For how long will the U.S. suicide rate remain elevated?

In the United States, Julie Phillips, a sociologist at Rutgers University, was among the first researchers to frisk these middle-age suicides for deeper meaning. In 2010 she and a colleague declared the age range a new danger zone for self-harm. Many commentators took this as another fun fact about the boomers, not a cause for general alarm. But earlier this month, Phillips presented the results of a second paper, an attempt to settle the question of whether the boomers were especially suicidal. She sifted through eight decades of U.S. suicide data, wrenching it to separate the influence of absolute age, peer effects, and the events of the moment, and she found something shocking: the boomers have the highest suicide rate right now, but everyone born after 1945 shows a higher suicide risk than expected—and everyone is on pace for a higher rate than the boomers.

Here is more on that topic.  There is also this:

In her next bundle of research, Phillips hopes to pinpoint the massive, steam-rolling social change that matters most for self-harm. She has a good list of suspects: the astounding rise in people living alone, or else feeling alone; the rise in the number of people living in sickness and pain; the fact that church involvement no longer increases with age, while bankruptcy rates, health-care costs, and long-term unemployment certainly do.

I would think also that these days committing suicide involves less shame than it used to.  Here is one of the cited papers.  Here is her home page.


This reminds of the Dixit/Pindyck assertion that committing suicide when the expected value of your future utility is negative is not necessarily optimal. The reason is that you must account for the option value of living, and that your expected value of future utility may become positive again.

In that sense, a social stigma against suicide helps to prevent unnecessary suicides amongst those who are unable to price options properly.

I don't mean disrespect, but this is a really assertion. The expected utility of your future, if defined properly, includes the option value of living.

silly assertion*

Yes, you are correct, excuse my mistake. Where I wrote "expected value of your future utility" I actually meant "net present value of your future utility".

In that sense, a social stigma against suicide helps to prevent unnecessary suicides amongst those who are unable to price options properly.

This is technically true, but one-sided: It also prevents suicides that would have left the person better off according to his or her own values. Without further evidence, you have no reason to assume that one effect is bigger than the other.

Also note that the stigma leads to legislation that actively reduces options, such as voluntary euthanasia or availability of reliable painless suicide methods. What is the value loss of these options caused by the stigma?

Not to mention personal resentment of people who take autonomy violations personally: The official involuntariness of my life and work has been the single most important factor in reducing my motivation, solidarity and subjective quality of life, which in turn led to a considerable increase in the probability that I will commit suicide in my thirties.

I strongly suggest policies that start with respecting autonomy and then work with voluntary options.

Does the Catholic Church still refused to bury suicides in consecrated ground? I guess if you're cremated it makes no difference.

There is no longer a categorical rule. As I understand, the theory was that suicides by definition died in sin, because their final act was a mortal sin. More modern interpretation is to allow that he may have repented of the decision to commit suicide prior to the actual death - eg, after jumping off the building, but before hitting the ground.

With the sort of stuff that's shadowing the Catholic Church right now, where to bury sinners ought to be pretty low on their priority list I'd think.

Suicide rates are typically lower in Catholic states and higher in not very religious Protestant northern western states.

Perhaps southerners are less likely to report suicides as suicides?

They are also higher in Protestant northern European states than southern ones, I believe. It might be the long cold winters, or it might be the culture (emigrants from northern Europe tended to settle in the northern US)...

How long: As long as we have wars and returning soldiers and do not have Obamacare per the Oregon study.

But, why stop at suicide when citizens do not create tornado proof schools. I guess it's not suicide when its not you, but your kid or neighbors kid.

Are you serious? Have you seen the pictures of telephone polls driven through two-feet-thick concrete walls? There are some things that can't be done.

Really, are YOU serious? Ever take a physics course. Or is it your belief that materials cannot stop the force of storm because it was in the Bible.

How much would it cost to fully tornado proof all schools in the US?

All schools are not necessary. The question is not total cost, but the loss of your child. How much is that worth to you if you live in Tornado alley.


How much would it cost to fully tornado proof all schools that you think are necessary?

Rahul, I don't see the logic of your point. Sometimes people make a broad statement--how much would it take to cover all X--without asking the question in the particular: how much would it cost in my school, or for my grandchildren. The choice framework that someone choose is not by accident: one tends to discourage, the other tends to force rational thinking.

Which path do you think you are on in the way you framed the question. Ask this question: How much do you think it would cost to immunize the American population for small pox versus how much would you pay to save your child from smallpox.

