Explore Harvard's Nieman network Nieman Fellowships Nieman Lab Nieman Reports Nieman Storyboard

Torture's attraction is not information -- it's retribution

COMMENTARY | November 06, 2008

It's not a good way to get information, so why did Bush administration officials endorse torture? A social psychologist's experiment offers evidence that those who endorse torture do so out of a desire for retribution, rather than because they think it will be useful.

By Kevin M. Carlsmith

How did the United States go from a champion of human rights to a state that condones and practices torture on detainees?  The present administration’s first line of defense is one of semantics: The United States has a policy against torture, ergo, actions taken in its name cannot be “torture.”  Its second line of defense invokes the utilitarian argument of expediency: It was necessary to obtain mission-critical information from combatants who would only divulge secrets under extreme duress.  In lay terms this is known as the “Jack Bauer” defense.  Sadly, reports indicate that episodes of the television show “24” were, in fact, used in the formulation of interrogation policy.

And yet, interrogation experts make clear that torture is a terrible way to obtain information.  Not merely from a moral perspective, but from a utilitarian perspective as well. Torture victims will always talk, regardless of whether they actually have any information.  That is, the obtained information is generally useless, and when it does have value, it is mixed in with so much false information that there is no reliable way to separate the true from the false.  We are left, then, with a puzzle:  Given the near unanimity that torture is immoral, and the expert agreement that it serves no intelligence function, how do we explain the broad support for enhanced interrogation techniques within the administration and within large segments of the population?

To understand this puzzle, one must first understand something about the psychology of punishment.  There are many justifications for punishment. One can punish a perpetrator to administer his “just deserts,” or punish to deter him and others from behaving similarly in the future.  Punishment can rehabilitate a person, coerce him into providing information, or simply change the cost-benefit analysis so that he will be inhibited from certain behaviors.  People are generally aware of these different justifications, and they like them all.  Indeed, numerous surveys reveal that people endorse all of the reasons, and if forced to choose, they generally split evenly between punishing someone because they “deserve” it and because it will serve some utilitarian purpose.

Psychologists have shown that there is a sharp divide between the reasons people express for punishment and the reasons that actually determine punishment.  That is, there is a discrepancy between what people say and what people do. It turns out that people are highly attuned to the factors that determine whether a person is deserving of punishment.  These are things like having the intent to harm, knowing right from wrong, and the severity of the harm.  At the same time, people largely ignore factors that would affect the utility of the punishment.  So people don’t increase recommended prison sentences when they learn that the perpetrator is likely to commit future crimes, or that the sentence is likely to deter other potential perpetrators.  When you examine behaviorally the reasons that people punish, it is all about trying to give someone what they deserve.

Torture operates along a very similar set of principles.  People treat the decision to torture in the same way that they treat the decision to punish.  They approve increasingly harsh techniques as their perception of the target’s moral culpability increases.  Put simply, if they perceive the target to be a “bad guy” who deserves to be punished, then they will approve of torture.  But if they perceive him to be a good guy (or, at least, innocent of wrong doing) then they generally won’t approve of torture.  It is only at the margins that people pay attention to the potential utility of the torture.  That is, the likelihood that a given target will divulge useful information under torture has far less impact on the decision to torture than does the perception of whether or not that target “deserves” to be tortured.

How do we know this?  The basics are laid out in any introductory text on social psychology.  The specifics, however, come from a series of experiments I conducted with my colleague Avani Sood (a fellow social psychologist who is also an attorney), some of which will be published shortly in the Journal of Experimental Social Psychology.  We surveyed about a 1000 people who were broadly representative of the United States.  We described the fictional case of Ahmad Farid, a 26-year-old Afghani who had been detained by U.S. and Coalition forces in 2007 on suspicion of terrorist activities.  Participants learned that at the time of capture, Farid was making his living by tending a small herd of goats. 

Some of the respondents learned that Farid had always been a simple goatherd, while others learned that years ago he had been a member of the insurgency and had killed members of the coalition force.  Thus in some cases he was innocent, and in some cases he was deserving of punishment.  Respondents also learned the likelihood that Farid possessed useful, actionable information.   The probability of his having information ranged from 0% to 95%.  We then asked respondents whether they thought Farid ought to be punished, and whether he ought to be interrogated.  We also asked how harshly he ought to be interrogated, with options ranging from mild questioning, to physical deprivations, to waterboarding.

What we learned was that people did not distinguish sharply between torture and punishment.  Indeed, the response to the punishment question and torture question correlated at .65, indicating that the two were largely interchangeable.  Moreover, the decision to torture was closely related to Farid’s prior bad acts and largely independent of the likelihood that he possessed any useful information.

So, what does all of this mean? First, it reveals that people use the same psychological process to form judgments about torture as they do for punishment.  Second, those processes revolve largely around retribution – the desire to give someone what they deserve – rather than the potential utility of the action.  Third, people are not aware of this process and have quite limited insight into the principles that guide their decision making.  Hence, they will often claim to be operating on the basis of utility, when utility in fact has little to do with it.

For those who are opposed to torture and to the practices the U.S. engages in, it is common to wonder how the practitioners and supporters can sleep at night.  After all, torture of another person is an ugly business, both practically and morally.  This analysis, though, can help us to understand why such practices seemingly yield so little dissonance.  For those who endorse torture, they are doing so because they believe the recipient is morally culpable and thus deserving of the mistreatment.  Their ostensible justification is utility, and this goal is in no way threatened by deontological concerns because only the deserving get tortured.

The greatest perversion of all comes from the “end-game” of this process.  Psychologists have identified a powerful and pervasive defense mechanism called the “just world” phenomenon.  People tend to believe that people get what they deserve;  good people get good outcomes, bad people get bad outcomes.  This belief serves a protective function against the reality that bad things happen randomly and can be completely out of our control.  One corollary to this belief is that a bad event is taken as evidence that the person deserved the outcome.  Hence, there is a human proclivity to “blame the victim.”  If something bad happened to this person, then surely they did something to deserve it.

Now apply this concept to torture.  When we hear of a person being tortured, it is common (if wrong) to assume that the person has probably done something to merit the torture.  The alternative, that the target was truly innocent in all ways, is too upsetting to contemplate.  The belief that the person is deserving solves the tension, and thus is much more readily accepted. 

In summary then, we find that people support torture on the basis that it gives bad people what they deserve.  When confronted with information that seemingly defies that belief – such as torturing the wrong person who merely shared a common name with an actual bad guy – they are unconsciously motivated to believe that the innocent is in fact not-so-innocent, and thus able to maintain their erroneous belief that only the guilty are tortured.

The NiemanWatchdog.org website is no longer being updated. Watchdog stories have a new home in Nieman Reports.