Illegal immigrants - do they commit more crimes?

In response to this post and others - I wanted to find some data. Do correct me if I calculated something wrong or misinterpreted the data in the following:

I found some data here: GAO’s Criminal Alien Statistics report

Page 58 in the PDF (52 in the report) provides the time scope of the convictions data: it’s for 4-7 years (7 years federal data, 4 for state data). Note that they were counting only the data for illegal aliens, since, as they explained, “The data we obtained represent a portion of the total population of criminal aliens who may be incarcerated at the state and local levels, since by statute SCAAP does not reimburse states and localities for certain criminal aliens, such as aliens with lawful immigration status”.
Page 27 in the PDF (21 in the report) shows various crimes and absolute numbers for them. Homicide shows 25,000 for that time period (!).

So - let’s very generously take 7 years as the time period - the bigger of the two periods of measurement in the report, even though the majority of homicide convictions in the US are state-level, so it really should be closer to 4 than to 7. But just to be fair, let’s make it 7. Total homicides in the United States for that period would be around 100,000. Thus, according to that report, illegal aliens have committed 25% of all homicides in the US for that period.

If you assume that there are 12 million illegal aliens in the US (those are the estimates, right?) that would mean they constitute less than 4% of the US population but commit 25% of all homicides.

That seems to be a staggeringly high level of crime from illegal aliens - so I am surprised at those numbers. So - did I misinterpret the report?

WSJ: The Mythical Connection Between Immigrants and Crime.

I presented the data. I calculated the result. It shows staggeringly high illegal alien homicide rate. What is the explanation?

The page refers to Arrest offenses. In the case of 5 guys killing 1 person you would have 5 arrests and thereby inflate the number if you assume 1 arrest offense = 1 homicide.

Point. But - is it possible that the average is 6+ arrest offenses = 1 homicide?

This is completely unresponsive to the OP.

The OP is talking about homicide convictions compared to population. Your cite is talking about overall violent crime rates - not separating the rate of decline between immigration status. Then it talks about overall incarceration rates, not just homicide.

Primary convictions would give you a better view than arrests. It would still rely on a N:1 relationship, but given only those primarily involved in a homicide would have that arrest considered the offense with the longests sentence it should be closer to the actual numbers.

Page 23 has the federal numbers where homicide sits within the Other category which itself makes up 2%.

Over 50% of all illegal immigrants commit the same crime: illegal entry, which is a crime under federal law.

[QUOTE=Wikipedia]
The Immigration Policy Center (IPC) is the research and policy arm of the American Immigration Council, a 501(c)(3) organization in the United States dedicated to promoting immigration to the United States and protecting the rights and privileges of immigrants in the United States.
[/QUOTE]

In other words, it’s an explicitly partisan source. Their claim that immigrants are responsible for lower crime rates also confuses correlation with causation.

The study population is entirely made up of aliens who have committed crimes. You might find that skews application to the general [illegal] alien population just a tad.

Are you responding to the OP? Of course that’s what the study about - illegal aliens that committed crimes, as they document in the study. What’s your point?

I misunderstood your extrapolation, but I found your problem.

[QUOTE=You]
Page 58 in the PDF (52 in the report) provides the time scope of the convictions data: it’s for 4-7 years (7 years federal data, 4 for state data).
[/QUOTE]

The conviction data is for short periods above. The arrest data (that you cited) is for…

[QUOTE=GAO]
Our analysis includes criminal aliens with arrests dating from August 1955 to April 2010. About 90 percent of the arrests in our study population occurred after 1990.
[/QUOTE]

I think if you’re trying to figure out how likely members of a group are to commit a crime, you’ve got to count arrests rather than crimes. If you go by crimes, one atypical individual can drive up the average rate.

It’s like when Bill Gates drives through a town and the average wealth of everyone in that town is doubled. Then when Bill Gates crosses the town line, the average wealth of everyone in town is halved. But if you counted what percentage of the people in the town had a million dollars or more, Bill’s passage through town would barely move the numbers.

If you want to give an accurate picture of the comparative crime rates, tell us what percentage of illegal immigrants commit a murder vs what percentage of native-born citizens who commit a murder.

It’s an improvement on the OP. Terr asked, “Illegal immigrants - do they commit more crimes?” – in general terms, that is – and then cherrypicked one stat, of, as has been shown in this thread, dubious relevance.

Given the content-free nature of your contribution, I’d characterize it as a **misdirection **rather than an improvement. YMMV.

(yay for haphazard bolding!)

Just to clear things up further, if we take the ~90% of homicide arrests in the study that occurred from 1990-2010 (22,500), and factor them into the ~379,000* homicides we had over that time span, we see that the criminal alien arrest/total homicide rate is 5.9%.

Of course, that brings us back to the fact that some of these aliens would have been arrested twice for the same crime, or that some will be co-suspects for the same crime, etc. If we could control for those figures I suspect we’d end up at around 4%, or the same overall rate as the general population.

In practice, I expect that undocumented immigrants commit a slightly higher proportion of crimes than the general public simply based on socioeconomic factors.

*The number is actually more like 390,000 but I didn’t feel like adding the hundreds, tens, and ones.

Ok, so 90% of arrests are for a 20 year period. That definitely changes the stat - so the % of homicides committed by illegal aliens, instead of 25%, becomes about 7-8%. Still way higher than the less-than-4% of the population but much less egregious.

That drew my attention to another table in that report - Table 8 - Number of State Criminal Alien Assistance Program (SCAAP) Criminal Alien Incarcerations in Each State, or, simply put, number of illegal aliens in each state’s prisons. State prisons I think would be a better indication of criminality than federal prisons, since most illegal aliens in federal custody are there not because of crimes but because of being illegal aliens.

So - according to this table, in 2009, CA had 102,000 illegal aliens in its prisons. Now, CA peak prison population was in 2006 - 163,000 prisoners. Does this mean that 62% of CA prisoners are illegal aliens? Considering that at estimated 2.4M, illegal aliens are less than 7% of CA population, that seems ridiculously high. Again - did I miss something?

Other states are not that bad, but still - Arizona is 17.5K out of 55K - 32% (with illegal alien population of less than 9%.

There is an error somewhere. According to the CDCR in 2010 ( via the Sacramento Bee ) there were 22,173 potential illegals out of a population of 168,830 which = ~13.1% of the prison population. Meanwhile the population of California in 2010 was 37,253,956 and the estimated illegal population 2010 was 2,570,000 or ~6.9% of the total population.

Well, even your number shows that illegal aliens’ imprisonment rate is double that of the state’s population. But yes, that table is a bit weird.

Ron Unz has published the definite article on the subject: http://www.theamericanconservative.com/articles/his-panic/