Annals of derp: The Texas Criminal Alien caper

Texas Department of Public Safety.

The chart of total Texas alien (red and green) vs. non-alien (blue) arrests for the past 30 months put out by the Texas Department of Public Safety has some interesting properties. The most interesting, in my opinion, is the way it is not to scale, so that the blue part, with numbers representing about 92% of the total, looks like it's a little less than half. [jump]
Whoever designed this chart is intending to make you believe that foreign nationals made up more than half of the criminal arrests in Texas since January 2012, even though they were actually less than a tenth.

You couldn't do this by mistake.

I found the chart through PolitiFact Texas, which got to the page when they were investigating Governor Rick Perry's claim that "illegal aliens" had committed more than 3,000 homicides in Texas since 2008. Another odd thing is that it comes from a Web page that you wouldn't easily find on your own, because if you went to the DPS homepage you'd first have to click a link for "Criminal History Reporting" and then another for "Secure Communities for Texas" that takes you there, even though the page isn't actually called "Secure Communities for Texas" but "Texas Criminal Alien Arrest Data". You'd have to know in advance that it was there (although if you put "alien" in the search box you'd find it).

PolitiFact found it by asking somebody where Governor Perry got his numbers from, and that's where they sent them, to the page's other chart, an attractive pie representing all of the criminal charges that have ever been faced, or maybe all the Texas charges, or possibly all the Texas charges for the specific time period, we're not told which, by foreign nationals arrested in Texas over the not quite six years from October 1 2008 to July 1 2014:

Yet another odd thing is that the statistics for the last 18 months haven't actually been published yet, outside of these tables (and the two tables are all the numbers this very sketchy little page provides); the most recent official Texas Crime Report is the one for 2012. So we have no way of looking at the official year-by-year numbers for the past year and a half to check how it adds up or how it relates to the total picture, or really anything.

You can, however, find some numbers to think with. For example, for total number of homicide arrests for each of the years with a complete report and a race-and-gender breakdown of the arrestees. Thus for the five full years from 2008 through 2012 there were a total of 3,903 homicide arrests, and around 40% of the people arrested were Hispanic, meaning something like 1,560 of them. And given that Texas's Hispanic population is 30% foreign-born, it would be reasonable to suppose that the number of Hispanic aliens documented and undocumented arrested for murder in Texas from 2008 through 2012 would be more or less 468.

That is a lot less than the 3,000 given in the chart from late 2008 through early 2014. And even if all the Hispanic people arrested for murder were foreign-born, instead of just 30%, it would still be a lot less. And you're just not going to tell me that all the alien murderers were from Vietnam (there are about one tenth as many Asians as Latinos in the Texas population).

So roughly 500 people were charged with over 3,000 murders, six killings apiece? Are Texas murderers really that productive? Or what's going on?

The PolitiFact examination noticed enough things wrong with the numbers to make them give the claim a Pants On Fire rating; for instance, that we aren't given any information as to when and where the homicide charges were filed (i.e. if someone had been arrested for murder in Alaska 40 years ago and then got busted for marijuana use in Texas in 2010, that murder charge would have made it into the pie chart but would have nothing to do with any Texas murders in that or any other year).

I think I've caught something else, though, a piece of prime derp, that accounts for the numbers fairly directly, in the first chart above: in the green bars representing recidivists, that is those arrested who had been arrested in Texas at least once previously, who made up somewhat over 40% of the "Criminal Alien" total. Every time our noble Texas cops busted one of those people they asked if they had ever been charged with a crime, and the people told them and the cops added that to their tally. If you were arrested in Austin for public nudity 17 times and you had once been charged with felonious assault in Galveston, then that would add up to 17 felonious assault charges in the pie. If there were 100 aliens with murder charges on their records each arrested 30 times over the period from 2008 to 2014 for public intoxication, that would be all 3,000 right there.

Because these statistics don't count people, they count arrests and charges, without distinguishing the two, and they have no way of knowing how many times they've counted any given charge. And so they vastly inflate how many people are involved.

Habitual criminals get arrested more often, obviously. There's another falsity conveyed by the first chart, making it look as if 40% of the Alien Criminals were habitual offenders, when it's in fact the case that it's 40% of the offenses that were habitual, committed by who knows what percentage of the offenders, but I'll bet it's well under 10%. I'll bet practically everybody in the red bars, the one-time offenses, was in there for Class B misdemeanors or lower, and practically all of the Alien Felonies were committed by just a thousand or so (out of a total of maybe 3 million foreign-born Latinos in the state or about 0.03%), most of them legal residents. The charts don't simply fail to tell you this stuff, they actively hide it.

Because they aren't just stupid, although they are certainly that (designed, I suppose, for Governor Perry's personal use in his foaming-at-the-mouth exercises). They are also deeply dishonest. They are devised (by employees of the Texas state government) in order to deceive: in order to make people imagine Texas crime is mostly immigrant-driven and Latino, when it isn't.

What scum, really. No wonder these people hate government; they only know the kind of government that they do, and it really stinks. Torturing convicts to death and constructing elaborate, multicolored lies.