Take the 2-minute tour ×
Skeptics Stack Exchange is a question and answer site for scientific skepticism. It's 100% free, no registration required.

This is somewhat of the converse of the question Is Gun Control Effective:

A quote of famous musician and gun advocate Ted Nugent has been making the rounds on the Internet again lately:

Where you have the most armed citizens in America, you have the lowest violent crime rate. Where you have the worst gun control, you have the highest crime rate.

Do larger numbers of legally armed citizens correlate to lower violent crime rates?

I'm not as interested in the converse portion of his quote, as the effectiveness of gun control is covered under the question I linked earlier. Instead, I'm looking for any evidence that an increase in private gun-ownership acts as a deterrent for violent crimes within a given geographic region.

share|improve this question
    
"Most armed" by percentage or absolute numbers? –  Stefan Jan 3 '13 at 17:42
1  
@KristofferNolgren You are correct that any correlation between increased gun ownership and decreased crime does not imply causation. However, a clear correlation would lend credence to the theory that there may be a causal relationship, or at least indicate that the quote has some basis in reality, rather than a claim made up on the spot. That's what most studies aim to do, after all: either provide support in favor of a theory, or disprove a theory. You can never prove a proper theory right; only strengthen the support for it, or disprove it. –  Beofett Jan 5 '13 at 20:54
1  
@KristofferNolgren The point I think you may be losing sight of is that a claim was made by a notable celebrity that is being repeated as an argument against gun control. This question is not about whether or not gun control is effective. This question is because I'm skeptical about that specific claim made. Whether or not evidence corroborating that claim is a good means of determining whether or not gun ownership is an effective deterrent is irrelevant. –  Beofett Jan 6 '13 at 0:36
1  
Likewise, do nuclear bombs promote world peace? –  siamii Jan 13 '13 at 20:34
1  
I am somewhat puzzled that one can find restricting oneself to United States statistics irrelevant when postulating an answer to a question that specifically states "in the US". Given cultural/social differences, I see the data presented as irrelevant; apples and oranges, if you will. Interesting, but irrelevant. I agree that using state boundaries as area definitions is not terribly useful; however, comparisons of major metropolitan areas within the US might be more relevant. Those units have a wider variance available in terms of regulation. However, it remains that you cannot get 'pure' dat –  user11739 Mar 2 '13 at 2:22
show 10 more comments

2 Answers

up vote 31 down vote accepted

In order to answer this question, I don't see any reason to restrict ourselves to comparison between states. The variation in gun laws between states are quite small, and the ease of crossing state boundaries make it easy for crime to flow across those borders. Instead lets look at comparison between countries, where there is marked different in gun laws and less likelihood of cross-border crime. Fortunately there is a study by Mark Reid, a machine learning researcher at Australian National University that does exactly that.

Graph of gun deaths versus gun ownership

The graph is for OECD countries, i.e. those with reasonably developed economies. There is a clear correlation between gun ownership and gun deaths. The outlier of Mexico is due to an ongoing drug war

of course gun deaths covers a lot of things other than homicides, including suicide, and it is possible that the prevalence of guns causes people to use a gun for a suicide that they would use something else for if it wasn't available. So let's restrict the analysis to gun homicides.

Gun homicides versus gun ownership

Again Mexico is the exception, but so is the US. It has a much higher level of gun ownership and gun homicide than any of the other countries. So much so that it obscures any trends in the graph. Let's replot without those two countries.

enter image description here

Now the trend is much less pronounced, but still detectable (especially if you consider Israel a special case, given its ongoing conflict). Without Israel there are certainly few countries with low gun ownership and high gun homicide. And it's probably worth restating that the US has rates of both gun ownership and gun homicide more than double any other country on that chart.

EDIT: I haven't been able to make a chart, but the figures for all homicides (more than just gun-related) follow a similar pattern to gun-related homicides. The US has double the homicide rate of the next highest country in the OECD, and four times the rate of most (with the strange exception of Luxembourg, whose homicide rate is very slightly higher than half the US).

EDIT:Several people commented that the questions was asking about violent crime, not just homicide. I found what look like reasonable figures for assaults at Nationmaster Encyclopedia . Plotting those against gun ownership gives this graph:

Assaults v Guns

The US is the data point at the top centre of the chart.

