Cooperation and Evolution. Why punish?

Behavioral economics is so fascinating, especially when combined with thinking about evolution and the ideal society. During globalization class today, Professor Thompson noted that to achieve justice in society, coercive government is necessary in order to punish as a way of guaranteeing cooperation.

It just so happens that yesterday, while reorganizing my files and folders, I came upon an article titled "Winners Don't Punish". According to this 2008 article from Nature Magazine, costly punishment is never a good idea. It helps neither the punisher nor the group as a whole, at least as long as reciprocity is possible. In other words, this article concluded that if tit-for-tat is an option, that's the way to go.


The question comes up, even in the article, why do we have an evolutionary tendency to punish, even at a cost to ourselves? The authors suggest that costly punishment works as a strategy for the strong to overpower the weak, resulting in societal structures that tend toward dominance hierarchies.

Anyway, there's a lot to discuss about this.

After investigating my web-browser history like a detective, I found out why I have the article. David Rand, one of its authors, actually came to speak at U-M in September, I got an e-mail about it, got interested after reading the blurb, Googled him, and downloaded the article.

The topic of his talk here was "Reward, punishment and the evolution of cooperation". In that talk and in the interview above, Rand discusses reward as an alternative to punishment.