Not sure how to extrapolate crime rate based on this survey of defensive gun use

Great Debates
1

According to this survey, gun owners used a firearm defensively about 1.67 million times a year. (On the low end, a study Politifact cited from the Rand Corp estimated about 100,000 times a year).

How would you do the math to see what effect that would have on the overall crime rate? In other words are 1.67 million crimes (or even 100,000 crimes) being prevented annually? Does this mean that people who don’t own firearms are more or less likely to be crime victims? I put this in FQ, optimistically, but I know how these things go, so it goes without saying that the mods will do what needs to be done…

2

One way to start looking for a meaningful statistic would be to create two number series - the 1.67M gun uses and the number of incidents of the relevant types of crime in the same jurisdiction. You’d probably record it fairly coarsely, like incidents per month or fortnight.

The claim is that there is a possible relationship between these figures in each month - more crime will increase gun use, or more gun use will suppress crime. At this stage we don’t know which way the influence goes or how strongly its correlated, so its purely exploratory to see if the series behave in tune.

The null hypothesis (which is what you are actually testing with the stats) is that your two number series are not actually correlated and any apparent pattern is just the result of random chance. Its important that your statistical test can shows that the patterning you have seen only occurs rarely, eg <5% or <0.1% of the time by chance, so is less likely to be happenstance.

People who do this sort of maths for a living can suggest suitable tests. If your test shows that there is no or weak correlation between gun use and crime rates, then you can at least answer one of your questions (no gun - more or less likely to be crime victim) with ‘probably not going to make a difference’.

3

I’ve not read the article but the 1.67 million seems high. It might come down to the definition of ‘used defensively.’

In one case,“if you hear a noise at night and grab a gun while you investigate”, will have a higher response, then: “Did you discharge your weapon at a human threatening you with bodily harm.”

4

Just skimming the article suggests that there’s not going to be detailed enough information (at least from that study) to try to calculate Defensive Gun Use (DGU) per month.

As I understand the methodology, the survey asked whether the gun owner ever had a DGU and, if so, how many. They concluded that there had been approximately 50 million DGU (by current gun owners) and divided by the average adult years of the gun owner to get to 1.67 million DGU per year.

Which seems to illustrate another problem, which is that I don’t see how you can extrapolate from this averaging to correlate to changes in crime rates.

5

The study is based on responses to a survey, but I didn’t get far enough to see whether or not they filtered out things where the respondent thought there was a threat and that turned out not to be the case. For example, one night in the mid 90s at 3am, a car followed me into my driveway. I had a pistol behind my back, but it turned out that the guys in the car had meant to pull into the driveway next door to see the neighbor. I don’t know how this would have been counted in the survey. I was really trying to get a sense about whether the 1.67 million figure would indicate crime was much higher than it is.

6

Seems to me I’ve read about this before, and that’s the rub. IIRC, these surveys simply ask people if they’ve used a gun defensively, and there’s no real way to verify it.

An anecdote: A co-worker once told me about the time he defended himself with his pistol. He was in his car and a shady group of guys was nearby, so he took out his gun and held it in his lap. The guys went away (not having seen the weapon). My friend apparently felt this was an important moment as he recounted it, I’m guessing from the comfort he got in having the weapon available. But I’m guessing such a story might count in the positive tally in some of these statistics.

7

I’ve done some reading in past years on the subject. As I recall, one reason for the wildly-varying estimates of defensive gun use is differences in definition. The studies on the high end always use a broad definition. The linked article in the OP says, “in most defensive incidents (81.9%) no shot was fired.”

That’s a far cry from the popular conception (either pro or con) of somebody blasting away at a criminal or perceived criminal.

8

Given that the factual question has been answered to the degree it may be possible/not possible to do so, I will float a hypothesis:

Self-perceived defensive gun use (the metric reported) will correlate fairly well with fear of violent crime victimization which only loosely correlates with actual crime rates.

I’d suspect there is a stronger correlation with a recent past report of a break in within a neighborhood and/or social circle and with recent past media reports of violent crimes than with actual crime rates either way.

