adolescents would be rated higher on a measure of expendability
than would be a psychiatric control group. The expendability mea-
sure specifically included a sense of being a burden on one’s family.
Results conformed to predictions: Suicidal youth scored higher on
the expendability measure than did a psychiatric comparison group.
Studies of youth like this one suggest that the concept of burden-
someness may affect a broad range of ages. Related to feeling a bur-
den on one’s family, suicide attempts among children have been
The Desire for Death ● 111
linked to perceived inability to meet parental demands.42 Again, the
emphasis on the term perceived bears repeating; the facile explanation that parents are responsible for their children’s death by suicide
because of high demands is hardly worth considering. However, the
explanation that people who perceive themselves as not measuring
up and as being a burden are prone to suicidal behavior is more seri-
ous and is supported by numerous research studies.
The potential importance of perceived burdensomeness emerged
in a study of low-income, abused African-American women. This
study identified protective and risk factors that differentiated suicide
attempters from those who had never made an attempt. Protective
factors associated with nonattempter status included self-efficacy as
well as effectiveness in obtaining material resources.43 In the current
framework, these results could be viewed as suggesting that those
who feel effective in general, as well as effective in providing material
resources in particular, are buffered from feeling a burden and thus
at relatively low risk for suicidality.44
Data on suicide in the Netherlands in the early twentieth century
are in accord with a role for perceived burdensomeness as well. A
higher suicide rate in rural as compared to urban areas was noted
and was attributed to “the peculiar conditions of the Dutch farming
system under which the aged find themselves a burden,”45 an experi-
ence not shared by older Dutch people in the cities.
One implication of the burdensomeness view is that suicides may
be more common during difficult economic times; when resources
are strained, the consequences of perceived burdensomeness may
be more acute (for example, the socially sanctioned suicide among
the Yuit Eskimos of St. Lawrence Island outlined earlier in this chap-
ter). Other data support this view. A significant association between
deprivation and the suicide rate in thirty-two London boroughs
has been documented. This study is unique in contemporaneously
studying closely contiguous and roughly culturally homogenous ar-
112 ● WHY PEOPLE DIE BY SUICIDE
eas that nevertheless vary with regard to socioeconomic variables
and suicide rate.46 Several others have reached the conclusion that
economic prosperity has a beneficial impact and economic down-
turns a detrimental effect on suicide rates.47 In a study of suicide in
African-American men, the risk of suicide was higher in areas where
occupational and income inequalities between African-Americans
and Caucasians were greater.48 At the individual level, too, several
studies have found that financial hardship constitutes a risk factor
for suicide.49
Suicidal Behavior in an Evolutionary Context
Some believe that suicidal behavior must have conferred some bene-
fit in the course of evolution, in part because it is common across
cultures today. But how would a behavior that leads to pain, impair-
ment, and death produce any sort of increased ability to pass on
one’s genes?
According to DeCatanzaro’s50 model of self-preservation and self-
destruction, there are conditions under which death may produce an
adaptive advantage. Specifically, death may produce an evolutionary
edge for an individual who has few prospects for reproduction and
who poses such a burden to close kin—who carry his or her genes—
that it reduces their prospects for reproduction.
For any case for adaptive benefit to be made, a lot of compelling
evidence would need to be marshalled (and even then, as I will point
out, questions will remain). For example, if suicidal behavior is an
exclusively human phenomenon not occurring in other primates and
animals, then the case for evolved behavior becomes more difficult.
If, by contrast, suicidal behavior is documented in primates and
other animals, the case is not made, but it may become a little easier
to imagine.
Is suicidal behavior an exclusively human phenomenon? Many
presume it to be so. For example, the promotional material for
The Desire for Death ● 113
Shneidman’s The Suicidal Mind 51 makes the claim that “suicide is an exclusively human response to extreme psychological pain.” In Conceptions of Modern Psychiatry, the American psychiatrist Harry Stack Sullivan stated, “So far as we know, there is nothing remotely approaching [suicide] in the infrahuman primates or any of the lower
animals. It is a distinctly human performance.”52
The most famous possible case—lemmings dying by mass suicide
in a kind of intentional population control effort—is not really a
case at all. Lemmings do die en masse, but this is because they mi-
grate en masse after they exhaust their food source—a type of slow-
growing moss—and many die in the hardship and chaos of the mi-
gration.
