Bryan Caplan  

An Unanswered Question on the Economics of Suicide

So Many Narratives... More on Matt Yglesias...
What don't more people with painful, terminal diseases commit suicide? 

Religion dissuades some of them - that's the point of Hamlet's famous soliloquy.  But most people quickly ignore religious teachings when they seriously cramp their style.

Some people might worry about hurting their families.  But is suicide at the end of a life well-lived really harder for your family to bear than watching you die a slow, horrible death? 

As an economist and a movie buff, the explanation that naturally appeals to me is that people refrain from suicide so their families can collect the life insurance.  But the idea that life insurance won't pay on a suicide is largely a myth.  My friend in the industry and all the sources I can find say that the suicide exclusion only applies for the first two years after the initial purchase of the policy.

The only explanation that I can't easily reject: By the time you want to commit suicide, you can't do it on your own anymore.  That's the premise of The Sea Inside, and it makes some sense.  For this story to work, people have to be extremely impatient, unwilling to sacrifice their last good days in order to avoid horrible pain in the foreseeable future.  But I can buy that.

Other explanations?

Update:  Several people point to the evidence that people with severe disabilities still have pretty high life satisfaction.  But as far as I know, these are studies of people who have lost capabilities, not people who live in constant, severe pain.  I can believe that people get used to being a quadriplegic.  I can't believe that people get used to physical suffering.  Furthermore, if you check with the GSS, subjective health does have a strong effect on overall happiness.

Comments and Sharing

COMMENTS (35 to date)
Matt Simpson writes:

Humans who had a tendency to commit suicide were far less likely to pass on their genes than those who didn't, so as time went on people evolved a preference against suicide. They are simply acting on this preference later in life.

Of course, they could have just as easily evolved a preference for waiting to have suicide until after they have mated. I'm not sure how to deal with this one.

Les writes:

For more information see this link:

OneEyedMan writes:

My understanding is that the disabled routinely rate their lives as much higher than outsiders so. Something similar could be in play here. It could be that many lives that seem from the outside as unbearable are better than nothing.

Joe Marier writes:

Actually the point of Hamlet's famous soliliquy is "I need to act crazy and suicidal to confuse the guy I'm attempting to kill."

Kevin writes:

As I understand it, religious affiliation is a strong predictor of whether one will commit an assisted suicide.

Zac Gochenour writes:
"But is suicide at the end of a life well-lived really harder for your family to bear than watching you die a slow, horrible death?"

While this question is ostensibly rhetorical (the supposed answer being no), I think the answer for most people is yes. Suicide is considered by most to be immoral and dishonorable, and for instance the Catholic tradition claims it damns your immortal soul to hell.

Say a Catholic man was dying of stomach cancer and had only months to live. His suffering level is high, and say he even goes against the grain of becoming increasingly religious as he approaches death and decides to abandon his religiosity in light of his extreme misery. He knows, however, that his Catholic wife and children have not abandoned these beliefs. He also knows that by killing himself, they will spend the rest of their lives thinking he is in hell, maybe never fully getting over that depression. So he decides to simply bear the pain and save them the suffering, probably telling his family everything is okay, and that he will see them again in the afterlife.

Further proof that religiosity is one of the worst, most despicable diseases of human culture.

Joe Marier writes:

So do you support the overturning of religious freedom protections in the Constitution?

Pat writes:

Dan Gilbert - "Stumbling into Happiness"

People who actually have bad things happen to them are happier than we imagine we would be in their situation.

Joe writes:

Most people fear change. You really can't get a bigger change than death.

It takes time for person dying to comes to grip with the reality and time for the family to come to terms with it also. By the time that has occurred they are already living in their new condition. They may be in pain, but they are now used to the new reality.

Blackadder writes:

I think Matt is on the right track. The instinct towards self-preservation is incredibly strong. It's much stronger, for example, than the sex drive (which itself can lead people to engage in a lot of seemingly irrational behavior).

John Jenkins writes:

Lack of certainty. Many people might be willing to die, but lack the access to effective methods (firearms) or distrust the efficacy of the methods to which they do have access (e.g., drugs might kill you, but might also just cause pain and suffering).

