Steven Malanga, a senior fellow at the Manhattan Institute and a senior editor of City Journal, argues that there is a crime wave caused by immigrants in his recent piece, “No, You’re Not Imagining a Migrant Crime Spree.” This is not accurate and the article overall is not a useful source for those interested in understanding illegal immigrant criminality.
Malanga’s piece leans heavily on rhetoric and individual cases of migrant crime and doesn’t give nearly as much attention to migrant criminal conviction, arrest, or incarceration rates as he should have. He begins his piece by recounting the tragic murder of Laken Riley in Georgia and a political debate that ensued when President Biden mentioned the murder in his 2024 State of the Union address. Malanga then brushed aside the best empirical evidence on illegal immigrant criminality when he wrote:
The elite press rode to Biden’s defense. The idea of a migrant crime wave was a myth, media outlets proclaimed, noting studies of Texas incarceration data from years ago, which seemed to suggest that illegals commit crimes at low rates.
What are those “studies of Texas incarceration data from years ago”? They don’t exist because those studies are of Texas criminal conviction and arrest rates by immigration status, not incarceration rates.
Researchers and the media focus on Texas state crime data because it’s the only state that records criminal convictions and arrests by immigration status. We’re fortunate that Texas does because it’s a border state with the second highest population of illegal immigrants after California, and it’s governed by Republicans, which short-circuits the common conservative argument that you can’t trust crime data reported from Democratic jurisdictions. The data are also more recent than Malanga makes it seem—although they weren’t available when Riley was murdered.
After her murder, I wrote a study of criminal convictions in Texas by immigration status focusing on homicides that includes 2022 data. My study shows that illegal immigrants have lower homicide conviction and arrest rates than native-born Americans but higher rates than legal immigrants.
The Texas data are pretty good but they contain some hidden landmines, as I have explained here. The most recent Texas data from 2022 were not recent enough for him, but a piece on the supposedly high rate of illegal immigrant gang membership published 20 years ago made the cut.
In 2022, native-born Americans had a homicide conviction rate of 4.9 per 100,000 in Texas (Figure 1). In other words, 4.9 native-born Americans were convicted of homicide in that year for every 100,000 native-born Americans living in Texas. In the same year, the illegal immigrant homicide conviction rate was 3.1 per 100,000 illegal immigrants living in Texas. The legal immigrant homicide conviction rate was the lowest of all at 1.8 per 100,000.
There were 1,336 people convicted of homicide in Texas in 2022. Native-born Americans were convicted of 1,209 of those homicides, illegal immigrants were convicted of 67, and legal immigrants were convicted of 60. Homicide conviction rates for illegal immigrants and legal immigrants were 36 percent and 62 percent, respectively, below those of native-born Americans in Texas in 2022.
The results of the above analysis of Texas crime data were criticized by the Center for Immigration Studies (CIS). Malanga cites CIS’ criticisms in his piece but he does not cite my responses. The major disagreement is that I maintain that CIS double-counted some illegal immigrant convictions based on how they requested the data. To be clear, this was not intentional on their part. CIS disagrees.
Furthermore, these data are presented without any controls for age, sex, race and ethnicity, and socioeconomic status (SES). Even if CIS is correct and my data analysis is wrong, the difference between illegal immigrant and native-born criminal conviction rates for homicide is small. Introducing normal controls for age, sex, race and ethnicity, SES, or any combination would surely produce a far lower illegal immigrant criminal conviction rate, as Richard Hanania points out. Although introducing controls would be best practice, it’s not necessary to have such controls because I’m happy fighting in the methodological landscape crafted by the nativists at CIS.
Malanga faults pro-immigration analysts and advocates for countering “that significant immigrant wrongdoing can’t be going on, as crime rates are now falling.” He’s correct about that. The foreign-born share of the US population is just shy of 15 percent, so fluctuations in immigrant criminality would have to be extreme to drive the nationwide crime rate. That’s why it’s important to look at illegal immigrant and immigrant crime rate data where they are available, which brings us back to his erroneous dismissal of the Texas criminal conviction data.
On this point, Cato also published research estimating legal and illegal immigrant incarceration rates. Our latest estimate of these is for the year 2018 (we’re updating it now), and there are methodological challenges that expose it to criticism. But the findings are in line with other research on immigrant criminality.