Still, Shneidman and Sullivan may have been wrong in thinking
that suicide is an exclusively human phenomenon. Perhaps the clear-
est examples of animal suicidal behavior involve a phenomenon
dubbed “adaptive suicide” in insects and possibly in birds. For in-
stance, researchers have studied pea aphids, which are parasitized by
a specific species of wasp.53 The wasp injects an egg into the host
aphid; the young wasp matures inside the aphid, feeding upon its or-
gans. When the wasp is ready to emerge as an adult, it chews a hole
out the back of the aphid’s body. Aphid populations can be devas-
tated by parasitic wasps.
Recall that DeCatanzaro theorized that self-sacrifice carries sur-
vival value for an individual who has few prospects for reproduction
and who poses such a burden to close kin—who carry his or her
genes—that it reduces their prospects for reproduction. The parasitized pea aphid could be viewed as just such an individual—prospects
for reproduction are few, because death is imminent; a potential bur-
den is posed to close kin, because the parasite wasp will live on to in-
fect other pea aphids. Aphids parasitized early in development, and
who thus would not produce any offspring, frequently engage in a
kind of “aphid suicide”—specifically, they drop from their host plant
114 ● WHY PEOPLE DIE BY SUICIDE
to the ground, where they are frequently preyed on by ladybugs and
other natural predators.
Worker bumblebees parasitized by a specific species of fly will of-
ten abandon their nest altogether, cutting themselves off from the
hive and ensuring their death. The bees’ behavior can be see
n as an
instance of adaptive suicide because the parasitized bee’s premature
death kills the parasite and avoids its spread in the bee colony.54
A similar phenomenon may affect birds. Under famine conditions,
evolutionary pressures may select for behaviors like fratricide (one
member of a brood killing its sibling), infanticide (a parent killing
the offspring), and suicide (by the nestling with the shortest life ex-
pectancy, which abandons the nest prematurely).55 This kind of self-
sacrifice is sort of what Durkheim had in mind (though from an en-
tirely different perspective) when he wrote about altruistic suicide,
which he defined as self-sacrifice for the good of the group. A key
difference, however, is that adaptive suicide takes place not for the
good of the group, but for the good of one’s own genes.
Perhaps an even clearer example of self-sacrifice for the good of
one’s genes occurs in the Australian redback spider. Male redback
spiders submit to being cannibalized by females after sex. They do
so apparently because they thereby gain advantages in the competi-
tion for females: Cannibalized males copulate longer and fertilize
more eggs as compared to males that survive copulation. Also, female
redbacks are more likely to reject subsequent males after consuming
their first mate.56 Self-sacrifice has been selected for in the course of
male redback spider evolution.
Self-injurious behavior has been documented in nonhuman pri-
mates; most studies are on rhesus monkeys. The form of their self-
injury is usually self-biting—at times of distress, some monkeys will
bite their arms or legs, sometimes causing injury. This body of re-
search suggests that self-injury in nonhuman primates is a way to
self-regulate at times of stress (e.g., accelerated heart rate decreases
when the behavior is enacted). Animals with early stress experiences
The Desire for Death ● 115
are vulnerable to the behavior, suggesting that these early experiences
disrupted the development of more normal stress-regulation abili-
ties. Self-injury as a means to regulate emotions has been docu-
mented in humans too, of course, particularly among those with
borderline personality disorder.
Apart from adaptive suicide in insects, birds, and spiders, there
is very little evidence of death by suicide in animals. There is at least
one claim that dogs sometimes die by suicide—by drowning or by
food refusal in response to being cast out of a household or re-
morse57—but this assertion would need more systematic scientific
support before it is given much credence.
I said earlier that I believe suicidal people are mistaken when they
view themselves as burdens. An adaptive suicide viewpoint would
suggest otherwise—specifically, that suicidal behavior evolved in hu-
mans to remove actual (not just perceived) burdens from kin and
thus facilitate kin’s survival. In fact, when researchers interviewed the
significant others of eighty-one people who had recently attempted
suicide, a majority of significant others reported that their support of
the patient represented a burden to them.58
Nevertheless, I do not much like this adaptive suicide view; my
own dad died by suicide and the idea that he was an actual burden
is offensive. My view is that self-sacrifice is adaptive in some ani-
mal species. It may have been adaptive under certain conditions in
the course of human evolution, but we will never really know. Most
important, it does not really matter now. What matters now is that
perceived burdensomeness—and, to the extent that it exists, actual
burdensomeness—are remediable through perception- and skill-based
psychotherapies. Death is no longer adaptive, if it ever was.
