Chad Perrin: SOB

27 July 2009

Statistics 101 Lab: Examples

Filed under: Cognition,Geek,Liberty — apotheon @ 12:49

For some reason, a number of people who read Statistics 101: US Gun Crime vs. UK Knife Crime completely missed the key points. In summary, those points are:

  1. Comparing crime rates of two locations with significant cultural, legal, and geographical differences, situated in very different surrounding environs, will not provide meaningful statistical evidence that legal restrictions on firearm ownership reduce any crime rates. This is because other factors (such as those cultural, firearms unrelated legal, geographical, and environmental differences) can very easily overwhelm the differences that may be caused by any legal firearms restrictions.

  2. There is nothing magical about “gun crime” that makes it necessarily a good thing to reduce the rate of “gun crime” without any regard for the effects on crimes related to other factors than firearms. As I said in discussion following Statistics 101: US Gun Crime vs. UK Knife Crime, if you get murdered:

    [You’re] not going to much fucking care whether the act was committed with a gun, a knife, or a sharpened kewpie doll.

In case clearly spelling out the point about regional crime rates here, in summary, is not enough to make the point obvious to readers, I’ll offer some more regional examples of the sort of statistical comparisons I would reject as not sufficiently rigorous to provide evidence as good as a comparison of conditions before and after a significant change in firearms laws:

  • Illinois and Indiana are neighboring States — much closer and, as two States within the US, at least theoretically far better choices for a comparison of violent crime rates and gun laws than the US and the UK. All crime rate statistics are per 100K per year.

    Statistic Illinois Indiana
    2007 Total Violent Crime 533.2 333.6
    2007 Murders 5.9 5.6
    2007 Forcible Rapes 31.9 27.5
    2007 Robberies 179.7 124.1
    2007 Aggravated Assaults 315.7 176.4
    1960-2007 Avg. Murders 8.59 6.07
    1960-2007 Max Murders 11.8 8.9
    1960-2007 Min Murders 4.8 2.8
    Carry Laws No Concealed Carry
    Carry Prohibited in Vehicles
    Open Carry: Unincorporated Areas
    Shall-Issue Permits
    Carry in Vehicles with Permit
    Open Carry with Permit
    Ownership Laws Firearm Owner ID
    Handguns Prohibited in Chicago
    No License to Own
    No Handgun Restrictions
    Sources IL Crime Rates 1960-2007
    IL Open Carry
    IL Concealed Carry [PDF]
    IN Crime Rates 1960-2007
    IN Open Carry
    IN Concealed Carry [PDF]

    Note as well that the murder rate was higher in Illinois than in Indiana every single year from 1960 to 2007. There is not one single exception to that rule.

Actually, I’m going to stop here, and make this a one-example list for now, because I’m getting too interested in this statistical analysis. That doesn’t mean I’m going to stop working on these stats, or that I will not produce more comparisons, but I’m going to stop working on it for this particular SOB entry because this one table is already getting too long. I’ll make sure to link to more comparisons when I have worked them out. Writing Ruby code to munge the data automatically is turning out to be too much fun.

Anyway, I’ll wrap up with what I planned to use as my final point for this table of data:

Judging by these statistics, the status of both gun laws and violent crime rates in Illinois (one of the most RKBA-restrictive States in the Union), compared with those in Indiana, shows a very strong correlation between strict gun control laws and high violent crime rates. These are neighboring States, within the same federated union of States, both operating under the aegis of the same federal firearms laws — a much closer comparison than between the US and the UK. Only certain municipalities within Illinois, most notably Chicago, even approximate the restrictiveness of UK laws regarding firearms.

Despite this — despite the apparently damning evidence of more restrictive firearms laws causing greater rates of violent crime — I do not consider that as good a comparison as a comparison of crime rates before and after the passage (or repeal) of a new, significant firearms law within a single jurisdiction. The reason is that, even though they are neighboring States, they are still different enough in a number of ways that many other factors are likely to interfere with a “clean” comparison than would do so in a statistical analysis of violent crime rates before and after a change in the law.

I believe that the difference in gun laws between these two states does contribute to the difference in violent crime rates, but that doesn’t mean I’d accept these statistics alone as proof of that. The fact remains that other variables are not sufficiently controlled to make this simple comparison between States and be certain of the results.

All original content Copyright Chad Perrin: Distributed under the terms of the Open Works License