One of the fundamental premises on which my argument in Ultrasociety is built is that structural inequality is detrimental to cooperative action. By structural inequality I mean large differentials in power. For example, between a slave and a slave-owner, or between a peasant and a lord. Other examples include cooperating with someone who is physically much stronger than you in a situation where they can beat you up to impose their will on you. Or, to take a more modern situations, cooperation with someone who is much wealthier than you–someone who can afford to hire expensive legal aid that you can’t afford.
Intuitively it seems clear that such situations are not conducive to cooperation. First, cooperation requires mutual trust and a basic degree of sympathy, and large differentials in power destroy that. Second, cooperation can yield high returns on effort expended by each cooperator, but that also means that dividing up the gains from cooperation becomes vulnerable to one cooperator taking all. A large differential in power would mean that the weaker party would have no recourse if cheated (in other words, there is no possibility of moralistic punishment).
All this makes sense, but if we want to develop the science of cooperation, we cannot go on introspection alone. We need data. It would be particularly nice to have direct experimental evidence of the effect that inequality has on cooperation.
Recently (22 December 2015) Nature published a study that provides such evidence (I only regret that the article came out after my book was published). Katherine Cronin and co-authors observed how participants played cooperative games while inequality among them was manipulated by the experimenters.
The experimental setup was a bit complex, and involved participants playing a combination of a public goods game and an ultimatum game. Here’s how it was done.
A group of ten participants was ranked on the scale of 1 to 10. In one version of the experiment, ranks were assigned randomly, while in another participants first competed for higher ranks by playing Tetris, solving arithmetic problems, etc. But whether ranks were earned or randomly assigned turned to be of no significance to the results, so the authors combined the two unequal treatments and compared them against those that were unranked—equal.
Next, pairs of experimental subjects, who knew each others rank, played a variant of the public goods game, in which they could gain large returns if they contributed enough to the common pot. The kicker came at the final stage when the gains of cooperation were distributed, using a variant of the ultimatum game. The higher ranked individual proposed a distribution (for example, she could reserve 70% for herself and offer 30% to her partner). The lower ranked subject could either accept, or reject. Unlike in the usual ultimatum game, however, rejection was not final—it did not mean that the pot was lost to both participants. Instead, a lottery was held, and the pot was given to one of the participants with the probability proportional to her rank. For example, if my rank was a lowly 1 and yours a hefty 9, then my chances of getting the pot after rejecting your offer would be only 10%.
So what would I do if I found myself to be a low-ranked participant playing against one with a high rank? Well, I would find the whole setup for dividing the gains completely unfair. Not only am I going to be in the position of low power because the other person is deciding how much to offer me. I couldn’t even get the satisfaction of rejecting an unfair offer, because then the pot would be given to the proposer with a high probability. No way would I cooperate (contribute anything) in such a case!
And that’s precisely what happened in the experiment. The greater was the difference between ranks in interacting pairs, the less cooperation–the less powerful players simply did not contribute to the common pot.
Inequality corrupts cooperation.
What are the implications of this study for the broad-scale, society-level cooperation in our modern societies? As I wrote in an earlier post, the capacity of the Americans for social cooperation declined together with growing inequality over the last three-four decades. The experiment by Cronin and co-authors suggests the following explanation of this trend.
As such social commentators as Kevin Phillips worried, increasing inequality of wealth gave rich people disproportionate power to influence politics in our democratic societies. The work by Martin Gilens has shown that public policy in the US reflects only the desires of the richest segment of the population—the top 10 percent, and most likely the top 1 percent. In other words, when it comes to dividing the pot, the 99 percent have no say anymore. Should it then surprise us that several measures of social trust and cooperation have been declining?