Tuesday, April 15, 2008

Virginia Tech Massacre Anniversary: Guns and Suicide

Tomorrow (April 16) is the first anniversary of the massacre/suicide at Virginia Tech. While moves have been made to restrict access to guns by the mentally disturbed, no other such moves have been made in Virginia. People like John Lott oppose such moves, arguing that allowing people to carry guns on campuses and in other public places reduces homicides (Lott has also been dumping on Obama for his supposedly bitter unhappiness with gun ownership). The data on that is a mixed bag, but another aspect of this is much clearer and a part of the VA Tech story, the very strong link between gun ownership rates and suicide rates within the US. It is just a lot easier to kill oneself if there are lots of guns around, and one does not have a chance to second guess one's intention.

I have done some digging around and guesstimate that if the US as a whole had the gun onwership rates one finds in the lowest states, we would see on the order of 10,000 fewer suicides per year in the US. That is more than the total killed on 9/11 or US military killed in Iraq. Here is a sources on suicide rates: http://www.suicidology.org/displaycommon.cfm?an=1&subarticlenbr=21. Here is a source on gun ownership: http://www.swivel.com/data_sets/show/100359. The five highest states in gun ownership are Wyoming, Montana, South Dakota, West Virginia, and Idaho. The highest states in suicide rates are Montana, Nevada, Alaska, New Mexico, and Wyoming. The five lowest states in gun ownership are D.C., New Jersey, New York, Rhode Island, and Massachusetts. The five lowest states in suicide rates are D.C., New Jersey, New York, Massachusetts, and Rhode Island.


Unknown said...

My guess about the states with the highest rates of suicide are also the states with large populations of original peoples. Seems I read somewhere that leaders of our original peoples are concerned about this. Maybe has something to do with rural poverty?

Unknown said...

Just want to add that people with guns scare the shit out of me and that includes the police and the military.

media said...

south dakota definately has a suicide problem on pine ridge reservation (especially youth). likely, some of the same issues apply on nearby reservations, including ones in montana and wyoming. so that might explain some of the data. alaska and new mexico may have the same spectrum. emile durkheim might have termed this 'anomie'. nevada has las vegas, so some people may have been distressed at losing at gambling. In places like DC, sometimes you have 'death by cop' or variants serving as an alternative to suicide, to keep the numbers low; NJ and NY may have some of the same thing.

i would imagine homogeneous, nonstressed out, and culturally traditional places like nebraska, iowa, and utah might have low suicide rates.

it has been noted elsewhere that internationally, countries with a high HDI (eg sweden) also have high suicide rates, while suicide is uncommon in places like bolivia and parts of africa with low HDIs. guns are a conveniance, and sufficient, but not a neccesary condition.

Bruce Webb said...

I suspect media has the right approach, much of the correlation might just boil down to the intersection of the isolation and alienation inherent in living in a sparsely settled large western state, particularly ones that have very cold, very dark and very long winters and where having a gun around the house is essentially a requirement. (When that bear is attacking your dog or your kid calling a County Deputy who might be a couple hours away, or in Alaska several days away, is not really an option.)

Combine cabin fever with loaded guns and bad things are likely to happen.

Nevada is a special case. Much of the state fits the sparsely inhabited plus gun model plus you get the added isolation of the casinos. In some ways you will never be more alone than you are sitting in a crowded casino.

I just did a quick Google on 'swedish finnish suicide rates' and see that you are twice as likely to kill your self in Finland than Sweden. And that the former country has the third highest rate of gun ownership in the world (after the US and Yemen). On the other hand the Finns have for centuries had the reputation for being moody. Winter nights? or Gun availability? Or national psychological trends? Which if any rules? (That damn correlation/causation problem never seems to go away, there almost always is a 'on the other hand' lurking).

rosserjb@jmu.edu said...

Over on marginal revolution John Lott has put forth a "rural isolation" thesis as a supposed explanation for the apparent relationship between suicide rates and gun ownership rates. Maybe, but I think it kind of falls apart on closer examination.

So, I went and looked at depression rates by state, which presumably are the main underpinning of suicide. I found that South Dakota had the second lowest rate of depression among youths of any state in the US, but there it is with a youth suicide rate twice that of the national average, with that rate four times the national average on the Indian reservations. I think the Indians have reason to be depressed, but Robert Putnam reports that eastern South Dakota has the highest rate of social capital of any place in the US. Being third highest in guns, even with low depression, there go the suicide rates.

