This week, I heard a member of the Obama Administration assert that “22 veterans” commit suicide in America every day. That’s more than 8,000 each year.
For 2013, the total number of deaths by suicide recorded was 41,159. Thus, putting the two numbers together, about one suicidee in five would be a veteran.
I find this very hard to believe, knowing that suicide affects particularly older people and that most current veterans today are not old.
Am I the only one who is skeptical? More alarmism? More underhanded propaganda against our collective resolve to defend ourselves ?
Update from the Los Angeles Times consulted 2/13/15
[published in the] “February issue of the Annals of Epidemiology, included all 1,282,074 veterans who served in active-duty units between 2001 and 2007 and left the military during that period.”
“One statistic has become a rallying cry: 22 veterans take their own lives each day.
That figure is a national estimate based on a Department of Veterans Affairs analysis of death records from 21 states. Though it is usually cited in the context of the recent wars, most of those suicides involved older veterans, who account for the vast majority of the nation’s 22 million former service members.
Among veterans in the current study, there was one suicide a day.”
(All boldings mine.)
“The rate was slightly higher among veterans who never deployed to Afghanistan or Iraq, suggesting that the causes extend beyond the trauma of war.”
Here is the link so you may judge for yourselves. (It was provided by Martin Anding; I thank him.):
My question has still not been answered but I am making progress and I have learned interesting things in terms of the underlying narrative.
First things first: The fact that there is a little more suicide among members of the military never deployed does not rule out the line pushed by pacifists right and left. The reason is that it’s possible that military authorities are watchful and careful to not deploy those who are in fragile mental health. (This is also noted by a critic of the article on veteran suicides in Wikepedia.) Thus those not deployed may commit suicide for their own reasons while those who are deployed commit suicide for a different set of reasons related to their deployment in a war zone.
Again, this could be true. I suspect it’s not because most of those deployed do not ever see actual combat. So, we are not dealing here with two discrete categories such as: combat/non-combat, but with a sort of near-continuum, like this: zero exposure to combat to much exposure to combat, with everything in between. If this description is correct, it does nothing to sustain the speculation I advanced above about different causes of suicide for non-deployed and deployed military. Most military personnel never hear a gun shot in anger. Only a few are in a position to have been broken by combat. Military personnel commit suicide – at whatever rate – for other reasons. This is not a final judgment. I can imagine additional data that would make me turn on a dime.
The finding that non-deployed military personnel commit suicide at a somewhat higher rate than those deployed does not kill the hypothesis that war is a cause of suicide but it sure as hell does not support it.
Now on to the actual rates. The study that provides the astonishing figure of 22 per day is based on data from only 21 states. So, there could be a big sampling issue with it. Nevertheless, I cannot wrap my mind around this idea: If data for the 29 missing states were included the number of 22 per day would shrink to the point where I, JD, would not pay attention. Frankly, I think that any number close to ten per day would seem too high.
I can think of only one study design that would give this kind of high number (22/day). That would be if the study designers had deliberately cherry-picked the states with the highest rate of veteran suicides and deliberately excluded the states with low numbers. I do not believe that anyone did this (not). Yet, the number 21, as in 21 states out of fifty, leaves me uneasy. There may well be a legitimate explanation in the fine print of the study. So, let’s not jump to conclusions
(Continued the next day.)
I have now spent more time on this issue than any ordinary s citizen should be required to do in order to decide whether a startling figure from his own government is to be trusted. Nevertheless, I am left with two incompatible estimates of the frequency of veteran suicides.
The two figures, 22 per day and 1 (one per day) are too far apart to be reconciled, according to my experience. One study of the two studies is flawed, or both are. I smelled a rat the first time I heard the figure of 22 per day, that rat has only become stinkier.
The two figures could not be that far apart if both studies were honest and competent. That’s because suicide is a a fairly well defined act. I am aware that there is a tendency, especially among Catholics, to camouflage suicide as natural death. I think it’s waning and plays a small role nowadays. If I am wrong, the most likely effect is an underestimate in both figures, 12 and one. This would not affect the gap that bothers me. I mean that suicide is “well defined,” as opposed to say. job discrimination based on race, sexual harassment, or date rape.
If I had to chose a rate, I would consider the sources. The first figure comes from a government agency, the Veterans Administration which is a natural lobby and has experienced a lot of political trouble recently. In addition (in addition), it’s part of an administration that has repeatedly offered us fanciful and grotesque figures. I repeat: a recent Pentagon study claims that 40% of victims of sexual abuse in the military are males. The lower figure, on the other hand, comes from a study published in a scholarly journal. Because of this publication, I trust that it has been subjected to a fair degree of critical appraisal through peer review. Peer review is not a panacea as scandals associated with climate change pseudo-studies have shown, also the fake medical study at the root of the widespread belief that autism is tied to vaccination. Yet, peer review as practiced in most American and British scholarly journal is a lot better than nothing. (See my essay on peer review, on this blog: https://factsmatter.wordpress.com/?s=peer+review ). To pose the question this way is pretty much the same as to answer it. The figure of one suicide of veterans per day is more likely to be correct than the figure of 22.
It seems to be that I should now allow my ideological preference for the latter figure deter me from using common sense and everything I know about quantitative research. It should only make be careful and open to criticism from those of a a different mind set.
Those who wish to push an avoidance of war agenda should not do it on the basis of dubious figures proffered by an administration now known for its light relationship with facts. It seems to me also that the undebatable draw on treasury of every war, the obvious fact that some soldiers die in battle, and the inevitable occurrence of “collateral damage,” the killing and maiming of innocent civilians in theaters of war should suffice. Pacifists who are also libertarians might also consider shouldering the task of explaining to the great unwashed masses of their imagination how every war inevitably enlarges the scope of government.