trump’s full nativist economic program was implemented, our living standards would collapse and the public would turn against him. Stephen Miller tried to deter migrants from crossing the border by separating children from their parents, but that policy caused an outcry, was widely unpopular, and had to be discontinued.

Basically, Americans want less free trade and migration but also a strong economy. They want to stop illegals from crossing the border but aren’t cruel enough to support truly effective deterrence, at least if it makes its way to their TV screens. Voters are in the end irrational and want mutually contradictory things. The proper role of Elite Human Capital is to decide which goals are humane and noble, and which attitudes and preferences should be ignored or even suppressed.