Jen pointed me to Level-Headed: Economics Experiment Finds Taste for Equality. In brief, people are willing to pay their own money to take from the rich and give it to the poor. The underlying Nature article mentions that:
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 behaviours, and may therefore be an important factor underlying the evolution of strong reciprocity and, hence, cooperation in humans.
However, I can see other explanations that don't require the explanation of altruism:
- Utility arbitrage: Utility is nonlinear: taking $1 when you have $10 of daily income is worse than taking $10 when you have $100 of daily income. This is used as an argument for progressive taxation, which might be nonlinear in money, but could be linear in utility (taxes giving everyone the same amount of pain). Those who take from the rich and give from the poor might effectively be doing arbitrage: the amount of gratitude from the poor minus the anger from the rich minus the cost amounts to a positive profit for Robin Hood.
- Insurance against slavery: There is an incentive for a commoner to prevent a powerful figure from gathering excessive power because letting this go on could lock the commoners into an under-caste.
- Power asymmetry: The rich can become richer only by increasing the imbalance in the income distribution. But as they become richer, they actually become fewer. At some point, increasing their riches actually reduces their power, and they get "taken under" (in a revolution or a revolt). Since revolutions are costly, it's adaptive to "equalize" without breaking things up.
As an aside, it's interesting to notice James H. Fowler among the authors: he's behind a chain of very interesting papers over the past few years.
I do think that "fairness" is a strong motivation which, actually, can work in different directions politically. On one hand, redistribution makes the income distribution more "fair" (in the sense of being more even). On the other hand, a non-redistributional flat tax seems "fair" (in the sense of treating all people equally). Depending on how you ask the question, you can find that people support redistribution and that they support equal tax rates.
P.S. More comments on Fowler's papers here.
Fairness is indeed a strong intrinsic motivator: people like fairness. But the same behavior would be generated by selfish extrinsically motivated behavior: societies that promote fairness beat societies that don't promote fairness in wars. And because the winner writes the history textbooks, the fit fairness meme lives on.
Aleks, all of the alternatives you mention might explain ostensibly altruistic behavior in other games, but they don't seem to explain it in this one. The key is repeated interaction: all of your alternative explanations require it, but it isn't a feature of the game in the Nature article. (Subjects did play the game multiple times, but no one played with the same person more than once.) Am I missing something?
John, the subjects expressed a preference for equality. In the sense of selfish behavior, this was suboptimal behavior in the context of this game.
But people have patterns of behavior (emotion of anger at inequality) that are adaptive in general life (sharing a cake among family members, for example), while perhaps not adaptive in the context of the game.
So, adaptive emotions evolve through repeated interaction (sharing of cakes), but they are not necessarily adaptive in an artificial setting that resembles the sharing of a cake.
adaptive emotions evolve through repeated interaction [...] but they are not necessarily adaptive in an artificial setting
Sure. And this explains why all three of your alternatives might account for the behavior observed in the Nature game. But it doesn't convince me that those alternatives are plausible, especially when we pit them against an explanation rooted in players' simple desire for equal outcomes. For example, I grant that the desire for “insurance against slavery” might cause the punishing behavior that we observe in the game, but the game itself provides no incentive to insure against slavery: slavery is impossible in the game, and the players know it. The same is true of revolt, which is essential to the “power asymmetry” explanation, and receiving gratitude or anger from other players, which is essential to the “utility arbitrage” explanation. On the other hand, if players just desire equality for its own sake, it's quite possible to satisfy that desire within the game. This makes egalitarian motives a much more likely explanation for the behavior that the authors observe—to me, at least. You seem to disagree. So how would you change the experiment to test whether pure egalitarianism is the explanation?
John, the chain of causation would be as follows:
insurance -> desire for equal outcomes -> observations in the experiment
Just as liking fatty and salty foods isn't adaptive in the present artificial circumstances, it is a preference that has evolved because it was beneficial in the past and now exists by itself.
The underlying dilemma, which isn't really discussed in the paper, is the where does desire for equal outcomes come from. The paper suggests that the desire for equal outcomes underlies strong reciprocity.
Now, strong reciprocity is usually defined as something that cannot be explained by Neo-Darwinism, and would require some sort of gene-culture coevolution. While there is a lot of evidence that gene-culture coevolution is present, it is not necessary. Selfishness with foresight can explain many of the phenomena.