Relative to native-born Americans, we found that illegal immigrants were 41 percent less likely to be incarcerated and legal immigrants were 74 percent less likely. But Malanga didn’t have to rely on Cato’s research (although he definitely should have); he could have looked at a wonderful working paper by Ran Abramitzky, Leah Platt Boustan, Elisa Jácome, Santiago Pérez, and Juan David Torres that took a long-term look at immigrant criminality in American history. They find:
We provide the first nationally representative long-run series (1870–2020) of incarceration rates for immigrants and the US-born. As a group, immigrants have had lower incarceration rates than the US-born for 150 years. Moreover, relative to the US-born, immigrants’ incarceration rates have declined since 1960: immigrants today are 60% less likely to be incarcerated (30% relative to US-born whites). This relative decline occurred among immigrants from all regions and cannot be explained by changes in immigrants’ observable characteristics or immigration policy. Instead, the decline is part of a broader divergence of outcomes between less-educated immigrants and their US-born counterparts.
Abramitzky and his coauthors don’t estimate illegal immigrant incarceration rates and their data only go through 2020, but the figure from their paper is striking.
Malanga’s more fundamental point is that there’s an immigrant crime wave in the United States. He recounts individual high-profile crimes like the murder of Laken Riley and others, but those don’t tell us whether there’s a crime wave—regardless of how tragic those examples of individual crimes are. Only crime rates, incarceration rates, criminal conviction or arrest rates, and similar measures can tell us whether there is a crime wave.
Furthermore, US-wide crime rates can’t tell us the source of the crime wave as Malanga writes. That’s why we need the detailed evidence above on immigrant criminality relative to native criminality to figure it out.
Crime rates are more important than relying on individual cases of crimes committed by immigrants because they allow us to see whether there is more crime controlling for the number of people, which is important because more people generally lead to more crime. Crime rates tell us whether Americans are more at risk of being a victim of crime; they tell us whether certain subpopulations are more or less likely to commit crime than others; and they tell us whether some jurisdictions are more dangerous than others. The criminality of different subpopulations can indicate where higher or lower crime rates are coming from.
As I wrote earlier this year, “the focus on crime rates matters when discussing the relative criminality of different groups and evaluating whether immigrants bring more crime than they add people to the United States.”
Malanga doesn’t spend much time discussing crime rates, criminal conviction rates, incarceration rates, or other similar measures over time, which makes it impossible to judge whether we’re living in the middle of a crime wave largely or partially caused by illegal immigrant criminals. The recent crime surge began in May 2020 after the murder of George Floyd, which was also the month with the second-lowest number of illegal immigrant border apprehensions in a very long time. We’re constantly told that Trump controlled the border in the preceding years, which is exaggerated for political purposes, but there were still many fewer crossings and apprehensions (“encounters” are what they’re called now) than after Biden was sworn in. Odd timing for an illegal immigrant crime wave, to say nothing of subsequently escalating border encounters and falling crime rates. So, where were these criminals born? Overwhelmingly in the United States.
Malanga writes that ICE removals of illegal immigrant criminals are down compared to 2019, although they are up since the last month of 2020 during the pandemic when Trump was president. Another wrinkle is that ICE has released fewer criminals under the Biden administration than during the Trump administration. There were 182,870 book-ins of criminal migrants into ICE detention in 2020, which rose to 211,450 in 2021, rose by another 100,000 in 2022, reached 273,220 in 2023, and was on track for more in 2024. The above administrative actions are complex and paint a confusing picture.
Regardless, migrant crime cannot be a major driver of nationwide crime rates and small changes in ICE book-in or removals policy can’t explain much of the variation in crime rates because the numbers are so small. Again, this is why focusing on illegal immigrant criminality measured by criminal conviction, arrest, and incarceration rates is so valuable.
Malanga then mentions data from the State Criminal Alien Assistance Program (SCAAP), which allows state governments to apply to be partly reimbursed by the federal government for incarcerating illegal immigrants. Before he mentions SCAAP data, Malanga mentions other surveys of federal multistate data that supposedly “show a far more troubling reality.” He provides no link to these surveys, but I assume he means the SCAAP data.
Malanga leans heavily on a report from the Federation for American Immigration Reform (FAIR) that analyzes the SCAAP data and finds that illegal immigrants have a much higher incarceration rate than other populations. I criticized FAIR’s interpretation of the data here and here. In the years analyzed, illegal immigrants were about 6.5 percent of Texas’s population, 4.7 percent of arrests, 3.9 percent of criminal convictions, and accounted for 3.6 percent of all days incarcerated. The patterns are similar for other states.
