This is a little weird

Ben Goldacre\’s noted an interesting little paper.

They took 48 students and got them all to play a game for money: players could earn cash if they co-operated, but a player who did not co-operate could make more money, at the other players\’ expense, as a \”free rider\”, by appearing to co-operate, and then reneging on a deal at the last minute.

The game, of course, was staged. The other players were computer algorithms and they were programmed to be nasty: on several trials, one of the players would warmly encourage the others to co-operate, coming over all friendly, but then, at the last minute, turn on their fellow players, making more money for themselves, and robbing the students of both reward and peace of mind.

Some of the students were then offered the opportunity to punish the person who had ripped them off, at cost to themselves: for every 5c they spent, 15c would be confiscated from the free rider. Asked how they would feel about punishing their adversaries, students said they thought it would make them feel better. They were wrong. The participants who were offered ? and took ? the opportunity to exact revenge actually felt worse afterwards than the ones who had no such opportunity.

The results are actually a little weird…..for there\’s been quite a lot of this sort of experimentation within economics and these guys are psychologists….meaning that there\’s no reason why they should be aware of the other research, divided as academia is (although of course they may be aware, I\’ve no idea).

Some of the stuff that comes out from behavioural economics shows that, in these sorts of games (involving the division of money.) people will punish the free riders at cost to themselves. This is especially so when multiple iterations of the game are played with the same contestants. From this we can build models of fairness and trust….and we do very much look at these experiments to build such. Similarly, if we play the Prisoners\’ Dilemma through multiple iterations the winning strategy is to do whatever the opponent did last time. If they screwed you, screw them, if they co-operated, co-operate…..it pretty quickly gets to co-operation.

This can then be used to explain all sorts of things about our own society, like how notions of fairness are socially imposed and so on.

However, what this paper seems to be indicating is that we feel bad about such imposing: what this paper calls revenge would, in the economics papers, be called something more like punishing those who breach notions of fairness.

All of which is a little weird. For we know that we do indeed impose such punishments, it\’s part of the grease that makes society rub along. So why do we feel bad about doing so?

3 thoughts on “This is a little weird”

  1. Perhaps it was the loss associated with inflicting the “punishment”. We like to think that “justice” and “revenge” bring closure and satisfaction – but it may be that they just prolong psychological involvement with the trauma. If one can’t do anything about it, one just “moves on”.

    I think of this in context of the Palestinian Arabs, who have spent the last 60 years sacrificing resources in futile pursuit of “revenge”. They refuse to move out of squalid camps into new housing because they might lose their “refugee” status which theoretically entitles them to “return” to pre-1948 homes (and throw those accursed Jews out).

    Meanwhile, Vietnamese “boat people” who lost everything and wound up in Louisiana have built up thriving shrimp-boat operations.

  2. It’s the repetition (iteration) that is the key. This is well-known. Reaching with one step and plucking my copy of The Selfish Gene (2nd Ed, published 1989, 1st Ed 1976) from my bookcase serves to confirm this. Goldacre’s normally sound. So why is he highlighting something that has been commonplace before either I or the author of this blog could sing baritone?

  3. Perhaps it is simply a matter of opportunity cost?

    The act of vengeance demanded the expenditure of resources that may have been better employed in winning the game or in their next move in the game. So perhaps keeping that 5c gave them greater utility than spending it. Notice that the article only says “worse” and not “bad”.

    As a general rule of thumb, I’ve heard that people hate losing something twice as much as they love winning the same thing. (Thaler and Sunstein for example, and it crops up in various other places). So perhaps the 5c they spent/lost simply had a greater negative emotional impact than if they’d not spent it, unless the 15c deducted from the free-rider were to be considered a gain?

    Alternatively, perhaps the 5c expenditure would be seen as paying more towards the free-rider problem, on top of what they’d already been cheated out of. They said they’d feel better about the free-rider being punished, but it doesn’t appear to suggest that they were asked how they’d feel if the punishment would incur a further cost on the punisher.

    Another possibility could be the immediacy of the feedback. It could be more about feeling worse after paying 5c in what is essentially a tax to pay for justice, and seeing what that tax is spent on. By associating cost with justice, it perhaps made it less desirable: whereas usually we wouldn’t necessarily consider the cost of taxes when being avenged.

Leave a Reply

Your email address will not be published. Required fields are marked *