> Ask this question: How much do you think it would cost to immunize the American population for small pox versus how much would you pay to save your child from smallpox

I guess you mean the second question to be how much you would pay to protect him before you know your child has smallpox - ?. It presumably
can't be a large sum of money!

It would be helpful for you to give an answer yourself: how much would _you_ pay to protect your child from the threat of a future tornado striking while he/she is at school. (I'd be somewhat surprised if the answer is more than a few tens of dollars, but perhaps I'm wrong.)

bxg, It's how you frame the question. Let me give you some insight: If I frame it as a $12 increase in your property taxes for 30 years...would you pay. If I framed it as $360 per household, would you pay. Often, when people try to aggregate a number across all X, for all X, as a lumpsum, they frame it to be dismissed.

Would you pay a $12 increase in property taxes? Would you pay a one time $360 assessment.

How you frame it matters.

If I framed it as: this building would remain standing so you could house your familiy if you lost shelter, would you pay?

bxg, Here is some further information:

With respect, I'm not sure what "information" one could clean from being referred to an article that uncritically cites such
unthinking gibberish as:

> Would you own car without seat belts or an airbag? Would you drive your one year old baby to day care without a proper child seat?
> NO because we want our family and loved ones to be safe at all cost!

You say:
> If I framed it as: this building would remain standing so you could house your familiy if you lost shelter, would you pay?

I get (I think) your point (am I right?) - that framing of the collective cost vs the individual is important. But you seem to also lean to framings that
elide whether something has happened vs made beforehand. How much do you think people should pay to have tornado proof schools
(you threw the answer back at me, but didn't answer yourself.) Whatever your answer, what is it's cost per expected life saved? If it's
>10M$, we have a problem. Or maybe not, but surely a disagreement. If it's >$100M, I think you _do_ have a problem. But I have no idea
what you think.


I'm trying to be rational here and I don't see your point at all. Are you really saying that project selection (when it comes to safety) is a cost independent venture?! Let's say the cost of protecting from highly unlikely tornadoes came to 30% of school budgets they should still go ahead?

Here's how I would do it (as a policy planner): Putting ideal tornado protection on all schools would save about 20 lives in two decades among the ~15-25 millions students that go to school in the tornado belt each year (tell me if you have different numbers).

As a very ballpark estimate, use $6Million as the value of a statistical life and tell me how much of a capital investment would be worth it.


Re your comment: "Let’s say the cost of protecting from highly unlikely tornadoes came to 30% of school budgets they should still go ahead?"

Let's say elephants are pink and they dance in your garden at night. When do you get to make up facts?? Is this an essay contest where you wish facts when you can't support an argument?

From an Oklahoma newspaper on a new school that is tornado proof:
"Locust Grove said the domed-design cost $2.5 million less than a traditional metal building, and the district estimates the super-insulated domes will save $25,000 a year on electric costs alone.

"They're the most highly energy efficient building you can do and they're rated an F-5 tornado, so an F-5 tornado can hit this building, and it's gonna still be here," said Superintendent David Cash. The campus can also serve as a community shelter in a worst-case scenario.""

Another source puts the costs for the safe room at $140 to $250 per square foot :

Don't forget workplaces, etc. My office isn't tornado-proof, and the stores I shop at aren't. If you want to make a stupid point, make it comprehensively, so no one misses it.

OK, I'll take you up on it: Your kids school should not be tornado proof, because your office isn't. Afterall, if we cannot do all X, we shouldn't do any X.

No. Better analogy: If you can spend $100 doing X which is more effective than Y we should spend on X.

Your tornado-protection kitty would be much much better spent on, say, anti-drug-education, better nutrition, more sports, reading tutors, scholarships, drivers-ed and many other things that seem a lot better use for these millions than some silly project that will save one kid a year from a scary tornado.

But what if the cost wasn't so big?

I read an article saying that the difference between 2 schools hit by the tornado was that one had an open ground so the kids ran out of the buildings and weren't inside when the school collapsed while they didn't in the other case...

Or somesuch. My point is this: If a bit of thinking about school construction can reduce the risk, it's not exactly expensive, at the very least for new schools.

(OTOH, open ground may good for tornado but bad for other things so I am not exactly reaching great conclusions... Just that, in tornado alley, just like in earthquake prone places, it wouldn't be a bad idea to think before building).


My post above this one shows the costs. Of course we live with opportunity costs, but that's not the point. You don't even have the costs, which I list in the post above.

There is an opportunity cost for responding to this message as well.

@Fredrick Mari

"But what if the cost wasn’t so big?"