The recent letter from ICE to Rep. Tony Gonzalez (R‑TX) on the 662,556 noncitizens convicted of crimes or with ending charges also makes an appearance in Malanga’s piece. He says that these criminals aren’t being deported, but that’s because they are on ICE’s non-detained docket. That mostly means the migrants are currently incarcerated, deceased, from countries where they can’t be deported because of the lack of a treaty, or for other reasons. It’s not as if the Biden administration just decided not to deport them to create a crisis that undermined the Democratic Party’s electoral chances.
Malanga links to a piece by Peter Kirsanow about a GAO report on SCAAP criminal aliens. Kirsanow makes the same error that FAIR did when he assumed that the number of criminal aliens incarcerated was a stock figure. In fact, it was the total number of criminal alien incarcerations—a sum of stocks and flows. In response to Kirsanow’s piece, I wrote that this means “if a criminal alien was incarcerated for 10 short sentences, released after each one, and then re-incarcerated, then that single alien would account for 10 incarcerations under the SCAAP figure for that year. But Kirsnaow counts that as 10 individuals [at a single point in time].
However, when it comes to estimating the incarceration rate of natives, Kirsanow compares the number of individuals incarcerated with their total population [at a single point in time].” In other words, he’s comparing the GAO’s measure of flows plus stocks of criminal alien incarcerations over a long period of time to the native-born stock of incarcerations in a single period. Kirsanow’s analysis is an apples-to-oranges comparison that can only yield a relatively higher criminal alien incarceration rate.
Much of Malanga’s piece recounts individual instances of migrants committing crimes. Almost all the cases he mentions are examples of individuals who should never have been allowed into the United States or who should be removed. But his mention of eight nationals from Tajikistan with supposed terrorist ties is really too much. Eight Tajik men who crossed the US-Mexico border in 2023 and 2024 were arrested in early June 2024 on immigration charges after the government learned they may have had contacts with ISIS or contacts with people who had potential ties to ISIS. There was no evidence to suggest that a specific terrorist attack was planned, no evidence of an imminent threat to the homeland, no terrorism charges have been filed against them, and “ties” is an ambiguous term with little consistent meaning. Zero people have been murdered in domestic terrorist attacks committed by migrants who entered illegally.
Zero Americans have been murdered by foreign-born terrorists in attacks on US soil since President Biden took office, and there has only been one injury in an attack, which was committed by Canadian-born David DePape, who was inspired by his right-wing ideology to attack Nancy Pelosi and instead injured her husband.
In contrast, 12 people were murdered and 38 were injured in attacks committed by foreign-born terrorists on US soil during the Trump administration. It’s certainly unfair to blame Trump or credit Biden for such different records, just as it’s unfair to blame presidents for the number of people with “terrorism ties” who cross the border even when none of them are terrorists.
Malanga’s piece begins and ends by criticizing the “elite press” and “media” for mentioning low illegal immigrant criminal conviction rates or describing Donald Trump’s immigration platform as “extreme.” It wouldn’t be a conservative piece without criticizing the media. But the only reason we know about these cases of illegal immigrants committing crimes is because of reporting by the media and elite press, so let’s give credit where it’s due.
More importantly, press reporting on illegal immigrant criminal conviction rates alongside cases of individual migrants committing crimes is a welcome development. Much of the press breathlessly reports tragedies without informing their readers of the baseline hazard, so I can’t be too upset when they provide that information. My only complaint is that the media should do this when they report every rare event, like the chance of being killed or injured in domestic terrorist attacks, the chance of being killed in a mass shooting, the chance of being unlawfully killed by a police officer, and other hazards.
Malanga and I agree on a few points. First, immigrant noncitizens who commit violent or property offenses should be punished and removed from the United States. Our immigration laws make this more difficult because immigration enforcement authorities waste most of their time trying to block peaceful migrants from selling their labor to Americans. Liberalizing immigration laws would reduce the black market and allow the government to focus on blocking, identifying, and removing migrant criminals. It’s unclear whether Malanga would agree with the last part, but he certainly would agree with the first.
The second area of agreement is on data availability. Every state should keep data on arrests and convictions by crime and immigration status. After all, it’s hard to adjust policy if we don’t know what’s going on. There’s no good reason for American law enforcement to not report these data.
Malanga’s piece ignores significant evidence, pays scant attention to actual crime rates, sometimes misdescribes research, is occasionally unspecific, and focuses overwhelmingly on individual instances of crimes committed by migrants. His piece is not a useful source for those interested in understanding illegal immigrant criminality and what to do about it.