At the other extreme, second in population density in the country Rhode Island turns out to be second in adult depression rates. However, it is in the bottom five on both gun ownership and suicide rates. I think we have the culrpit cornered, if not captured.

In the meantime, the US Supreme Court is about to overturn the strict gun ownership laws that D.C. has, with its lowest rate of gun ownership and lowest suicide rate in the country. I hope the NRA and its allies take responsibility for all the increased dead as the suicide rate there goes up after this happens.

BTW, there was a moment of silence around the state here in Virginia in honor and memory of those slaughtered at Virginia Tech a year ago.


Bruce Webb said...

Barkley just by geographical situation a depressive in Rhode Island is no more than minutes from an outlet where you can self-medicate in the company of others who are similarly self-medicating. In a hugely rural state the triangle of depression/guns/social self medication may be a little different than just jumping in your car in Providence Town and going down to the local bar. Access to Bud, booze, broads and bros all with unlimited access to Busch series racing might mean the difference in the rates.

media said...

looking at the various 'swivel' charts, assuming they are true (and not Lott's propoganda---the west virginia numbers for homicide don't actually look right, but who knows ), it is interesting to note that while NY NJ, and Md. have low numbers of gun ownership (and suicide) they have high homicide rates, so the 'elasticity' of the homicide welfare function is lower than for the suicide welfare function, for a given bundle of guns---that could be phrased better. I wonder which symmetry group is involved---probably E8 since its trendy.)

(Of course it is interesting to consider that maybe homicide and suicide are actually 'goods'
which only because of traditional prejudice, due to poor upbringing, are considered 'bads'. Suppose half the people in the US chose one of these options today; global warming might be reduced 15%immediately. Has Obama considered this as a policy----for republicans-- get out, the vote. But then you'd have the 'free rider' problem, so one would have to convince china to do it too, perhaps via confus/cionist rhetoric. )

the japanese apparently have high suicide rates, yet few guns. bad feng shui? if a 6 were a 9, maybe eastern civilization could become western, and balanced.

I am sure that the NRA will take full responsibility for any possibly (pure speculation) bad choices they have made in life; those are the most bestest peoples. (I hear that actually the DC gun case is not about guns, its about fundraising for the NRA by keeping the issue alive, the way Jesse Jackson does with 'hope'. All those lawyers need something to do, also, to justify $500/hr. Then you can afford prozac. don't worry, be happy. Then we have the law schools, journals, etc. Maybe instead they could deliver ice and bottled water to antarctica?).

perhaps with all the firearm profits the NRA have made flooding the inner cities with them (a fashion acccesory, like an ipod!!!) they can sponsor a scholarship to GMU or UC to study with Lott or fellow travelers, on peaceful conflict resolution, via conceal carry or some equally worthy variant---unless it was abortion that caused crime, or maybe darwinism, or adam and st/eve and snakes. Find the one deserving person who might benefit from avoiding depresssion or homicide or suicide, and then see if they can study and find a way to solve the problem. 'school choice!!!' The Philanthropy administering the scholarships alone might be like the Gates foundation, and offer employment in perpetuity for its administrators and their families. Bush and Cheney might like this as a competetitor to the carter center.

I wonder if via the LTV one can compute the relative quantity of bad decisions and associated profits made by gun sellers, NRA lobbysists, and poorly thought out bestseller writers. In heaven, then you get a merit badge based on rankings.

I am sure Lott, et. al. could use a new humvee and a few conceal carry permits to protect it with, so laundering gun money through scholarships seems to be a reasonable way to go. The grease gets the squeaky wheel, bit by it. Or as the americans say, don't tread on it, cuz in full effect, it flies like a butterfly, or Lorentz invariant quetzelchuatls.

rosserjb@jmu.edu said...

Across countries, culture is very important for suicide. It is honorable in Japan, the old samurai tradition of hara-kiri. It is frowned upon in heavily Christian US.

There were major outpourings at VA Tech and on our campus yesterday over the memorial. Many here knew people there, with a colleague of mine having a son there last year who upsettingly did not answer his cell phone for a long time when things were going bad. And two people from Harrisonburg were injured, the one most seriously a friend of my daughter's. So, this was personal for us around here.

Anyway, back to the gun issue, Lott is partly right that in the US the relationship between guns and homicide rates is murkier. It is stronger internationally, while the suicide-gun relation goes the other way, weaker internationally because of all the cultural factors (Japan, etc.), whereas it looks very strong in the US.

