The Ultimatum Game
Posted: December 14, 2004 at 10:17 pm in cognition, people ~ Permalink ~ TrackBack

I mentioned that there would be cases when people would answer the question “Do you want $2 or $0?” with “$0″. This is what actually happens in the Ultimatum Game, described here, with references. The basic idea is that there are two players, asked to split up a pot of money, say $10. The first player gets to decide an appropriate split between the two players. The second player is given the option of accepting the split as offered, or rejecting the split, in which case neither player gets anything. In the isolated case of “Do you want $2 or $0?”, a responder would almost always take $2. But if the responder is playing the Ultimatum Game, and they know that when they’re getting $2, the first player is getting $8, half of all respondents reject the $2 and take nothing.

This is a fascinating result to me. It demonstrates the all-encompassing importance of context. In an isolated context, people answer one way. In a social context, they answer differently, where feelings of fairness are brought into play. In fact, one study used MRI scanning to demonstrate that unfair offers activated a part of the brain that is associated with negative emotions, including, one would assume, spite. The paper goes on to point out that the MRI results demonstrated a conflict between “the emotional goal of resisting unfairness and the cognitive goal of accumulating money.”

One might wonder where humans have learned what “fairness” is, and why it is built into our brain chemistry. This paper gives some insight into how such an instinct evolved. In it, they run computer simulations and demonstrate that the fairness instinct can evolve in the Ultimatum Game if participants are given a history. If it were a one-off game, the first player would always make the split uneven, and the second player would decide that something is better than nothing. However, if there are repeated iterations, the second player can spite the first player by holding out for a “fair” split, and enhance the likelihood of getting a better deal in the future. In other words, fairness only matters when you are likely to interact with the same people repeatedly – “When reputation is included in the Ultimatum Game, adaptation favors fairness over reason. In this most elementary game, information on the co-player fosters the emergence of strategies that are nonrational, but promote economic exchange.” And the MRI studies demonstrate that such strategies, such feelings of fairness, are actually built into our brain chemistry.

This leads to an important result in my mind. Because we are primates, and prisoners of the monkeymind, things like fairness and social justice only matter when we are dealing with those within our monkeyverse. If we are dealing with people within our social universe, people who we are likely to run into again, even if they are only Familiar Strangers. We don’t rip off the guy at the corner convenience store because we stop in there regularly. We pay our fair share at dinners with our friends because we know we will be going out to dinner with them again.

However, if we are dealing with strangers, with people we don’t feel are part of our world, and with whom we will not have to interact with again, then all the rules of fairness go out the window. We are returned to pure self-interest. It’s like a one-off game of the Ultimatum Game. We feel fine cheating the people we don’t know, because in an emotional sense, they aren’t people to us. They don’t evoke our rules of fairness. They are objects in the world, to be used and disposed of.

How do we expand our monkeyverses, keeping us from doing stupid things like stealing from strangers, committing hate crimes, and invading foreign countries? My answer is probably not surprising: we use stories as a way of giving us the details about other people that change them from cardboard cutouts into people. By turning them into real three-dimensional people, stories can activate our monkeybrain and all of the accompanying emotions of fairness and guilt. Such emotions leverage the way our social brains have evolved, hopefully getting us to treat each other better. It’s a theory. And one I’ll probably return to at some point.

Previous: Conservative postmodernism | Next: Context in modern physics




  1. Eric Nehrlich, Unrepentant Generalist || Transactional exchanges || November || 2006 commented on November 1st, 2006 at 1:21 am :

    [...] One other aspect of the experiential exchange is that unlike the transactional exchange, there is a continuing relationship between the buyer and the seller. There is history. There is likely to be a future. If all exchanges were between buyers and sellers meeting only once, there would be no incentive to provide anything but a transactional exchange. It’s only worth the extra effort to construct an experience for the buyer if that work will be rewarded by a continuing relationship in the future. A reputation has to be maintained (shades of the ultimatum game). This actually relates to some ideas I’ve been wanting to write up about community but I’ll save that for another night when I don’t have class. [...]

  2. Eric Nehrlich, Unrepentant Generalist || Sharing || November || 2006 commented on November 4th, 2006 at 4:21 pm :

    [...] Social capital does have an element of selfishness. Our brains have evolved to keep track of social favors (as the ultimatum game suggests), and if I do favors for others, then they owe me. I can hoard that social capital and collect favors to use as leverage when needed. [...]

  3. Eric Nehrlich, Unrepentant Generalist || Community Capital || November || 2006 commented on November 7th, 2006 at 10:54 pm :

    [...] And being able to keep score in our heads of things like fairness and our history of interactions with others is part of what has let us decouple ourselves from time and space. I can do you a favor now, and you’ll do me a favor later. We even created money, a way to keep track of this stuff outside of our tribes, so that we can interact with those outside our personal MonkeySphere. Since we have neither a history nor a future of relationships to maintain with outsiders, we use money as a generalized score keeper, as we interact with them via transactional exchanges. Within our tribe, we leverage our social capital; outside the tribe, we use money for that purpose. And it can’t work the other way; using social capital outside our tribe doesn’t work (”you want to pay me off with your sweet personality? No way!”), and using money within our tribe is considered crass. If you have to pay off your friends, it’s questionable that they’re your friends. [...]