Monday, July 02, 2007

By Christopher T. Dawes, James H. Fowler, Richard McElraeth and Oleg Smirnov. Nature: 446, April 12,2007, Issue 7137, pp 794-796.
The authors of this paper begin by introducing the question they wish to answer ie,
"Participants in laboratory games are often willing to alter others incomes at a cost to themselves, and this behavior has the effect of promoting cooperation. What motivates this action is unclear; punishment and reward aimed at promoting cooperation cannot be seperated from attempts to produce equality. To understand costly taking and costly giving, we created an experimental game that isolates egalitarian motives. The results show that subjects reduce and augment others' incomes at a personal cost, even when there is no cooperative behavior to be reinforced. Furthermore, the size and frequency of income alterations are strong;y influenced by inequality. Emotions towards top earners become increasingly negative as inequality increases, and those who express these emotions spend more to reduce above-average earners' incomes and to increase below-average earners' incomes. The results suggest that egalitarian motives affect income-altering behaviors, and may therefore be an important factor underlying the evolution of strong reciprocity, and , hence, cooperation in humans."
Strong reciprocity is the tendency of social organisms to not just cooperate between kin, where it can be explained by relatively simple calculations of genetic similarity or beween unrelated individuals where there is expectation of reciprocity- mutual altruismirect reciprocity - the effect of "reputation" amongst the group. It is the tendency to cooperate/reward and to punish non-cooperators even when there is no expectation of reciprocity and when both the reward and the punishment are costly. Groups that evolve a tendency towards strong reciprocity enhance cooperation within the group by this mechanism. Not only is cooperation rewarded, but defection, the "free riders" are punished. Both mechanisms are prior to any laws or even customs amongst those animals which have culture, but they influence the direction that customs evolve towards.
The authors of this paper criticize other experiments in the "public goods" set of experimental economics "games" as being unable to differentiate " cooperative norm enforcement" from the "egalitarian motive". The experimental design that they have set up supposedly eliminates any "cooperative norm". The subjects were randomly divided into experimental groups of 4 people and just as randomly assigned a computer generated "income". They were allowed to see the results for the three other anonymous people in their group and were given the opportinity to give either "positve" tokens increasing another person's income or "negative" tokens decreasing them. Giving positive tokens increased the recipient's income by three tokens and decreased the giver's by one. Similarily, giving a negative token decreased the punished person's income by three tokens but decreased that of the punisher by only one. After each round, for a total of five rounds per session,participants were randomized into new groups of four in each round of a session None would be in the same group again and this was made plain to the participants. Evolution of cooperation was impossible.
Yet, income alteration was common in the experiment even though self interest would say that the best strategy is to never do it. 68% of participants reduced another player's income at least once, 28% did it at least 5 times and 6% did it ten times or more. Similarily, 74% of subjects increased other players' incomes at least once, 33% did it five times or more and 10% did it 10 times or more. The results of the game clustered around a tendency of subjects to attempt to bring the random incomes towards an arithmetic mean. The tendency to engage in this behavior did not decline over time as subjects would supposedly learn that it decreased their individual income- something that they were informed about in the beginning but might take some time to "sink in". Neither was the tendency to give negative or positive tokens influenced by the actions of other players towards a subject in previous rounds of the game.
Further details of the experiment can be seen in the apper cited above. The authors conclude that they have demonstrated an "egalitarian tendency" seperate from other mechanisms of reciprocity that "cause individuals to engage in costly acts that promote equitable resource distribution", the Kropkonian idea of "Justice" (see Ethics by Peter Kropotkin). They admit that "concerns for equality are clearly not the only motivation for human behavior" in the context of "public goods games"that attempt to model human social behavior.
Now, Molly would not be displeased at all if the authors' contention were true, but she is of the opinion that they have failed to prove their point. The first thing is that such an experiment should be repeated across various cultures as other experiments concerning cooperation in public goods gams have been. Those experiments have shown a wide variation between cultures in terms of the two "arms" of strong reciprocity, that of reward and that of punishment. The differences do not correlate to any "mode of production" and the two arms may vary randomly between cultures. Strong rewarders may be either strong or weak punishers and vive versa. Like any other aspect of human sociobiology the tendency to engage in what some have called "the evolution of spite" is "permissive" rather than "proscriptive". The "phenotype" will exhibit varying degrees of "penetrance" of the "genotype". Does the frequency of "spiteful" and "emphatic" acts vary significantly across different cultures ? Still an open question.
Second, and connected with the first caveat, the students that the experimenters recruited as their "experimental animals" came from a rather homogenous social group- undergraduate students at the University of Davis, California. They came to the experiment with a rather narrow spread of social attitudes towards equality, empathy and spite already fully packed into their mental baggage compartment. Even though the experimenters think that they have isolated a "pure" egalitarian motive the reader is invited to imagine how much different the results would have been if these students came from a culture that strongly disapproved of spite or one that considered the "gifts" to be degrading to the recipent.
So, in Molly's mind the results are suggestive but not yet definitive. It would be very pleasing to think that there is an internal "hard-wired" propensity for people to automatically disaprove of and try to reduce inequality. Certainly experiments in Communist dictatorship where inequality was in fact retained while being denied by a massive effort of propaganda show that people can pick up on the obvious fact despite being lied to by the best liars human history has ever seen. It also shows that this inequality was generally resented,particularily as it hardened into caste differences that reduced social mobility. This is also evidence for the sort of thing that the authors suggest, but ,like their work, it is not yet proof. Molly looks forward to the replication of this experiment (the authors' not the Communists') under different conditions.

No comments: