We have already responded to many comments on Twitter and updated the earlier copy of the research. The latest version can be downloaded here. The Cato Institute and others have raised new attacks, which we respond to below (including responding to their responses). We have also added a response to Cato’s response and will continue to update. Among the comments already addressed in the updated version:
1) A concern is that illegals will commit crime as they pass through Arizona on their way to other jurisdictions. But if the census estimates pick up those who are temporarily in Arizona, this effect shouldn’t matter. Even though there may be different people in Arizona in June than in December, the total population is all that matters for the purposes of our calculations. However, it may be more difficult to count people who are only passing through Arizona. The Census’ small confidence intervals reflect recognition of this as a serious problem.
If this theory is correct, the size of the effect should vary with the intensity of enforcement on the Arizona-Mexico border. A more porous border should cause undocumented immigrants to make up a higher share of newly incarcerated inmates.
A section at the end of the paper tests this “porous border” hypothesis.
2) There is some concern about the accuracy of the listed citizenship statuses of criminals. In particular, some Hispanics are listed as being born in the US but are not listed as citizens. There are some explanations for this result, but it’s worth examining what happens to the crime rate when we simply remove these cases from the dataset. An earlier version of the paper found that this makes only a small difference. We have since added a more detailed discussion, especially in Appendix 2.
3) The question was raised of what would happen to the US citizen crime rate if naturalized citizens were omitted. Again, this had little effect on the results.
Responses to new comments:
Yesterday afternoon, the Cato Institute’s Alex Nowrasteh posted these comments on what he calls the “Fatal Flaw” in our research. He then provides his own estimates, using ICE detainers that he claims show a low rate of crime by illegal aliens.
Cato then blasted out an email to all congressmen and their staffs, and to who knows how many other people across the country. For a month, Dr. John Lott asked Nowrasteh if he had any comments. He had been given the “do file” so that he could check Lott’s calculations as well as the codebook. Nowrasteh might have saved himself some embarrassment if he had conferred with Lott. Below, we will also analyze some comments from others.
Nowrasteh uses detainers from ICE to estimate whether someone is likely to be an undocumented immigrant.“In June 2017, only 38.3 percent of criminal aliens had ICE detainers on them and, thus, were more likely to be illegal immigrants. As a back-of-the-envelope estimation, I assumed that 38.3 percent of “non-U.S citizens and deportable” are actually illegal immigrants in the ADC’s larger 1985-2017 dataset. This back-of-the-envelope calculation turns Lott’s finding on its head. Whereas he found that 11.1 percent of the admissions to Arizona prisons in 2014 were illegal immigrants, the real percentage is a maximum of 4.3 percent, below the 4.9 percent estimated illegal immigrant share of the state’s population.”
Detainer orders are hardly a reflection of changes in crime rates by illegal aliens, and Nowrasteh’s work could be reasonably interpreted to show the opposite of what is claimed.
Nowrasteh ignores Lott’s results showing that the number of incarcerated criminal aliens is disproportionately large compared to the noncitizen share of the Arizona population. Total criminal convictions for incarcerated criminal aliens (documented and undocumented immigrants, as well as temporary visa holders) equals 12.5% of all convictions and 11.9% of prisoners.
Take the 12.5% estimate first. If only 38.3% of criminal aliens had ICE detainers against them, Nowrasteh would have to infer that about 4.79% of convictions are for undocumented immigrants, while the other 7.7% must be for documented immigrants and temporary visas holders. Given that undocumented immigrants and temporary visa holders only makeup about 3.9% of the population, this group’s share of incarcerations would be 98% greater than their share of the population.
A similar analysis is possible for the 11.9% share of first offenses. In this case, documented immigrants and temporary visas holders make up an 88% greater share of incarcerations than their share of the population.
Now, as we will explain in the next section, we don’t actually believe that documented immigrants are committing crimes at high rates. These documented immigrants and temporary visas holders are also relatively old compared to both U.S. citizens and undocumented immigrants (see Table 2), so their age-adjusted crime rates are even proportionally higher.