Then we should go ahead and build them. Which is why I started out asking how much it would cost. I just found Bill's initial assertion puzzling that "it doesn't matter how much it would cost!"

I agree with the rest of your points but you very reasonable "Let's build new schools wisely (i.e. tornado safe)" is very very different from someone saying let's retrofit all existing schools to make them tornado proof.


First, you make up facts with the statement: "Let’s say the cost of protecting from highly unlikely tornadoes came to 30% of school budgets they should still go ahead?"

Next you state something I didn't say: "I just found Bill’s initial assertion puzzling that “it doesn’t matter how much it would cost!”

Not true: I dare you to point to that "initial assertion" that it doesn't matter how much it would cost. I talked in terms of framing, all X v. no X, and price over time, and then showed you the costs, which you assumed to be 30% of a school districts budget to advance an argument.

When you are losing an argument, you seem to make up "facts".

Economic uncertainty is another big one, at least according to the father of sociology. Busts and booms both increase the suicide rate.

Incidentally, if current trends hold, depression is set to overtake cardiovascular disease as the leading cause of death in industrialized nations.

In how many years?

Approx. 700,000 people die every year in the US from Cardiovascular disease. Approx. 40,000 die from suicide.

I'm skeptical of your assertion.

"Did you know that disco record sales were up 400% for the year ending 1976? If these trends continue... A-y-y-y!"

My vote is the ever-increasing prevalence of recreational drug use. Color me un-cool.

You are not including alcohol?

I am referring mostly to drugs that leave a lasting impact on the dopamine/serotonin pathways. Alcohol doesn't qualify, nor would we have any reason to believe changes in alcohol consumption would have occurred at about 1945. Marijuana, cocaine, amphetamines including MDMA, and virtually ever major recreational drug including Rx drugs that are sold on the street all have similar effects on how the brain uses/expects to use dopamine and/or serotonin. (Often times both.)

It's just a crackpot theory I have, but it's an informed one, based on how these drugs interact with brain chemicals, and how those brain chemicals relate to human moods, expectations, levels of depression, etc. etc. etc. I'm no doctor, but the circumstantial evidence is strong enough to warrant that people avoid these chemicals like the plague.

What on earth would make you think that drug use has gone up? Cocaine, opium, speed, all used to be free. You could buy coke in the Sears catalog. I'd need to see some sort of proof that the use of cocaine and amphetamine has increased significantly over the last century.

Income has increased over the past century. Is cocaine an inferior good?

Matt -- The price change might be too large for that to matter. There didn't used to be a massive federal bureaucracy devoted to making cocaine hard to get and dangerous to possess. 10,000% profit margins don't make themselves!

...very weak evidence that the suicide rate is "elevated".

And highly arbitrary choice of the 'baseline' time period for comparison (..elevated over what ?)

CDC basically says the suicide rate among Americans aged 35–64 years increased 3.9 percent (from 13.7 per 100,000 population in 1999 to 17.6 per 100,000 in 2010). That increase is trivial, even if accepted at face value. But does anyone really believe the suicide-rate can be accurately measured to a tenth-of-percent (?) -- that premise is absurd. The real-world margin-of error is easily +/- 10% ... and likely much larger.

And why 1999 as the baseline ? Why not 1979, or 1929, or 1889.... etc (??) -- if boldly forming such grand conclusions about suicide rates.

Also, "recorded" suicides are a dicey data point. Even today the social stigma of suicide is very strong... with many suicide Death Certificates avoiding that term as cause-of-death (...uncle Bill was just cleaning his gun).

So CDC doesn't enlighten us much about suicides. And exactly what would we do if we had perfectly accurate suicide statistics ?

Perhaps suicide rates are already astonishingly low, in the grand scheme of human existential being.

Indeed. The Bahamas, a nation of 350,000, typically records a single-digit number of suicides annually. It's a conservative society with a strong bias against acknowledging uncomfortable facts. A more honest accounting would almost certainly change this graphic:

Another consequence of "the sixties"?

"We have yet to reap the so-called benefits of the Acid Generation."
-- Frank Zappa


An amusingly self-contradictory melange of bad statistics and factually wrong statements.

When will it become the leading cause of death? Self driving cars and medical advances might mean that day is closer than expected.

Of course, that is a very good thing, not a bad thing.

Hypergamy kills.

Bzzzzzzt. We were looking for "the Mexicans." "The Mexicans."

Better luck next time in the Marginal Revolution Comment Thread Blame Game.....where you've always got a 1/3rd chance of winning!

