Debunking "A Tale of Two Cities"

You know, now that the holidays are over, I can spend a little time clearing out my email. And what do I find? A message from my dad! Let's see what he has to say!

Aw, it's cute. It really is. I love it when unsubstantiated facts and statistical anomalies are stirred together and turn into fertilizer. And this one has tables and everything! In fact, it looks something like this.

A Tale of Two Cities

	Chicago, IL	Houston, TX
Population	2.7 million	2.15 million
Median HH Income	$38,600	$37,000
% African-American	38.9%	24%
% Hispanic	29.9%	44%
% Asian	5.5%	6%
% Non-Hispanic White	28.7%	26%

Pretty similar until you compare the following:

	Chicago, IL	Houston, TX
Concealed Carry gun law	no	yes
# of Gun Stores	0	184 - Dedicated gun stores plus 1500 - legal places to buy guns- Walmart, K-mart, sporting goods, etc.
Homicides, 2012	1,806	207
Homicides per 100K	38.4	9.6
Avg. January high temperature (F)	31	63

Conclusion: Cold weather causes murder

Now, my dad is a reasonably smart person, so I can't assume that this is evidence of incipient Alzheimer's or anything. In fact, mostly, it looks like he just forwarded somebody else's data, without bothering to fact-check it (I'm sure most of us have relatives who do that). But since it's sitting there stinking up my inbox, I guess it deserves an answer.

First off, let's start with the fact that any time the NRA tries to claim that Chicago's "unreasonable gun laws" don't do any good, it ignores the fact that Chicago is surrounded by unreasonably loose gun laws, and anybody who wants a gun just needs to drive an hour to get someplace where they can buy one without a problem. So, you know, that part's crap - Chicago's laws have minimal effect because those laws have been nullified. Or, if you really want to look at how it works:

More than a quarter of the firearms seized on the streets here by the Chicago Police Department over the past five years were bought just outside city limits in Cook County suburbs, according to an analysis by the University of Chicago Crime Lab. Others came from stores around Illinois and from other states, like Indiana, less than an hour’s drive away. Since 2008, more than 1,300 of the confiscated guns, the analysis showed, were bought from just one store, Chuck’s Gun Shop in Riverdale, Ill., within a few miles of Chicago’s city limits.

Now, let's look at the statistics as presented. Assuming they're accurate (and we'll get to that in a second), remember the phrase "pretty similar until you compare the following." Because, just taking them at face value, you have a 15% difference in African American populations, and a 14% difference in Hispanic populations. Anybody who thinks those numbers are "pretty similar" either failed statistics, or never graduated high school.

But you can just feel free to pull out your Klan membership card and claim that the higher number of blacks explain the difference in the murder rate. (Trust me, the argument has been made.) Of course, you'd then also have to explain how the lower percentage of Hispanics has affected these statistics, and I'd LOVE to hear you try to argue around that corner.

But then, just for fun, let's consider the REAL facts. (You remember "facts," right? Those things Fox News has no time for?) First of all, this link here goes to a Cost of Living calculator. Now, I want you to do a little homework (calm down, it isn't difficult). Compare the costs of living between Houston and Chicago.

Done? Did you notice that tricky little 22% percent (average) difference in the cost of living? So that a person making $78,000 in Houston would need to earn $100,000 to live in the same style in Chicago? Hmmm... I wonder if that has any effect?

But, you know, those numbers in the chart still seem a little off. And statistical analysis is probably a real pain when you're working with incorrect data, isn't it?

So I went looking, and it seems that there's this thing the census bureau does, and it's called the American Community Survey. But those are all these tables, filled with numbers and stuff, and I don't want to make anybody's head hurt worse than it probably does. So I found a website that extracts numbers from the Census Bureau's American Community Survey, and you know, it's funny. There seems to be a discrepancy here. Just a slight one.

Because, as it turns out, the median household income for the Chicago-Naperville-Joliet Illinois metro area was $59,261 in 2012. Not $38,600, as claimed. Wow, that's a little bit of a difference, isn't it?

And look here: the median household income for the Houston-Sugar Land-Baytown Texas metro area was $55,910 in 2012. Not $37,000. That's kind of interesting, too.

But, you know what else? I seem to remember just a couple of months ago, when it was big news that Chicago was the "murder capital of the USA." But, funny thing. The number of homicides wasn't 1,806, like that cute little table claimed. Seems like it was more like 500 or so. Isn't that odd?

But let's check that, shall we? How about we look at the FBI's official data? And we poke around for a while, and we see that, sure enough, the number under "Murder and non-negligent manslaughter" for Chicago was exactly 500. Kind of a round number - you know, the kind of number that might stick in your head if you had any interest in actual facts, instead of... well, I don't want to call it "fecal matter," because that would be rude. But still...

So they were... well, maybe they were off by a little bit. Roughly 1306 homicides off, to be exact: they were wrong by almost three times the actual figure! I wonder how they did with the number of homicides in Houston? Well, right there, they were MUCH closer! Houston had 217 homicides, instead of the 207 in the table! That's so much closer! I mean, it's still wrong, but it's so much better than they've been doing!

But still, it seems like a lot more people have been killed in Chicago than in Houston, doesn't it? That's just weird. Is there some sort of difference between the Chicago mobster and the Texas cowboy that could account for these numbers? I wonder if anybody has looked into this problem?

Efforts to compare the strictness of gun laws and the level of violence across major American cities are fraught with contradiction and complication, not least because of varying degrees of coordination between local and state laws and differing levels of enforcement. In New York City, where homicides and shootings have decreased, the gun laws are generally seen as at least as strict as Chicago's, and the state laws in New York and many of its neighboring states are viewed as still tougher than those in and around Illinois. Philadelphia, like cities in many states, is limited in writing gun measures that go beyond those set by Pennsylvania law. Some city officials there have chafed under what they see as relatively lax state controls...

"The way the laws are structured facilitates the flow of those guns to hit our streets," Garry F. McCarthy, the Chicago police superintendent, said in an interview, later adding, "Chicago may have comprehensive gun laws, but they are not strict because the sanctions don't exist."

So, really, even if you used accurate numbers and factored in socioeconomic data, the numbers wouldn't really mean a thing, would they? It's almost as if this email was comparing two completely unrelated things, isn't it?

I wonder if whoever put together the original chart knew that when he wrote it?