Tuesday, March 09, 2010
The Joy of Nastiness
Schadenfreude, taking pleasure in someone else's pain or misfortune, seems to appear a normal thing to most people. It occurs in literature, comedy, our every day lives. But what do Economics and Social Psychology have to say about it? We know from previous research by Simon Gaechter, Benedikt Herrmann and several others that people display 'anti-social preferences' - they punish people who contribute more than they do to a public good. Now, a recent experimental paper, 'The pleasure of being nasty' by Klaus Abbink and Abdolkarim Sadrieh, tries to take some of this work forward, they investigate a game in which players can destroy the output of others without reaping any benefit from the destruction, or without incurring any cost for doing so. What do you think people might do?
Abbink and Sadrieh nested an experiment that allowed players to destroy the output of others in an experiment in which subjects were asked to evaluate adverts for a marketing campaign. They then had the option to 'destroy' the production of another person who was doing a similar evaluation. The destruction treatment came in two forms: in the first treatment (open) subjects could destroy the produce of their partner and the information was shared perfectly, in the second treatment (hidden) the information was shared imperfectly because a random amount would be destroyed. The procedure was repeated ten times. The incidence of nastiness for the two treatments differed substantially (see Fig. 1 adjacent) and people evidently took advantage of the opportunity to destroy their partner's production more when there was less chance of being found out.
So, when there is no benefit to destroying, people destroy anyway. When there is less risk of being found out by your counterpart - even though they could retaliate - people destroy a bit. Moreover, the subjects tendency to destroy did not correlate with whether they had experienced their produce being destroyed in previous rounds. Also, the subjects did not seem to be driven by inequity aversion because they often destroyed their counterpart's produce even their counterpart had the same as or less than they did. Abbink and Sadrieh make the point that we need to reconcile these results of nastiness with the many results we have about prosociality, cooperation, inequity aversion and others. The view also needs to be reconciled with our conception of rational homo economicus which might say that no rational individual would just destroy things without a pecuniary benefit. So, because I take pleasure in the misfortunes of homo economicus, I'll end by linking to the song schadenfreude from Avenue Q - 'Cos when I see how sad you are, it sorta makes me happy!'
From the abstract: "Two players receive an endowment."
In choosing to destroy their partner's endowment or "production", are participants destroying something that has actually been laboured over by their partner, or only a kind of 'imaginary production' which is of no emotional significance to them and also doesn't really matter to their partner?
The experiment seems to have a tone similar to a game, with neat 'production' units standing in for the fruit of blood, sweat and tears in the real world. But destructive choices in the real world are usually made in emotionally complex scenarios: the destruction one wreaks is usually somehow meaningful. In this experiment, except in the retaliation scenario, it's hard to see how the production or the destruction is meaningful.
Also, it seems premature and invalid to extrapolate from these laboratory observations to human nature. Thus, a statement that seems to be about human nature, such as "when there is no benefit to destroying, people destroy anyway", is misleading.
Firstly, we don't have the evidence to make such a claim except insofar as it refers to this specific, lab-controlled instance.
Secondly, real-world instances bearing out the statement "when there is no benefit to destroying, people don't destroy" are too numerous to count.
These dark pronouncements on human nature are rather less serious than we might fear.
They laboured at a task for which they were paid - hence 'production'.
The point of the research was to show that people aren't always generous, prosocial and inequity averse. Science needs to be able to explain and experiment with the 'darker' sides of human behaviour. So it's not a 'dark pronouncement' but rather a comment on attempts to have a fuller, more complete science of human behavior.
This experiment seems extremely artificial. So they were *told* they could destroy other people's production? Doesn't that prime them rather seriously? Were I a participant, I'd think 'fuck it, let's destroy some and see what happens'. That doesn't mean I go around keying BMWs, though.
You have the option to do something or not to do it, to be found out or not. Read the paper if it fascinates you remotely and then judge. 'Giving the option to destroy' is at least as artificial in my mind as 'giving the option to give' as in other experiments. The thing is to see whether people do it or not and to assess the conditions under which they do it more or less, as with bargaining experiments.
Well, sure. But that just shows all these experiments have artificiality issues.
My main concern is the priming, though. I don't have time to read the paper, but I have serious doubts.