Bryan Caplan  

An Evolutionary Model of Depression

New Interview with Milton Frie... The Marginal Myth...

Depressed people - what is their problem? Edward Hagen of Humboldt University has a fascinating answer: Getting depressed is a good way to get the people around you to give you more for less. Feel underappreciated? Then mope around non-stop, and the people who depend on you will pick up the slack. Hagen explains it with a tidy economic metaphor:

Given that the principal cause of major unipolar depression is a significant negative life event, and that its characteristic symptom is a loss of interest in virtually all activities, it is possible that this syndrome functions somewhat like a labor strike. (emphasis mine)

The idea is that during humans' evolutionary history (and even today, inside the family), you couldn't easily hire a replacement for a malcontent:

It would have been difficult, for example, for mothers to raise offspring without help from the father and/or other family members; conversely, the fitness of the father, parents, and other family members depended critically on the mother successfully raising offspring. Abandonment of one party by another would have entailed a significant fitness cost to all...

In evolutionary terms, then, getting depressed is basically a way of saying: "Treat me better, or you'll have to get by without my help." According to Hagen, it usually works:

A number of behavioral studies have demonstrated that although depression in one family member prompts negative feelings from other family members, it nonetheless appears to deter their aggressive behavior and to cause an increase in their tendency to offer solutions to problems in a positive or neutral tone and an increase in their solicitous behavior (e.g., caring statements), consistent with the bargaining model.

Post-partem depression is a case in point:

[T]he spouses of individuals experiencing PPD should report increasing their investment in parenting, and in fact they do. Depression scores for one spouse were positively correlated with reports of increasing investment in childcare by the other spouse (Hagen 2002). High levels of help from spouses and better interactions with infants in one study were also the only variables associated with remission of PPD...

How can suicide be functional? It can't be. But suicide threats can be highly functional:

Suicide permanently removes oneself as a source of valuable benefits for the group. Suicide threats are therefore threats to impose substantial costs on group members and can be viewed as a means to signal cheaply and efficiently to a large social group that it may suffer such costs if assistance or change is not forthcoming. Suicide attempts are necessary to underwrite the credibility of suicide threats and must therefore entail a genuine risk of serious injury or death.

Need I point out how Szaszian this "bargaining model of depression" is? Hagen doesn't cite Szasz. But it would be easy to sum up his paper with a Szaszish aphorism: "They're not depressed; they're depressing!"

Comments and Sharing

TRACKBACKS (3 to date)
TrackBack URL:
The author at amcgltd in a related article titled Speaking of Evolution... writes:
    Econlog is featuring a paper which attempts to link depression with evolutionary and economic pressures. The main gist, as I read it: acting depressed is a way of withdrawing your support from the group in a way that draws... [Tracked on September 5, 2006 10:56 AM]
COMMENTS (26 to date)
Robin Hanson writes:

Of course the most common objection raised to this theory is that depressed people don't feel like they are striking for a better deal. It is hard for many people to accept the idea that the social functions of our actions may be far from the conscious thoughts that accompany our actions.

Dr. T writes:

This paper has many problems that render it worthless as anything other than an exercise in mental masturbation. I will highlight what I believe to be the biggest problems.

Biggest problem: The author fails to distinguish among reactive depression, depression secondary to a primary medical problem, and chronic depression. Treating all these as the same negates many of his conclusions. Reactive depressions are just that: reactions to crappy acute circumstances such as the death of a child. Such depressions cause subchronic or chronic stress which then affects the limbic-hypothalamic-pituitary-adrenal/thyroid/gonads axes. The net effects are hypothyroidism, hypercortisolism, and decreased sexual function. But, as I indicated, such changes can occur with chronic stress without depression, so it is inappropriate to classify them as depression-associated evolutionary adaptations. They are physiologic adaptations to stress which possibly (though I doubt it) confer advantages to the person or to humans.

Second biggest problem: The author excludes all bipolar disorders. Why? The depressive phase of classic bipolar disorder has symptomatology no different than that of chronic depression. The author also fails to acknowledge recent evidence that some persons with chronic depression as young adults convert to some form of bipolar disorder in middle age. This strongly argues in favor of a neurobiochemical model of chronic depression and argues against an evolutionary adaptive model.

Third problem: Some persons with depression successfully hide this from almost everyone around them. In such circumstances, there is no secondary gain or evolutionary advantage.