That variable is broken down into seven categories. Lott erroneously assumed that the third category, called “non-US citizen and deportable,” only counted illegal immigrants. That is not true, non-US citizen and deportable immigrants are not all illegal immigrants. A significant proportion of non-U.S. citizens who are deported every year are legal immigrants who violate the terms of their visas in one way or the other, frequently by committing crimes.
According to the American Immigration Council, about 10 percent of people deported annually are Lawful Permanent Residents or green card holders—and that doesn’t include the non-immigrants on other visas who were lawfully present in the United States and then deported.
We estimated that undocumented immigrants accounted for 11.8% of prisoner convictions in Arizona, and that undocumented immigrants make up about 4.8 percent of the population. If 10.5% of the people that we have listed as undocumented immigrants are actually documented immigrants or temporary visa holders, that reduces undocumented immigrants’ share of convictions leading to incarceration from 11.8% to 10.6%. That would still imply an undocumented immigrant conviction rate that is 121% higher than their share of the population. But let’s be cautious and assume that rather than 10.5%, the reduction should be 21% in Arizona. Undocumented immigrants would still make up 9.32% of convictions and their share of convictions would still be 94% greater than their share of the population.
Again, remember that as the percentage of Arizonans who are undocumented immigrants falls, the percentage of documented immigrants and visa holders increases. So the ultimate question is whether Nowrasteh believes the calculations that he posted yesterday are the “real” results.
UPDATE: Nowrasteh has tried to respond to the points raised above. Nowrasteh is trying to have it both ways. The data show that either legal immigrants have twice the incarceration rate, or illegal immigrants do, or a combination. For reasons discussed in my response, only a small 10% fraction if any of the one category that Nowrasteh focuses are likely to be here legally. He spends a lot of time explaining that his estimates shouldn’t be taken seriously because they are only “back of the envelope” (BOE) estimates. But the problem with his BOE is a simple one: in trying to make extreme assumptions to show that the crime rate by illegal aliens is low, those same assumptions (if you believe them) imply that crime rate by legal immigrants is very high. Will Nowrasteh admit that if you believe his BOE assumptions, the incarceration rate by legal immigrants is twice their share of the population?
Nor does he attempt to respond to either the impact that relying on pre-sentencing report or that even if he is somehow correct the effect is trivial.
— On February 1st, Nowrasteh posted another comment on Twitter:
However, Table 1 [in the Lott paper] makes even less sense now. Table 1 has 20,645 total prisoners sent to AZ state prisons in 2014. But according to BJS, there were only 14,634 prisoners admitted to AZ state prison in 2014. Can u explain the ~6000 prisoner discrepancy?
After we got this question, we reached out to the BJS and received this answer on Monday. As we suspected, the BJS data had left out some cases from that particular Table because they had included those numbers elsewhere and didn’t want to have double counting or to count people who had since escaped from prison.
The tool includes AWOLs/Escapes and transfers, but the report states that those are not included in the report table.
Obviously, neither of those issues are relevant to our research. We are still counting crimes, even for those who have escaped before their term was over. And while the BJS is concerned about double counting, if someone was previously in incarcerated in another jurisdiction, we are still interested counting them for crimes that they committed in Arizona.
— On February 5th, I got a comment from Tom Wong, an Associate Professor of Political Science at UC San Diego, former Advisor to the White House Initiative on Asian Americans and Pacific Islanders under the Obama administration.
“In Table 1, they’ve counted each row in the data as a unique individual. This is incorrect, as this means that the same person is counted multiple times (one can confirm this by a simple tabulation by race without removing duplicates)And looking at the number of times each “IDNO” number appears by “commit_number_for_inmate” shows that ADOC is cleanly counting individuals in the data who they encounter multiple times, however”
Does the ultimate question concern how many crimes people committed or whether the number of people who have ever been to prison? Both are useful questions. The first part of the study examines the first question. Later the issue of recidivism is addressed, and that directly answers this second question. The finding is very clear that there is a significant, large group of U.S. citizens in prison who keep committing crimes over and over again.
The previous comments regarding Now’s claims are also relevant for evaluating other claims by Wong.
More will be added later as we have time.