Just for the record I do not believe that illegal aliens have any particular predisposition to violent crime and have never made that claim. I have said that homicides perpetrated by illegal aliens would not have taken place but for the presence of the illegal alien, something that seems self-evident to me. Homicide is not a random occurence. It requires a perpetrator and without the perpetrator it would not have taken place.
But the study described in this New York Times article, however well-intentioned, is junk science:
A lot of research has shown that there’s no causal connection between immigration and crime in the United States. But after one such study was reported on jointly by The Marshall Project and The Upshot last year, readers had one major complaint: Many argued it was unauthorized immigrants who increase crime, not immigrants over all.
An analysis derived from new data is now able to help address this question, suggesting that growth in illegal immigration does not lead to higher local crime rates.
In part because it’s hard to collect data on them, undocumented immigrants have been the subjects of few studies, including those related to crime. But the Pew Research Center recently released estimates of undocumented populations sorted by metro area, which The Marshall Project has compared with local crime rates A lot of research has shown that there’s no causal connection between immigration and crime in the United States. But after one such study was reported on jointly by The Marshall Project and The Upshot last year, readers had one major complaint: Many argued it was unauthorized immigrants who increase crime, not immigrants over all.
The clearance rate for violent crime in the United States is around 45%, that is to say arrests are made for 45% of violent crimes that are reported to authorities. But the situation is even worse. According to the FBI only about 37% of violent crimes are ever reported to the authorities. That means that we have absolutely no idea of who committed more than 80% of all violent crimes and making any claims of any variety about who committed them is more than the available data allows. We certainly don’t have enough data to draw any conclusions about year-to-year variance.
The study in question is dependent on a very commonplace fallacy: false precision.
I don’t know what proportion of violent crimes are committed by illegal aliens. No one does. I don’t believe they’re responsible for a larger proportion than they comprise of the population and could even be responsible for a lower percentage. But that’s a belief, not something I can prove.
“In 2012, Mexico had a murder rate of 21.5 per 100,000 population. There were a total of 26,037 murders in Mexico in 2012.”
“According to a 2013 report by the United Nations Office on Drugs and Crime (UNODC), between 2005 and 2012, the average homicide rate in the U.S. was 4.9 per 100,000 inhabitants compared to the average rate globally, which was 6.2.”
Any analysis pretty much has to address this particular elephant.
The FBI has written about this and notes that they believe the reporting rate for homicides is pretty close to 100%. The clearance rate is about 60%. We dont know for sure who committed the other 40% of those murders. So, for example, we think most homicides are committed by men, but maybe all of the unsolved murders are by women. So assuming that the gender make up of the unsolved murders, by your standards, is junk science.
“Any analysis pretty much has to address this particular elephant.”
Violent crime in border cities is lower than other cities.
https://www.cato.org/blog/crime-along-mexican-border-lower-rest-country
Most of the murders in Mexico are drug related. The homicide rate in Tijuana is 140. In San Diego it is 2.
https://www.latimes.com/world/mexico-americas/la-fg-mexico-tijuana-drug-violence-20190130-htmlstory.html
Steve
Violent crime is not synonymous with homicide. Every woman who get beaten up by her husband (or vice versa) does not report it to the police.
Inductive logic (which is what they’re using) requires that you be able to prove uniform distribution or a relation between the 39th percent and the 40th percent. They do not meet that standard; what they’re doing is a misuse.
“Violent crime is not synonymous with homicide.”
Nope, but since reporting is at almost 100% it acts as a useful proxy. Since they dont really find any differences in rates for other violent crimes, it suggests that the lack of reporting is probably not a big factor.
“Inductive logic (which is what they’re using) requires that you be able to prove uniform distribution or a relation between the 39th percent and the 40th percent. ”
Which is hard to do since the population is illegal and so many crimes are not solved. However you can still have sufficiently powered studies and you can look at the results of other studies for confirmation. That said, all this study (or report as they call it) did was look at crime rates from 2007-2016, then graph that against metro areas by percentage of illegals. Areas with a higher percentage of illegals did not have more reported violent crime. A perfect study? No Science? Probably not. Useful? You bet, especially because it confirms what other studies find. Doing a perfect study on this would be incredibly expensive and will never be done. (You should try managing some research budgets for a while.)
Steve
The proper form of the argument is that low-income, low-skilled immigrants increase violent crime, because low-income, low-skilled men in the United States commit most of the violent crime.
In the U.S. almost all of the low-income, low-skilled immigrants come from Mexico and Central America. That’s a product of proximity and third world economies.
If a study controls for demographics, it cancels out the effect. It’s quite possible that an adult Mexican male with a 7th grade education making minimum wage is no more likely to commit a crime than a similar American male, but it doesn’t mean we should encourage their migration.
PD- There are a few weak points in that argument, but it is not a bad one. However, the reason we are discussing reports like this one is because of the repeated claim that immigrants are violent criminals. Instead it looks as though they are no more likely, probably less, than Americans to be violent criminals.
Steve
“I have said that homicides perpetrated by illegal aliens would not have taken place but for the presence of the illegal alien, something that seems self-evident to me” is a specious argument. Alcohol poisoning wouldn’t take place without the presence of alcohol, but banning it is not a solution, as we found out at our great expense.
Sorry, but I just don’t trust official crime statistics. IMO the amount of violent crime committed by illegal immigrants is significantly underreported, because much if not the majority of it is committed against other illegal immigrants, who are understandably reluctant to report such crime to the authorities.