In any case, regarding gun control, what I think is fully defensible is banning major assault weapons that were banned here in the 1990s, including of the type that were used a year ago at VA Tech. I simply have no use for the John Lotts who resist that.

Unfortunately the Dem prez candidates have pretty much caved on this issue, much like they did after Dukakis on the death penalty, even though I am sure they all know better. But they want that bitter, gun-toting, Bible-thumping, immigrant-hating, small town rural vote. So, there was Obama last night defending the idea of the Second Amendment conferring an individual right to own guns, even as he fended off the questions about flag pins, William Ayers, bitterness, Rev. Wright, and so on and so forth...


Kaleberg said...

Remember, the suicide rate being reported is actually the SUCCESSFUL suicide rate. Most people who try to commit suicide fail. Men, in particular, are much more likely to successfully commit suicide than women. Most people who try to commit suicide recover, but many are seriously injured, often permanently.

One of the reason men tend to commit suicide successfully more often than women is that they are more likely to use a gun. Women are more likely to try overdosing or poison, which are less reliable methods.

There is obviously a cultural baseline which determines the overall suicide attempt rate, but the method used has a big impact on the success rate. Easy access to fire arms makes for easier success at suicide.

Unsympathetic said...

The second amendment is about the right of the citizens of a country to exercise their right to overthrow the government if the government isn't responsive to the needs of the people.

Suicide is not relevant to this argument.. suicide is something that will occur with guns or without. The guy 2 doors down from me in my undergrad dorms at an ivy league institution hung himself with a bedsheet -- no gun needed.

This is simply a correlation v causation argument.

If you don't want to have a gun, you'd better not live anywhere more than 3 minutes from a police station. How many of you have prevented yourself from being robbed by pressing a pistol against a burglar's head? If you remove gun laws, the only people with guns will be criminals.

Happiness is most highly correlated with social involvement AND a sense of the ability to achieve a higher standard of living. The five highest states in suicide rates are... the 5 most sparsely populated with the least economic activity.

Guns and happiness don't belong in the same discussion.

rosserjb@jmu.edu said...


When I read the Second Amendment, I do not see anything about it being about overthrowing the government. Indeed, about half of it is about maintaining a militia, an arm of the government, although implicitly at the state level. Traditionally the courts have said that is the main meaning of it, not some right of the individual to own guns no matter what, although we may be about to have that changed so that D.C can have higher suicide rates.

Correlation is not necessarily causation, but often is. Statistically one can never "prove" causation, so this is a red herring argument, despite "Granger causality."

In any case, 90% of those who attempt suicide with a gun succeed. Only about 3% of those who try with other methods succeed. This is almost certainly the source for the very strong correlation, I would submit. Most of those who do not use a gun have another day later to rethink what they were up to, with many deciding it was a mistake and they can work things out.

BTW, this is personal. A good friend of mine recently attempted suicide. He used a razor blade in conjunction with a bunch of pills and alcohol. He was found in time, and he has since rethought his position and has a different view on life. If he had owned a gun, I have little doubt he would be dead now.


Unsympathetic said...

" Correlation is not necessarily causation, but often is."

Barkley, you're wrong.

Correlation has a very specific definition, which the first post - and your post above - have not bothered to address.

How is the gun the sole motivator of the suicide? There's incidents of it happening, yes, but it's your burden to show how banning guns will stop suicide..

and all we have is whimsical statements about how a person with a gun killed themselves. The problem is SUICIDE, not guns.

If you want to crack down on gun ownership based on the suicide rate, you get to make the connection.. not stipulate that there is a connection (and invent statistics that never existed to support your erroneous stipulation) as the first step in proving that guns are the main reason for the suicide, not the person's overwhelming unhappy condition which would have led to suicide in any other form.

No post yet has addressed the issue.

Random stories don't show correlation. The next extension of this is to find one person who was foreclosed on and make a sweeping policy generalization about every foreclosure in the country.

How does your personal story "prove" your point yet mine is wholly irrelevant? :)

Suicide happens more frequently in areas where low happiness exists -- gun ownership has no relation to the discussion.

rosserjb@jmu.edu said...


Your misrepresentation of the Second Amendment, which even refers to a "well-regulated militia," suggests that you are not worth bothering even replying to, but I shall simply note that your statement that I am wrong about correlation and causation is itself wrong. I do not know how much statistics or econometrics you have studied, but allow me to assure you, I am the wrong person to pick a fight with about such matters, and definitely the wrong person to make such an ignorant and stupid remark to.