Every human being on this Earth has a right to end their own lives at the time and in the manner of their own choosing. We need to learn to respect that right.

I agree. Since you are the owner of your own body and life, this right is as essential as your right to life. Unfortunately, not everybody sees it that way.

That's all fine in principle. But sometimes humans aren't temporarily rational.

I read somewhere that they interviewed "jumpers" that survived suicide attempts and many said that in the short moment after they jumped and before they blacked out they suddenly realized how fixable their life was.

What do you say to that?

It's a) anecdotal, and b) obvious bullshit.

Do you seriously expect people to have insightful analytic thoughts during a fall? And then correctly remember them after the fact?

Even worse, society expects you to not consider suicide. So those who still do will keep their mouths shut, or lie about it. Your freedom and legal autonomy rights are both contingent on this lie, after all. So all you get is this anecdotal evidence from people who happily proclaim that they are now glad and grateful that their original intentions didn't pan out.

But hey, I even believe those. People make mistakes all the time. I bet a lot of suffering people wish they had killed themselves when they still had the chance. I'm not suggesting that everybody jump off bridges at a whim.

But you could have a system where the best suicide drugs are available for everyone at cheap cost, under the condition that they simply wait for a couple of weeks. In the meantime, you can give them voluntary counselling options. This would not only establish a basic legal right to opt out of life, it would also prevent whimsical suicides by creating an alternative to unreliable methods, which introduces a time delay and offers a chance of mental health professionals to intervene on a consensual basis.

I'll admit that it's anecdotal but I don't think it is bullshit.

Nevertheless, I do like your last idea with the caveat that I'd make the cooling off period many months, or a year or two rather than weeks. I sympathize with the idea that for some people it may indeed be the only solution but I think that's a rather small fraction of all suicides.

Most suicides are sad because they are impulsive, irrational actions in response to a hard phase. With a bit of support they could've gotten over the bump.

"Nevertheless, I do like your last idea with the caveat that I’d make the cooling off period many months, or a year or two rather than weeks."

And during this time, will you be the one who pays the cost of living for these people? Or are you going to expect them to accept forced labor or homelessness while they are not even here voluntarily any longer?

I agree with respect to public policy, but disagree from the standpoint of ethics. I will respect a person's right to make that choice so long as they will also respect my right to attempt to verbally persuade them otherwise.

I think the linked article suffered from bad statistics and also no proper consideration of correlation vs causation bias. For instance they quote higher suicide rates for lower income people, implying that envy is maybe the cause of some suicides. But equally it could be that people who attempt suicides have some tendency that also lowers their income (maybe they are more prone to hyperbolic discounting due to lower IQ). The researchers theory also seems very vague, in his Venn diagram of the three attributes he says are needed to combine to cause suicide, he quotes loneliness as one factor, how is that measured? What is the strength of that factor? Could it be that the factor that also causes loneliness also causes suicide, many mentally ill people are unpleasant to be around and so that could account for both loneliness and suicide with one factor. Its hard to see how we could address suicide with such a weak theory.

Given that the will to survive is genetic (anyone who didn't have this gene wouldn't leave many descendants) it is inherent in the design of our brains. My suggestion is that a problem with this module (either in the design or in the operation, perhaps due to stress chemicals, or perhaps a pathogen) will result in suicidal behavior.

Like other commentators, I am not convinced we are seeing a rise in rates so there is nothing really to explain in terms of modern life causing higher rates. This in itself is an interesting result, since we have such a different life today than say 20 years ago (at least in the West).

to show some correlation, this recent Gallup poll shows the de-stigmatization of suicide in the past 12 years: a 3%-age point increase in the 'moral acceptability' of suicide:

ht: @DouthatNYT

Note that the rise is really among white middle aged men. Others are not killing themselves even though their lives on the whole are worse.

I wonder if there is any correlation between the war on tobacco and suicides? Cigarettes were a cheap stimulant and anti-depressant. But now are too expensive.

I think this is actually something that's been known since the early 1990s. I remember reading a study that showed similar results. Apparently this is one of those "you value more what you've worked hard for" propositions; when people had to put a lot of work into staying alive, it was worth more to them so they clung to life so as not to have wasted the effort. IIRC they mostly looked at people who had lived through famines.

I've been browsing online more than 4 hours today, yet I never found any interesting article like yours. It's pretty worth enough for me. Personally, if all site owners and bloggers made good content as you did, the web will be much more useful than ever before.

Comments for this post are closed