Equality preference and inequality aversions

In a letter to Nature , a group of political scientist and anthropologist report an experiment designed to test equality preference and inequality aversions. the design was simple:

Subjects are divided into groups having four anonymous members each. Each player receives a sum of money randomly generated by a computer. Subjects are shown the payoffs of other group members for that round and are then provided an opportunity to give 'negative' or 'positive' tokens to other players. Each negative token reduces the purchaser's payoff by one monetary unit (MU) and decreases the payoff of a targeted individual by three MUs; positive tokens decrease the purchaser's payoff by one monetary unit (MU) and increase the targeted individual's payoff by three MUs. Groups are randomized after each round to prevent reputation from influencing decisions; interactions between players are strictly anonymous and subjects know this. Also, by allowing participants more than one behavioural alternative, the experiment eliminates possible experimenter demand effects—if subjects were only permitted to punish, they might engage in this behaviour because they believe it is what the experimenters want.

The results support what is often referred to as the "Robin Hood effect": richer individuals were heavily penalized, while poorer received more gift. This would support the hypothesis of Strong Reciprocity (SR), put forth by Fehr, Camerer, Gintins, and many other scholar in behavioral economics. SR implies that individuals will cooperate with cooperator (reciprocal altruism), will not cooperate with cheaters, and are even ready to punish those who cheat others (altruistic punishment):

“people tend to behave prosocially and punish antisocial behavior at cost to themselves, even when the probability of future interactions is low or zero. We call this strong reciprocity." (Gintis, H. (2000). Strong reciprocity and human sociality. Journal of Theoretical Biology, 206(2), p. 177)

And of course, SR implies that individual will be inequity-averse. SR proponent go further, and state that we an innate propensity for altruistic punishment. That’s all well and good, but why so much moral optimism? Couldn't it be that we are selfish agents and that our mechanisms of decision-making aims primarily at maximizing positive outcomes and minimizing negative ones? We feel good when we punish bad guys, we feel bad when someone make unfair offers to us in the ultimatum game (neuroeconomics studies showed that). I agree with SRers that we are not cold logical egoists, but would favor another approach I call "Hot Logic": (from an abstract of a forthcoming talk):

human agents are selfish agents adapted to trade, exchange and partner selection in biological markets (Noë et al., 2001). Cognitive mechanisms of decision-making aims primarily at maximizing positive outcomes and minimizing negative ones. This initial hedonism is gradually modulated by social norms, by which agents learn how to maximise their utility given the norms. The ‘hot logic’ approach provide a simpler explanation of cooperation and fairness: subjects make ‘fair’ offers in the ultimatum game because they know their offer would be rejected otherwise. Responders affective reaction to ‘unfair offers’ is in fact a reaction to the loss of an expected monetary gain: they anticipated that the proposer would comply with social norms. This claim is supported by other imaging studies showing that loss of money can be aversive, and that actual and counterfactual utility recruit the same neural resources (Delgado et al., 2006; Montague et al., 2006). This approach explains why subjects make lower offers in the dictator game (an ultimatum game in which the responder make an offer and the responder's role is entirely passive) than in the ultimatum, why, when using a computer displaying eyespots, almost twice as many participants transfer money in the dictator (Haley & Fessler, 2005), and why attractive people are offered more in the ultimatum (Solnick & Schweitzer, 1999). In every case, agents seek to maximize a complex hedonic utility function, where the reward and the losses can be monetary, emotional or social (reputation, acceptance, etc.). SR is thus seen as cooperative habits that are not repaid (Burnham & Johnson, 2005)