Sorry, rules are rules.

Donald Trump's much anticipated (so they say) speech on illegal immigration used the word "kill" six times. "Murder" came up three times. There was one "rape," three "violent," a whopping 24 "criminal," five "horrible," just one "terrible" (!), five "drugs" (enough to share?), two "deadly," two "homicide." Seven "death."

I'm not going to joke about this (except where I already did, and a few later). This is detestable. It's beyond detestable. My command of English feels unequal to the task of adequately describing how revolting Trump is when he talks about undocumented immigrants.

I would be revolted by this even if undocumented immigrants were committing a disproportionate number of serious crimes. If that were true, Trump would still be guilty of smearing the innocent along with the guilty and feeding fears that could easily lead to serious violence.

But it's not true. Undocumented immigrants do not commit a disproportionate number of crimes. This has been reported before. It's not news. But what can we do but keep saying it?

Well, we can make a chart:

photoshopped, u can see the pixels

As you can see by the words on this chart, the blue is the undocumented (or "unauthorized" per Pew) immigrant population, and the red is the rate of violent crime as reported by the FBI.

This chart doesn't prove that undocumented immigrants aren't committing violent crimes. Undocumented immigrants are a small percentage of the total US population, so conceivably they could commit crimes at twice the rate of legal residents and you could still make a chart that looks like that, more or less. The only thing it says definitively is that as illegal immigration has risen, crime has fallen. It's a good place to start.

It's better, anyway, than the evidence Trump cites. The closest he got to data in his immigration speech was when he read off the Teleprompter that "our prisons and jails together, had around 25,000 homicide arrests to their names, 25,000."

That certainly is more than zero. But without relating it to other data for context, it's like saying "We have a tremendous problem with gerbils because 16."

I too can make up a bunk statistic, one that's even better than Trump's "big number, scary!" one, because it sounds mathematical. Observe: The violent crime rate declined by 31% between 1995 and the peak of the undocumented population in 2007. During that time, the undocumented population grew by 114%.

Glance at that, and you might think there's no way undocumented immigrants are committing a disproportionate amount of crimes. But of course, as with my chart, there's really not enough information to draw that conclusion. It seems likely, but it's not proven.

So to refute Trump effectively, I'll need to find something to refute, some evidence to stand in for his lack of any. There was a Fox News article last year around this time that will be perfect. In retrospect this article seems almost liturgical in its xenophobia, given what's developed since. It wouldn't be shocking if Trump's "platform" on immigration came largely from this single source.

The headline was "Elusive crime wave data shows frightening toll of illegal immigrant criminals." This is filed under US News (it has numbers in it), yet it's also larded with Opinion-section descriptions of "wildly disproportionate number of murderers, rapists and drug dealers" crossing the border, "explosive figures" in the crime stats as a result, and of course with words like that you'll certainly believe that illegal immigrants commit "far more crimes than their 3.5-percent share of the U.S. population would suggest." And of course the "crime wave" and "frightening toll" in the headline itself.

You know, "news."

So I was curious: Most of what I've read said that undocumented immigrants don't appear to commit crimes at disproportionate rates. But might this just be George Soros's liberal media devils on each of my shoulders and in 15 browser tabs whispering "All is well, sleep, sleep"? Rather than sleep, I decided to check some of the statistics in the Fox News article, because I couldn't check Trump's, because he doesn't use them.

Florida was easy to check. The Fox News article says that "[i]n Florida, there were 5,061 illegal immigrant inmates in state prison facilities as of June 30 [2015]."

You have to check the Florida Department of Corrections' annual report to learn that there were 100,050 inmates in Florida state prisons then, which means that about 5.1% of them were undocumented immigrants. That still doesn't mean anything unless you know how this compares to the overall percentage of undocumented immigrants in Florida's population.

The most recent data I could find was a Pew report from 2012 that put Florida's undocumented population at 925,000. Florida's population was 19,352,000 in 2012, meaning that about 4.8% of Florida's population was undocumented then.

So Florida's state prison population is more or less 5.1% undocumented, and the overall state population is more or less 4.8% undocumented. I would hardly say that's a "wildly disproportionate" "crime wave" among "illegal immigrants" that is taking a "frightening toll."

Texas is an even weirder case. The Fox News article leans heavily on analysis and quotations by J. Christian Adams, who wrote a PJ Media post headlined "Illegal Alien Crime Wave in Texas: 611,234 Crimes, 2,993 Murders". That post quotes a Texas Department of Corrections report:

"From October 2008 to April 2014, Texas identified a total 177,588 unique criminal alien defendants booked into Texas county jails."

First, that's a bit of a weird period: five years and six months? It gets weirder. Apparently, The Texas DOC issues this report regularly, and they've changed the wording since the PJ Media post was published. The current one says:

"According to DHS status indicators, over 199,000 criminal aliens have been booked into local Texas jails between June 1, 2011 and August 31, 2016."

In the earlier report, we were only looking at "criminal alien defendants booked." But apparently now, we're looking at "criminal aliens booked." There's no mention of "defendants" anymore, but the number they give seems like it's measuring the same thing. If you break it down by month, that's 2,691 bookings/month during the earlier period, 3,157 bookings/month during the later period. It seems like we're dealing with arrests the whole time.

That's good, because we can easily compare it to Texas's arrest data and look for any indication of an illegal immigrant crime wave.

Long story a little bit shorter, if "criminal aliens booked" means the number of arrests, then during the 2008-2014 period cited in the PJ Media post, about 3% of all arrests were "criminal aliens." About 6% of Texans are undocumented.

It's likely that there are problems with my analysis. For example, it's not clear what exactly Texas is measuring when they report on "criminal aliens booked." Does that include only certain offenses, or all offenses? It's not explained, and there's no data set or references on the Texas DOC's Criminal Alien Arrest Data.

Basically, you need to make guesses. The data is bad. The Fox News article calls the data "elusive" in ominous Big Brother terms: "Critics say it is no accident that local, state and federal governments go to great lengths to keep the data under wraps."

No, the data is just bad, or not there to begin with.

Trump could be arguing for stricter immigration enforcement and tougher laws without evoking blood and death and suffering. But he is a salesman, and he knows that fear is a great motivator. So he'll use that, and if it means whole ethnicities start to feel threatened in their own land; if it means a few more people take beatings or are killed; well, he was trying to close a deal. Which is, of course, vile beyond my power to describe.

But gosh, did he look statesmanlike up there with Enrique Peña Nieto the other day, huh?

[ LA Times / USA Today / Pew / FBI / Fox News / FL DOC / Pew / TX DOC / TX DOC ]