Now it appears that high gun ownership is not correlated with high levels of assault. But that wasn't the claim. The claim was that high gun ownership reduced levels of violent crime. This graph certainly indicates there is no inverse correlation between gun ownership and assaults.

share|improve this answer
    
I think this is great info, and I appreciate you taking the time to post it. However, my concern is that the claim involves specific data grouping that winds up being lumped into the aggregate values you list. The crux of the claim is that states the have lots of citizens legally armed will have lower overall crime than states that have less gun owners, and combining the data across all states precludes determining the accuracy of this claim (i.e. if areas with low gun ownership have higher incidents of gun crime, you cannot tell from this data). –  Beofett Jan 4 '13 at 17:18
2  
I believe the key claim is "Do larger numbers of legally armed citizens correlate to lower violent crime rates?". Looking at state data is one way to answer it, but in my opinion not the best, for reasons I give in the answer. –  DJClayworth Jan 4 '13 at 17:21
2  
-1 as this equates violent crime to gun death which is not accurate at all, and even all homicides is not a totally accurate measure. –  Ryathal Jan 4 '13 at 18:35
2  
“trend is still detectable” – I’d like to see a correlation coefficient and a regression for that. In fact, I think it’ll be hard to argue that there’s much more than noise in there. The only thing we might conclude is that with increased number of handguns the variance of gun homicides increases (that itself would already be interesting, and indeed expected – at least by me). –  Konrad Rudolph Jan 4 '13 at 22:59
1  
@Dunk Do you have evidence to back up the idea that the definitions are different? –  DJClayworth Jan 14 '13 at 16:31
show 9 more comments

UPDATE 4/4/2013:

The claim is somewhat supported by graphing the US Census Bureau's 2006 Violent Crimes Data and the "Gun Rights Index" calculated here:

Scatter Diagram

But, statistically, the correlation is very weak.

Here is the data I used, scraped from the cited sites:

data = {
{Alabama, 425, 6.8},
{Alaska, 688, 8.0},
{Arizona, 501, 9.3},
{Arkansas, 552, 6.2},
{California, 533, 4.9},
{Colorado, 392, 7.7},
{Connecticut, 281, 5.1},
{Delaware, 682, 6.3},
{Florida, 712, 6.5},
{Georgia, 471, 7.1},
{Hawaii, 281, 2.6},
{Idaho, 247, 8.7},
{Illinois, 542, 3.0},
{Indiana, 315, 6.8},
{Iowa, 284, 5.8},
{Kansas, 425, 7.4},
{Kentucky, 263, 7.4},
{Louisiana, 698, 6.9},
{Maine, 116, 7.4}, 
{Maryland, 679, 5.1},
{Massachusetts, 447, 3.7},
{Michigan, 562, 6.3},
{Minnesota, 312, 6.1},
{Missouri, 546, 7.4},
{Montana, 254, 8.7},
{Nebraska, 282, 7.4},
{Nevada, 742, 7.4},
{New Hampshire, 139, 7.4},
{New Jersey, 352, 4.1},
{New Mexico, 643, 7.4},
{New York, 435, 4.6},
{North Carolina, 476, 6.9},
{North Dakota, 128, 6.2},
{Ohio, 350, 7.4},
{Oklahoma, 497, 6.2},
{Oregon, 280, 7.4},
{Pennsylvania, 439, 7.4},
{Rhode Island, 284, 4.6},
{South Carolina, 766, 6.3},
{South Dakota, 171, 8.7},
{Tennessee, 760, 8.1},
{Texas, 516, 6.8},
{Utah, 224, 7.5},
{Vermont, 137, 8.0},
{Virginia, 282, 6.9},
{Washington, 346, 6.7},
{West Virginia, 280, 7.4},
{Wisconsin, 284, 6.2},
{Wyoming, 240, 8.7}
}

This is really a comment on @RedGrittyBrick's answer, but I do not know how to put graphics in a comment...

The claim is not supported by the data provided by @RedGrittyBrick (which are incomplete -- note FL's lack of gun crime data, which is a pretty important data point)...

Here's the plot of "Gun Murders / 1M Population" vs "Gun Ownership %":

gunmurders_vs_own

And the linear model parameter table:

enter image description here

This is totally back-of-the-envelope stuff, I haven't checked the sources, I don't assert a logical connection between "% gun ownership" and weaker gun laws, I'm dubious about the logical connection between "gun murders" and all gun crime, I shudder at the use of "The Guardian" as a primary source, etc.


Edit by RedGrittyBrick (I've now deleted my "Answer")

The data came from US Liberals, The Guardian, ipl for kids but primary data sources were claimed to be


share|improve this answer
1  
Same problem as the other answer; you're focusing on 'gun deaths' or 'gun crime', which was not the question. –  kbelder Apr 4 '13 at 20:00
    
I updated the answer to use Census Bureau data on "violent crime," which is directly from the question. I tried to use the same wording I had used previously, to avoid editorializing. But since I'm not a talk-show host or trying to justify my own research, I think that it's obvious that, one way or the other, the thesis has little explanatory power. –  Larry OBrien Apr 5 '13 at 19:55
    
Interesting. Thanks. It looks like maybe this argument doesn't have a whole lot of power for either side of the debate. –  kbelder Apr 5 '13 at 22:39
add comment

protected by Oddthinking Mar 3 '13 at 3:54

Thank you for your interest in this question. Because it has attracted low-quality answers, posting an answer now requires 10 reputation on this site.

Would you like to answer one of these unanswered questions instead?

Not the answer you're looking for? Browse other questions tagged or ask your own question.