Fourth problem: Females are twice as likely as males to become depressed. The author ascribes this to females lacking the power of males and resorting to depressive effects to gain favorable behavioral changes from the people around them. The more likely explanation is different neurobiochemistry. Women are more prone to depression, but men are twice as prone to schizophrenia as women, and men and women are equally prone to bipolar disorder. How would an environment of evolutionary adaptedness explain those findings?

Fifth problem: Post-partum depression is a favorite topic of social biologists and evolutionary psychologists who observe that animals abandon or kill their young in response to stress. Those scientists extend their theories about that behavior to humans. But, it is a long stretch to compare a half-starved bird that kills all her nestlings to a well-nourished woman who just delivered a baby and gets depressed two days later. Depression would be maladaptive: cold-hearted infanticide would better aid survival. Depression is anti-survival: the symptoms are apathy and indifference, inability to think and concentrate, slower reactions, etc. The social biologists and evolutionary psychologists gleefully point to the lack of correlation between hormonal changes and post-partum depression and claim that is evidence that post-partum depression does not have a neurobiological cause. What they fail to accept is that people differ biochemically and physiologically, and some are more sensitive to hormonal changes than other. There are well-document cases of acute psychoses that occur when patients receive large doses of corticosteroids. This happens in a small percentage of patients, and it is largely unpredictable. Some men suffer depression after acute reductions in testosterone levels due to anti-androgen drugs or surgical castration. Is post-castration depression an adaptive response? It is quite probable that similar biochemical and physiological mechanisms cause some women to experience post-partum depression or psychoses while most women in similar circumstances do not.

This paper exemplifies the dangers of having a little knowledge.

Matt C writes:

Certainly some depressed people do play their depression for sympathy, but others actively withdraw from social contact when they feel down, or are otherwise too distant from other people to get any social leverage from their depressions.

One could argue that this kind of isolation was impossible in the ancestral environment when the tendency to mopey depression evolved.

Art DeVany has suggested that seasonal depression developed as a way for several men to stay shut up together in a crowded cave during the winter months without killing each other. I think this is an entertaining notion.

I don't see that depression has to have only *one* single function in human evolutionary fitness. Clearly dogs can grieve and feel sad; I'm guessing sadness is a basic mammalian response that has gotten pressed into service for lots of purposes.

Eric Hanneken writes:

I can't answer all of Dr. T's objections, but I'll try to respond to two of them.

Third problem: Some persons with depression successfully hide this from almost everyone around them. . . .

I don't think it's fair to require Hagan to find an adaptive rationale for all of the behavior that depressed human beings exhibit. They do have brains, after all. Some of what they do is chosen, influenced by culture, etc.

Fifth problem: . . . [I]t is a long stretch to compare a half-starved bird that kills all her nestlings to a well-nourished woman who just delivered a baby and gets depressed two days later. Depression would be maladaptive: cold-hearted infanticide would better aid survival. Depression is anti-survival: the symptoms are apathy and indifference, inability to think and concentrate, slower reactions, etc.

Nature selects genes that promote their own survival; the organism's survival is subordinate. If the hierarchy were reversed, no one would become pregnant or spend resources to raise and protect offspring.

Bob Knaus writes:

Am I the only one who thinks a few of the posts on this blog are statements of the obvious? People treat you nicer when they see you're sad... imagine that!! Knowing other people are more successful than you causes envy, possibly spurring you to work harder... who'da thunk it??

What's next? The discovery that being hungry makes it likely you will soon eat? A lengthy paper on the direction of causality between feeling sleepy and getting ready for bed?

Or am I missing the point entirely? Are these kinds of papers and articles really just academic jokes, written as filler between more serious work?

yourcommentsrock writes:

>>Are these kinds of papers and articles really just academic jokes, written as filler between more serious work?

all academic works are 'jokes' in the sense that you need to take them with a pound of salt. Nothing better than collections of interesting ideas that are evidently no more true than anything else.

brian writes:

Clever foray into speculation. I guess that's science?

Barkley Rosser writes:

I would say that Dr. T.'s opening rather overdoes it with his overweening pomposity. Yes, Hagen does not distinguish different kinds of depression and some kinds might fit his story better than other kinds.

However, the fundamental problem with Dr. T.'s commentary is his assumption that somehow neurobiochemical phenomena have nothing to do with evolution. How can one assume such a thing? Where did our neurobiochemical processes come from if not as part of our evolutionary process? If our neurobiochemical patterns lead us to be so dysfunctional that we do not successfully reproduce, then those patterns will not reproduce and survive. Others will.