JP writes:

Some people might worry about hurting their families. But is suicide at the end of a life well-lived really harder for your family to bear than watching you die a slow, horrible death?

Yes, it's harder.

Dr. T writes:

The self-preservation concept mentioned above makes little sense: the person already faces a painful, lingering death.

Some people in this circumstance avoid suicide because they don't want to disappoint loved ones. They don't want their loved ones to say "I still had more to tell him, and I'm angry that he killed himself."

Other people are just too afraid of dying to kill themselves.

Some people won't commit suicide due to religious or philosophical beliefs.

Other people just keep postponing. They plan to commit suicide when the pain and degradation gets to be too much, but they wait too long and become incapable of committing suicide unassisted.

Sarge writes:

"But most people quickly ignore religious teachings when they seriously cramp their style."

Very true, but I don't think that on this basis we can dismiss the possibility that religion prevents people from commiting suicide. Many religious people act in a way their religion forbids by assuming they will live long enough to get back on the straight and narrow. But for obvious reasons, suicide doesn't allow this kind of wiggle room, so this rationalization fails and you're pretty much screwed.

barghest writes:

Poeple do not seem to want to kill themselves, even when it seems unreasonable to remain alive. The question is why. I find the thought that those with an irrational desire to live, are more likely to do so and hence more likely to reproduce thereby causing a greater fraction of the population to possess this trait. But, can an irrational desire to survive be transmitted genetically this way. It seems that hunger, fear, sexual desire, the way people respond to having water enter their lungs, are easily explainable this way. But does it feel that good just to be alive? It could be that the reason people do not often kill themselves is largely social pressure from adaptive social norms, the problem being that sometimes people for selfish reasons, or no particularly good reason at all decide to off themselves. But does a group gain any competitive advantage by keeping members around who don't want to be? Is the suicide of net contributers a greater problem than the suicide of net consumers of group resources?

Sister Y writes:

John Jenkins above is correct, I think, in his explanation of why people irrationally continue to live. I disagree that firearms are, for most, an acceptable means of suicide - overdose of barbiturates is the method used in (civilized) states that allow limited assisted suicide, and those are illegal without a prescription and hard to obtain even on the black market. Every other method available is either likely to be painful or likely to fail and render the person less likely to succeed at a later time.

There's a risk - but still, even with the risk, wouldn't it be rational to attempt suicide if the expected suffering of continuing to live exceeds the expected suffering of attempting suicide? My explanation for why people still don't do it has to do with time horizons: it's hard to weigh the total suffering of one's expected life against the risk of suicide; instead, it's tempting to weigh the expected suffering of waiting another day/another week/another month against the expected suffering of a suicide attempt, in the hope that a better method might become available. But it never does. This is irrational.

ericyu3 writes:

Some people might worry about hurting their families. But is suicide at the end of a life well-lived really harder for your family to bear than watching you die a slow, horrible death?

If you prefer for someone to have severe but non-life-threatening pain for the rest of their life over them dying a slow, painful death within a year, then you would probably prefer that over them committing suicide.

Norman writes:

For most people who genuinely believe in an afterlife, suicide (as a forbidden activity) promises to significantly reduce one's flow of utility for the infinite future. Further, this flow in the future will presumably matter more since religious people may also believe that their level of patience, as a virtue, will be substantially increased in heaven, thus raising the influence of expected future utility flows on current decisions. In most cases I don't think it's surprising that current suffering pales in comparison to the lost utility of suicide for these people.

Even for the irreligious, though, expected future utility flows after death are constant at zero. Because you're dead. Suicide in this case only becomes a good idea if expected utility flows while still alive reach zero or negative (that is, continuing to live actually sucks the enjoyment out of previous experiences). I doubt very much there are many cases of even the most painful terminal illnesses which reduce all future utility flows to zero or below.

Shakes writes:

I want to become a heroin addict when I am on death's door. I would like to explore all of the drugs that I was never allowed to do. Heroin, maybe a some LSD. Why not go the gamut and let me do them all and repeat the ones I like if I have months to live. Who cares if I am a junkie in my last 6 months of life?