9

The problem is- we just do not know what would have happened if there had been no gun.

I used my gun (I used to have a CCW) twice- once when walking to the train station in the winter dark in SF, escorting a couple of female coworkers. Two homeless guys lurched out of an ally, saying “Give us some money”, one had a board with a nail thru the end. I simply showed by gun, said something, he dropped his club, and nothing occurred. So- could we have just thrown a couple bucks at them? I dont know. But the women were scared shitless.

Next time was at a 24 hour newstand/bookstore. We heard noise, the clerk grabbed his baseball bat, and I followed. There was a guy pulling a women by her purse into a dark alley. The perp saw us coming, with my gun out, let go of the purse and ran away. Would the clerk with the baseball bat be enuf, or maybe he and I (with no gun out) as a show of force? Sure. Maybe. She was shaken up, the clerk gave her a cup of coffee, she didnt want to wait for the police, but she was effusive in her thanks.

So, you can say in both cases, the gun did nothing. Or you can say in both cases the gun prevented a violent crime. I honestly can’t say either way, except we did feel more secure.

That’s the issue- did the show of firearm force prevent a crime? Or was it unnecessary? My WAG is about half the time, it was unnecessary- except for the feeling of security. But that’s not nothing.

However, in neither case was anyone shot, either. So, maybe not necessary, but no harm was done.

Now, we do have some cases where a good citizen or off duty cop or security definitely prevented a mass shooting, murder, rape etc. Those are, admittedly- rare.

10

Maybe “defensive gun use” should be broken into two categories, “displayed weapon, threat stopped” and “fired weapon, threat stopped”.

11

Hmm. That could work, but still.

12

Considering how few times “good guy with a gun” makes the news, I suspect the number is grossly exaggerated if it refers to actual weapon discharges.

13

[Moderating]

I don’t think there’s really sufficient data here for a factual answer. Moving to GD.

14

It doesn’t refer to actual discharges. Those are, IIRC, a small subset of the quoted figure. Homicides in self defense are an even smaller portion.

15

Gotta be at least one more category: “actually, it turns out the gun use was offensive and may even have constituted a crime, but the gun owner was too oblivious and indifferent to the rights and safety of others to notice.”

16

Correct. And many defensive gun displays are not reported to the police.

17

This was my immediate thought upon hearing the results are based on a survey. George Zimmerman, the McMichaels, Susan Lorincz, and Andrew Lester would all have answered they successfully used a gun defensively.

18

Respectfully, I think there’s an FQ answer to this.

Are there sufficient incidents nationwide to make this figure plausible (or even the lower figure of about 100k)?

I am trying to respond to someone who cited the study, but I want to see if it’s legit or if the self-reported incidents are mostly the result of paranoia. Dr. Deth’s incidents don’t seem like paranoia to me, but I didn’t read far enough in the study to learn if the questions asked attempted to ferret that out.

19

I guess I question the result, because many times that a situation arises that would require using a gun to prevent further danger, a crime has already happened and would be reported. For example, if some guy got into my car threatening to steal it from me, but my visible gun (I don’t own a gun) scared him off, I would still call the cops and report it.

Does the survey specifically ask for prevented crimes that were never reported, or could there be tons of double counting?

20

Just for the record,the “paper” is actually a report that was been put in a repository for releasing working papers and hasn’t been peer reviewed (although it may be published later).

Also the author does appear to be firmly on the guns rights side of the spectrum.

He wrote a paper entitled "“Economic and Ideological Corruptions of the Regulatory State” with the following abstract

What accounts for the excesses of the regulatory state? Economists have shown that bad regulations are often due to “rent seeking” and “regulatory capture” and have provided compelling accounts of the incentives that drive these processes. However, there remain many unwarranted regulations, such as recent gun control measures, that cannot be explained by an economic rationale. Rather, it is essential to understand the distinctively ideological origins and functions of such regulations. This paper examines both the economic and ideological corruptions on the regulatory state, suggesting that those concerned with the growth of state power will have to address both of these phenomena