All "Dr. T." has done is relabel things. So, depression is neurobiochemical. So what? That does not make it something that did not arise out of evolutionary processes or pressures. This is ridiculously misplaced arrogance.

mjh writes:

My wife suffers from depressive flair ups (for lack of a better phrase). This article makes me wonder if I'm the problem. That she needs to communicate "Help!" to me and that the only mechanism she has left is depression.


ChiSean writes:

Actually Dr.T DIDN'T just relabel things. He gave the diagnostic criteria for the different types of depression and then explained in a non-turgid fashion how the writer is incorrect in their thinking.

Barkley Rosser writes:


Yes, he noted that there are different kinds of depression. But he clearly argued that if some kind of depression is "neurobiochemical" in origin, then it is not evolutionary in origin. Last time I checked, the sex drive is a neurobiochemical phenomenon. Do you wish to argue that how it operates has nothing to do with evolution? That is the level of Dr. T.'s argument.

Eric Hanneken writes:

Perhaps I was being charitable, but I assumed that Dr. T would agree that human neurobiochemistry was shaped by evolution. I thought he (she?) was just arguing that depression is not an adaptation, but a pathology that we're vulnerable to as a side effect of the way our brains evolved.

Matthew Cromer writes:

Now I am supposed to believe that depression is really evolutionarily adaptive behavior in disguise even though it often leads to suicide which is almost by definition non-adaptive behavior.

Let's see some evidence in terms of increased life expectancy for depressed individuals, increased health, or increased fertility. I suspect the numbers for each of these will point in the wrong direction.

Epicycles, anyone?

James writes:

I'm always skeptical of these sorts of "Such and so trait that we once thought was bad may have served a purpose and is here today as a result of natural selection" claims. As I understand it, there are burdensome characteristics that reduce the odds of an organism's survival and there are beneficial adaptations that increase those odds. But why suppose that depression is one of the adaptations and not one of the burdensome characteristics that natural selection has yet to eliminate?

Barkley Rosser writes:


That is probably what Dr. T. is arguing, but look at Problem 4. While the evolutionary explanation supposedly has women getting depressed because they lack power, we get instead that they just have a different neurobiochemistry. Well, this just begs the question does it not? Why is it that their n.b.c. is different in a way that leads them to get more depressed?


That we do not see depressed people necessarily doing better than non-depressed people in our society does not mean that it was not an evolutionarily adaptive trait in the context in which our ancestors evolved. Consider homosexuality. Gays do not in general survive better than straights, and they certainly do not personally reproduce better. However, arguments have been put forward regarding how having some gay relatives around may help family members survive and reproduce.


Of course Hagen has not proven his case. It may well be that at best this is interesting speculation. My point was that Dr. T,'s snide remarks did not disprove Hagen at all, although they may have indicated that Hagen would have trouble publishing in a medical journal because of failing to discuss the different kinds of depression (which indeed may well vary in the degree to which they arose from evolutionary forces/pressures, whatever).

Just for the record, one and all. I am not a Dawkins type who sees all biological reality as due to natural selection/optimization. I personally knew Sewall Wright and take very seriously his random drift theory. This in effect allows all kinds of things to appear and then hang around, in effect if they are not too much of a hassle for the species in question, neutral mutations, and for differences in populations to arise as these mutations vary across space due to migration patterns.

What is interesting about Hagen is that indeed there are good reasons to think of depression as not very adaptive, as contributing to an individual not doing well at surviving or reproducing. His argument may be wrong, or speculative, or obvious (to note Bob Knaus). But they do point out that there are possible offsets in which depression could be evolutionarily adaptive rather than what appears to be the obvious opposite case. Dr. T.'s saying that it is neurobiochemical failed to refute argument at all.

James writes:


I wasn't trying to lend support Dr. T's position.

Martin G Smith writes:

I read with constant interest, often amusement and
sometimes anger at the commentary, too often
rhetorical on varying degrees of assessment and
attributed expertise related to mental States
and/or illness. I have in my files a case
transcript which provides verbatim, the query made
of an expert regarding an assessment, and
subsequent diagnosis of Psychopathy. After the
learned expert had given his testimony the Judge
asked, ‘The n sir, i take by the evidence you have
just given under oath, that the finding of a
diagnosis of Psychopathy is dependent in large
part on whether the subject of your assessment was

We, today, depend too often on the evidence of
experts, rather than the evidence of fact.

My bias stands on history, a history of near on 35
years on the front lines, cleaning up the messes
mad by experts, carrying the agenda of whomever is
paying the bill. I have reduced my assessment
skills set to three simple rules.