Take my pain away and put me into hyperspace.

I watched my father die from cancer. I nursed him through the kemo and then he had a relapse and the cancer spread all thoughout his body and into his brain. They gave him some morphine to ease the pain. Thank God for the Opium poppy. I don't like kemo, especially for older folks. I had an uncle go that way too. The family puts an intense pressure on people to fight to live. "You need to fight it! We can win!" and then they go through kemo. Kemo is a bitch and my dad may have lived longer without it.

Don't give me kemo, give me cocaine, heroin, lsd and give me the ride of my life when I am heading out the door. Make it pain free and make it a blast. I think we have the technology! It is all about choices.

Kemo is great for someone in their 50's who is fit and could benefit from beating the odds. But for someone who is in their 70's and not a great piece of physical work to begin with, I advise against it. Maybe if they catch the bug early enough, but please take caution. We are all going to die. Might as well make it enjoyable.

redbud writes:

To address the question of why folks with painful, terminal diseases and vested life insurance don't off themselves, I suggest that pain management and other drugs wipe out personal power.

A drug-free person may determine that a few seconds of pain -- to provide large sums to dependents without a burden of caring for the terminal patient -- are doable.

But how does one be drug-free when doctors are in charge? All traditional medicine knows is cuts & chemicals, right?

Prakhar Goel writes:

God... listen to some of the comments here: "irrationally continue to live...".

Wanting to live is a moral statement. Are the commentators here (and this includes you Bryan) really so absurdly childish that they can't even imagine that some people just really like to live? I.e. that people have different preferences? I would have expected better from a bunch of economists.

Hollywood_Freaks writes:

"Are the commentators here (and this includes you Bryan) really so absurdly childish..."

Yet, you are the one engaging in ad homium attacks.

Pat writes:

Prakhar, not only are commentors having trouble imagining that other people have different preferences, they're also assuming that their predictions about how they would actually feel are accurate. They're not.

Rather than wonder why people don't kill themselves more, we'd be better off paying attention to their revealed preferences and factor it into our own predictions.

Jeremy, Alabama writes:

It would be interesting to find out whether it is more or less common now than in previous decades.

With modern medicine there is always a chance science will find something to ameliorate your condition. Perhaps this lingering hope - compared with the starkness of the alternative - keeps people in the game. But of course, this belief in medical progress is a very modern phenomenon, so it might be a testable hypothesis.

Prakhar Goel writes:


Thats not my primary argument. My argument is: wanting to live is a moral imperative and therefore cannot be considered irrational and people always have different preferences and this is just another example. The rest is an insult to the other commentators on this thread because this discussion is such a departure from the normally good quality stuff that goes on here. Its my attempt to get their attention.

This discussion is also imho, deeply disturbing and disgusting as so many of the other commentators are just blatantly asserting their own view of when it is "rational" to live and die for what is a deeply personal decision and thereby implying that any choice that deviates from this norm is somehow illegitimate. They are in effect denying one of the core aspects of human individuality and yet, nobody has called them on it (except for me).

baconbacon writes:

"Rather than wonder why people don't kill themselves more, we'd be better off paying attention to their revealed preferences and factor it into our own predictions."

"I.e. that people have different preferences? I would have expected better from a bunch of economists."

Preferences, and revealed preferences rely on two things

1. information at hand
2. ability to act.

For the first part people diagnosed with cancer are facing two unknowns- the choice between death and pain well beyond their normal experiences. To claim that one can make a truly informed decision in such a case is hubris. It does not surprise me that many people take the non choice (living and suffering) over the choice of immediate action prior to experiencing the suffering.

As for #2- the pain that comes with cancer also comes with exhaustion, dehydration, and the disconnect that comes with painkillers. Presuming that one is able to reveal their preferences under those circumstances is dubious at best as well.

In terms of understanding the decision faced by individuals I think it must be best to look at those who have experienced some similar level of pain- those who survived an earlier cancer or are relapsing to begin to understand the issues at hand.

Snark writes:

[Why] don't more people with painful, terminal diseases commit suicide?