A Note on Life Insurance
Perceived burdensomeness implies a kind of calculation along the
lines of, “my death will be more valuable to others than my life.” It is
interesting to consider the issue of life insurance in this regard. The
116 ● WHY PEOPLE DIE BY SUICIDE
usual life insurance policy will pay for death by suicide provided that
the death occurs two years or more after the initiation of the policy.
If the death occurs before two years have passed, a usual practice is
that premiums plus interest are returned, but the policy is not paid.
If people are considering suicide and they imagine that their death
will confer benefits to loved ones, then life insurance could enter into
suicidal people’s thinking. A higher benefit could conceivably en-
courage suicide. It would be interesting to know the percentage of
those who die by suicide with life insurance versus the percentage of
those who die by other means with life insurance. Any comparison
like this should adjust for demographics, which are very different be-
tween those who die by suicide versus those who die by other means.
One way to do this would be to limit comparisons to one ethnicity
and age range, say, white men over fifty. If, after such adjustments,
more suicides are insured, it would suggest that life insurance possi-
bly figured into the calculations made by suicidal people. It would
only be suggestive, though, because other factors could have influ-
enced the data. Perhaps a personality variable like lack of optimism
would explain both the tendency to take out life insurance policies
and a tendency toward suicidal behavior.
Perhaps not surprisingly, data on this issue are hard to come by. In
the 1920s and 1930s, Metropolitan Life paid out more life insurance
money per claim on suicide versus other deaths, though it was not
clear that this was corrected for gender and ethnicity. The figures in
1925 were $2,283 (suicide) versus $1,867 (nonsuicide). In 1931, the
respective figures were $3,580 versus $2,216.59 These numbers are ob-
viously quite dated, but suicide often remains the most expensive
category of death for life insurance companies today.
As with the findings related to the acquired ability to enact lethal
self-injury, alternative explanations exist for some of the results link-
ing perceived burdensomeness to suicidality (e.g., the studies on in-
creased suicide in difficult economic times are amenable to many in-
The Desire for Death ● 117
terpretations). Relatively few studies have explicitly tested the specific
role of perceived burdensomeness in suicidality. On the positive side,
every study specifically examining burdensomeness and suicidality
had stringent features, and each produced results supportive of the
role of burdensomeness in serious suicidality. As with the findings on
acquired ability to enact lethal self-injury, the results reviewed on
burdensomeness were from a wide array of topic areas, samples, and
methodologies. That such diverse data are at least consistent with the
parsimonious construct of perceived burdensomeness inspires some
confidence in its role in suicidality.
The current model asserts that to perceive oneself as so ineffective
that loved ones are threatened and burdened is a so
urce for the desire
for suicide. Perceived burdensomeness, not actual burdensomeness, is the key variable in this account. Empirical data to date, including
studies explicitly testing this assertion, are supportive. Assuming the
capability for suicide, perceived burdensomeness removes one of the
two key barriers to suicide. Even for a person who has acquired the
capability for suicide and perceives him- or herself to be a burden,
there remains one “saving grace”—belongingness. In my view, if the
need to belong is satisfied, the will to live remains intact.
Thwarted Connectedness: The Sense that One Does Not Belong
William James, in The Principles of Psychology (1890), wrote, “No
more fiendish punishment could be devised, were such a thing physi-
cally possible, than that one should be turned loose in society and
remain absolutely unnoticed by all the members thereof. If no one
turned around when we entered, answered when we spoke, or
minded what we did, but if every person we met ‘cut us dead,’ and
acted as if we were non-existent things, a kind of rage and impotent
despair would before long well up in us, from which the cruelest
bodily torture would be a relief.” I think James’s use of the phrase
118 ● WHY PEOPLE DIE BY SUICIDE
“cut us dead” is telling, as is the insight that bodily torture can, under
some circumstances, be a relief.
The need to belong is a fundamental human motive. When this
need is thwarted, numerous negative effects on health, adjustment,
and well-being have been documented. It is interesting to note that
the pain of thwarted belongingness may activate similar brain areas
as physical pain. It has long been known that a brain center called
the anterior cingulate cortex is important for the processing of physi-
cal pain signals. Eisenberger and colleagues60 obtained brain scans
of undergraduates as they played “cyberball.” “Cyberball” is a com-
puterized ball-tossing game. Participants believed they were playing
with two other players, and during the game, the two players ex-
cluded the participant. In reality, there were no other players; partici-
pants were playing with a preset computer program.
If someone is in a brain scan machine and receives a pain stimu-
lus, the scan will show activity in the anterior cingulate cortex. Simi-
Why People Die By Suicide Page 15