1.You look for what is there that should not be

2.You look for what is not there that should be

3.You leave your ego/agenda on a peg at the door*

*– That is to say, you walk into a room with a
group of people talking and everyone goes quiet –
It is not because you have walked into the room –
It is because ‘someone’ has walked into the room.

Each day, as we walk the byways and
flash along the Highway of Light:
There are those who are left
standing, as the march to flight
goes by some to stop and wonder
at the enormous potential at what
has been created:
But others who are left, standing
in a rut at the side of the road,
a rut too soon to become an abyss.

So Then? What to do?

Have said before, too many times to count,
yet shall say again, as often as need be,
until there is clarity in the air:

The time has come to put a fence at the top of the cliff
instead of a net at the bottom:
Thus giving a chance to build a bridge over the abyss.

Kendra writes:

I think that in a lot of cases people use depression to get attention. If you think about it, people who are out of the "norm", they are the ones who get the extra attention. If someone is sad and depressed all the time, people are going to be concerned about them and maybe take extra time to listen to them. I agree with the article about how people use depression as something that gets them out of being relied on. Personally I would hate to know that I was never able to be relied on, and that I hate to count on everyone else to get my work done. I beleive laziness starts relating with that and how people want to rely on everyone else. Depression is a big concern among americans today, especially in women. It needs to be takin more seriously then what it is.

Dr. T writes:

Barkley Rosser makes the absurd claim that diseases with neurobiochemical causes must have arisen through evolution. Disease is not an evoluationary adaptation, as Eric Hannekin and Matthew Cromer point out in their comments. Susceptibilities to serious diseases are maladaptive and will die out (on evolutionary time scales) unless something related to the disease confers some benefit. The best example of this is sickle cell disease. The full disease (hemoglobin SS) is fatal in childhood (without repeated transfusions). Sickle cell trait (hemoglobin AS) confers resistance to malaria. Lack of trait (hemoglobin AA) is associated with a higher rate of malarial deaths. The deaths from malaria among AA persons were balanced out by the increased survival among AS persons, which is why sickle cell disease remained in malarial areas.

The author of the paper has no reliable evidence that depression confers a survival advantage in animals or humans. Depression is a disease with disadvantages that far outweigh his speculated advantages. He omits or glosses over situations and evidence that contradict his hypotheses. He appears to have no understanding of the differences among subtypes of depression and makes claims that any physician with a halfway decent background in psychiatry would find absurd. If stating these things makes me overweeningly pompous, so be it.

jemaura writes:

There can be no doubt that depression has roots in evolutionary psychology. I tend to believe that it serves more than one master. One such master may have been revealed when I tried the atkins diet. After my initial period of zero carb intake, I immediately noticed a shift in my daily energy level. The peak highs and lows were gone and a much more even energy level developed - awake during the day and dead at night. I also noticed that my mood was much more up-beat and I was motivated to do alot of things I had been putting off. I have suffered from mild anxiety and depression on and off for a long time. After quitting the diet, I noticed my mood changing and depression began sinking in again.

My theory is this: In our ancestral environment where food would have been scarce during cold seasons, metabolism would shift to allow for increased activity needed to forrage afield hunting or whatever. During the harvest months, when trees and fields would be laden with fruits, nuts and grains, pickings were easy, no need to travel far to find food, and a depressive state - sluggishness - would allow a person to build a layer of fat to sustain him/her through the cold season. The presence of high levels of blood sugar would be the trigger.

Future study needed: Effects of low carbohydrate intake on depression - all kinds.

Matt McIntosh writes:


You picked a really bad example: the just-so stories offered up in order to shoehorn preferential homosexuality into a crude adaptive framework are uniformly theory-free, and make absolutely no sense in the context of everything we know about population genetics. In order for the "worker bee" theory of homosexuality to make sense, gays would have to spend all their efforts helping their relatives. They'd have to posess an even more compelling and obvious caretaking impulse than mothers do. Needless this is not in evidence. And there's also no reason why this should be connected to sexuality at all -- why aren't they merely sterile? Rutting with one's own sex takes away from valuable time that could be spent helping the family. It makes no sense at all, so it's a mystery to me why I see someone more intelligent than me repeating it.

hanmeng writes:

How can suicide be functional?
Well, if you "succeed", you're dead. There's a function for you.

Barkley Rosser writes:

This will be my last comment on this thread. I have already said too much, and have probably been overly harsh to Dr. T.