Needless to say, it has nothing to do with guns being the "motivators" of suicide. It is that people, both men and women, often attempt suicide on an impulse at a point when they are maximally depressed. If they survive, they overcome their depression, especially if they can get some help (as my friend has), and do not try it again. But, there is no reversing it if they succeed, and there is no question that success occurs at a much higher rate if one uses a gun than if one uses some other method, and one is a lot more likely to use a gun if one owns one. All this is pretty straightforward and obvious and in full agreement with the available data in the US.

You claim that my personal anecdote is irrelevant, but yours is not. But you did not provide a personal anecdote. Did you know somebody who was not allowed to own a gun who then went and said, "I don't need a gun, watch me jump out the window," and then did. Something like that is what you need.

As for suicide happening where low happiness exists and being disconnected with gun ownership, I simply remind you of Rhode Island, second highest state in the US in adult depression, but fourth lowest in gun ownership and fifth lowest in suicide rate.

I could say a lot more, but I think this will do.


Unsympathetic said...

Or possibly you don't want to actually define and defend a position.. so that you can move the goalposts wherever you'd like to declare victory.

Correlation and causation have zip to do with each other. To say something is definitively "caused" .. it's up to you to establish that relationship, not simply use the words in the same sentence. But keep talking big on teh Interwebz - I'm sure someone cares!

Are guns bad because they kill people on their own? Or is depression bad because it might involve reaching for a gun? Or are guns bad only when depression is involved? Or is depression bad by itself? Or is the conflation of depression with guns a bad thing?

Until you find a published study showing people kill themselves with guns when depression isn't present, you haven't shown that guns CAUSE people to kill themselves. You've shown that guns and suicide exist in the same areas. Congratulations! Unfortunately, you've thrown numbers out there.. that aren't linked, because you haven't shown how you're correcting for the fact that the suicide attempt was inevitable with or without the gun.

Suicide attempts, additionally, have no bearing on gun ownership, because the relevant statistic is the number of suicides per gun in the US, not raw number of suicides. It's not the fault of safe gun owners that a very VERY small percentage of people are stupid.

By way of comparison.. are you proposing to ban all cars and trucks from the road due to the number of teenagers who drive poorly?

Depression causes suicide, nothing else. Guns aren't relevant.

rosserjb@jmu.edu said...


You are seriously incoherent. For starters, you are the one who started using "correlation" and "causation" in the same sentence. I was careful to avoid this precisely because they are so poorly, well, correlated. However, I will note that when there is a causal link between two phenomena it often shows up in some sort of correlation in actual data, whereas that there is a correlation in data between two phenomena, does not prove causation.

Again, it is very, very hard to "prove" causation of anything. Do we really know for sure that smoking causes lung cancer? I and most people by now think so, even though there are people who smoke and never get lung cancer and people who get lung cancer and never smoked (or were around second-hand smoke). Although they have mostly faded away, there were some very well paid statisticians who essentially repeated over and over that the strong correlations in the data between smoking and lung cancer did not prove causation. And, they were right, technically. But, so what?

Regarding the case of happy people who have a lot of guns committing suicide a lot, I remind you of youth suicides in South Dakota, who have a suicide rate twice the national average, the third highest rate of gun ownership of any state in the US, but the second lowest rate of youth depression of any state.

I suspect that you are one of these gun nuts who will just run in circles endlessly no matter what is said, especially given the incoherence of a bunch of what you have said (just what does the Second Amendment say again???), but I will simply point out again: it is a whole lot easier to kill oneself with a gun than by pretty much any other readily available method. There is no particular reason to believe that there is some direct link between depression (or some other fundamental tendency to suicide) and gun ownership, even if there might be some correlation due to other variables, such as "rural isolation" or whatever. So, ceteris paribus, there is every reason to believe that once we correct for the sorts of exogenous factors such as national cultural differences, we would expect that having more guns around will in fact cause more suicides to happen because more people will use them when they try, and we know that using them leads to greater success rates of suicide. No, we have not proven a causal relationship, but we see very strong reasons for ones, and the cross-section data in the US is very strongly in accord with the supposition, even if it does not "prove" it.

Finally, let me ask you again. How do you explain Rhode Island's low rate of suicide and gun ownershop coinciding with its high rate of depression, while South Dakota's low rate of depression coincides with its high rates of gun ownership and suicide? If you cannot come up with a sensible explanation for these data, I suggest you should call it quits.