As silly as it may sound, they're probably afraid of what the neighbors will say.

Statistically, 3 times as many women attempt suicide as men, while 4 times as many men succeed. Women may be more rational, but men are more reliable.

Bill Abbott writes:

Though not as good as The Sea Inside, another good movie that deals with this (and other issues that appeal to my libertarian sensibilities) is the French film Invasion of the Barbarians:

ZH writes:

Avinash Dixit and Robert Pindyck in Investment Under Uncertainty actually briefly discuss rational suicide while critiquing the Hammermesh and Soss rational suicide model. Dixit and Pindyck point out that the ability to commit suicide is actually an abandonment option on your own life and that as with most abandonment options it is usually better to wait until things get extremely bad as opposed to commiting suicide when expected net utility of life is negative. Now even under Dixit and Pindyck model, it should still be rational for a terminally ill patient who is not religious to commit suicide. However, if we add in cognitive biases, namely that due to anecdotal evidence of the terminally ill being miraculously cured (possibly due to new experimental cures) the terminally ill overestimate the probability that they will miraculously survive, and therefore the terminally ill overvalue the option, leading them to not commit suicide.

Alex J. writes:

People don't like giving up. Russ Roberts had a discussion with Tyler Cowen on an econtalk about how it was difficult to get over his compulsion to finish every book he started. TC pointed out that we want to seem to ourselves and others that we see things through. Perhaps we take it to excess at the end of life, but that wouldn't be selected against, if you're going to die anyway.

Lane C. writes:

I agree with Norman above: Utility with terminal disease > utility when dead.

I also think Brian is right about suicide is much harder on the family than the expected death, primarily due to one being sudden and the other expected. One could mitigate this though by informing the family you plan to commit suicide.

I think the single largest factor though is that people with terminal diseases still hold out hope they might get better. Lifetime utility is a LOT higher under the no-suicide plan if you even assign a 1% probability to the disease going away.

Sister Y writes:

Snark mentions the different rates of male and female suicide attempts and successful suicides. I think this is highly relevant here.

People always want to explain the male/female difference in suicide as primarily motivational. I'm sure to some degree it is, based on evo bio factors. BUT, there are a few pieces of data that make this explanation suspicious:

1. Female physicians commit suicide successfully at almost the same rate as men.

2. Suicides in the high-stress finance sector are not elevated, but suicides are elevated among doctors, police, veterinarians, and chemists.

3. Females in countries where lethal poisons (pesticides) are easily available commit suicide at much higher rates relative to men than in countries where they are not available. In some of these countries, female suicides exceed male suicides.

4. More men are familiar and comfortable with, and own, guns than females.

My argument is that much of this - unusually low suicides of terminally ill and unusually suicides of women - comes down to lack of access to an appropriate method, rather than lack of motivation.

Bill writes:

What kind of illness are you thinking of that would cause intense 10 out of 10 pain even with modern medicine's pain killers? Although extremely uncomfortable, physical pain seems to me to be a relatively minor reason for wanting to die.

I could see suicide to escape being tortured to death. But being really nautious and having intense and aches and pain, well, the pain lets you know you're still alive. If you kill yourself you will never have pain, but only because you will never have another thought.

It makes sense to me that many suicides are of people suffering mentally, and not physically. The mind can overcome what the body tells it.

George writes:

Brian wrote: But most people quickly ignore religious teachings when they seriously cramp their style.

That's astonishingly glib. Do you really think that coming face-to-face with their own suffering and death makes people less concerned with religion?

Does the Book of Job ring any bells? Or the roughly half of Buddha's teachings that deal with suffering? I don't know my suras and sutras, but I'm betting there's a lot about suffering and death in those, too.

The post goes downhill from there, while still leaving a substantial stretch of hill for the comments to descend.

Chip Smith writes:

Prakhar Goel,

In what sense is "wanting to live" a "moral statement" or "moral imperative"? Unless it is bound to other obligations, a desire to live seems to me to be a simple expression of a preference that may be either rational or irrational. Where is the moral content?

Can "wanting to die" be a "moral statement" or "moral imperative" as well?

Comments for this entry have been closed
Return to top