I did not say definitely that this was true about homosexuality. I noted that the argument had been made. It is very hard to definitively prove anything about human evolution as it happened long ago and cannot be readily observed. Our evidence is all backward speculation based on a variety of evidence of varying quality.

Dr. T.

I most definitely did not say that diseases with an n.b.c. origin must have an evolutionary foundation. They all have appeared as part of the evolutionary process, however. You yourself provide examples of ones that some have argued might have arisen out of natural selection pressures within the evolutionary process, although none of these have been proven to have done so. These are all merely not unreasonable speculations based on some evidence. This is essentially what Hagen has done, although again, I fully agree the paper would not pass muster in an medical journal because of the failure to distinguish types of depression.

Let me note a paper of a similar type, although of higher quality, that showed up on the cover of The Economist magazine in May, 2005, and is now mentioned in an early chapter of Mankiw's Principles textbook, "How trade saved humanity from biological exclusion: an economic theory of Neanderthal extinction," by R.D. Horan, E. Bulte, and J.F. Shogren, Journal of Economic Behavior and Organization, Sept. 2005, 58(1), 1-29.

Noting some anthropological evidence that humans traded with each other and Neanderthals did not (scattered and not definitive), the paper develops a model that shows how this could give humans an advantage over the Neanderthals even if the Neanderthals were individually more adapted in some ways, bigger, stronger, and more individually intelligent (they did have larger brain cases). The paper does not prove the result. It simply shows that it is reasonable.

George writes:

1) Breaking your leg is a good way to get the people around you to give you more for less, too. It's also apparently less painful, since people rarely choose death to avoid living with a broken leg. Nonetheless, it doesn't seem like something that would be selected for.

2) Odd that a libertarian economist wouldn't notice that if the people around you are giving you more for less, there's less overall as a result for "the people around you" -- on an evolutionary scale, your family -- the ones who share half your genes. This makes me think evolution selects strongly against getting depression (and getting your leg broken).

3) Narrowing focus to reactive depression, perhaps the evolutionary point is pain: it's not actually physically painful to have your offspring devoured by a lion, so little apes on the savannah might not avoid it as much as they should (from their genes' point of view). Fear of awful (but non-physical) feelings might motivate more care.

4) Switching to chronic depression, apparently sufferers make more accurate assessments of their abilities and performance than non-depressed people. Perhaps evolutionary pressure to accurately model the world goes overboard and interferes with gene propagation.

5) Scurvy tends to clear up when you increase vitamin C levels in the patient. Depression likewise tends to clear up when you increase serotonin levels in the patient. This it why people tend to view it as a disease, and not as Bryan's mother being a pain in the ass for him and his family just because she's a slacker.

Adam writes:

A few more points suggesting that the linked article is not based on a good understanding of the topics it addresses:

1) The abstract doesn't postulate any selective factor...why would people who get depressed when abused be more likely to procreate? Depression has been shown to be an unattractive quality, and predicts a lower libido.

2) Also, though the principle cause of major depression IS a significant negative life event, this "significant negative life event" which the author notes in his abstract, is defined and validated as something like marriage, divorce, birth of a child, or death of a loved one. A protracted time period during which a "powerful other is benefiting from an individual's efforts, but the individual herself is not benefiting" does not constitute a life EVENT, or parallel any of the stress/depression literature I've read.

3) The sorts of depression research which the author cites to support his theory is all chronic depression. Almost nobody talks about "acute depression;" this is just called sadness. More importantly, a protracted period of abuse would not be expected to produce anything but chronic depression. For depression to serve in the manner theorized, the "labor strike" has to end when an agreement is reached. What use is a perpetual striker?

4) It also doesn't make sense to put the "powerful other" in the position of the "family member" or "spouse." This would be necessary for the depression of the oppressed to elicit bargaining power.

5) For this connection to be an evolved quality, the conditions under which depression is helpful must be consistent over thousands of years. In this case, that would require a market with a serious dearth of labor--a market in which a depressed individual must be recruited back to working status, instead of simply fired or replaced. Has this really been the case for most of human history?

Matt McIntosh writes:


On the other hand, it's not so hard to definitively rule out certain things. Evolution is like economics in that sense: it's hard to tease out specific predictions, but you can say with a great deal of confidence that e.g. no producer is going to sell below marginal cost and stay in business for long.

There are a lot of things people try to shoehorn into adaptive stories where it simply makes no sense to do so; much more sensible would be to start looking for alternative biological explanations that *aren't* adaptations. Infection is a major suspect, since evolution is playing both sides of the conflict.

Comments for this entry have been closed
Return to top