|
|
Economics, Altruism and Manners
Economics, Altruism and Manners
by A. Allan Schmid and Lindon Robison
Experiments with ultimatum and dictatorship games have provided economists with a rich set of empirical observations which shed light on economic behavior. Camerer and Thaler (1995) have provided a meta-analysis of this literature. The ultimatum game literature provides overwhelming evidence that people offer positive sums to split a pot of money with a partner who may reject the offer in which case neither person receives anything. It is indisputable that people are doing something more than maximizing their own incomes. What is disputed is whether the behavior can best be modeled as altruistic (interdependent utility function) or as an expression of manners and etiquette. Camerer and Thaler prefer the latter explanation.
The authors argue that the altruistic utility function explanation is not satisfactory because \"it is not possible to say whether the average participant puts a positive or negative value on the other subject's payoffs.\" In dictator games where the proposer may keep all of a windfall regardless of the acceptance of the recipient, the offer of money to the other party appears to be consistent with a positive relationship to the other party's payoffs. In contrast, in ultimatum games, respondents who reject the proposed split which denies any income to the proposer may be interpreted as receiving no utility from the proposers receiving some income. If I cared about someone, I would both propose (offer) them a large share of the pot and accept any proposed split from them. The dollar amount of the gift could be the same in both cases. On this reasoning, the authors conclude \"that the outcomes of ultimatum, dictatorship and many other bargaining games have more to do with manners than altruism.\"
However, caring need not be invariant of other's actions. If I am the proposer and care about the other party, I offer a fair split. My prior feeling about my partner is not affected by the transaction itself. But if someone I initially cared for (and would have offered a fair split) sends a signal that they do not care for me by offering a poor split, I may change my attitude toward them and will not be willing to use the proposed split as an opportunity to make a gift. Large differences in the affinity each feels toward the other are unstable. So contrary to Camerer and Thaler, the fact that I may have a positive interdependence with other's utility as a proposer and a negative interdependence as a responder does not call the altruism concept into question. Altruism is predictably different depending on the occasion and opportunity of others to signal their care for the subject. The evidence is consistent with a model that says the average player puts a positive value on other's payoffs if that person has not signaled negative feelings toward the player and puts a negative value on other's payoffs when the other person signals lack of caring. The average player does not feel close (altruistic) toward an unfair proposer.
There is other evidence that is consistent with this interpretation. Robison and Schmid (1991) conducted a mind experiment in which people were asked to set a price for selling a used car of known value to people whom the subject might be expected to have varying degrees of affinity such as family member, friend, stranger, and nasty neighbor. People indicated willingness to lower the price to those close to them (make them a gift). When the game was switched from selling to buying, a different pattern emerged. If you pay more for a friend's used car, you have the opportunity to give them a gift of the same dollar amount that you made by accepting a lower price for a good you are selling. The data however revealed that people did not offer the same break when selling as when buying to the same affinity group. Here as in the ultimatum and dictatorship games, there is an asymmetry. We reasoned that in the used car experiment, the gift was much more obvious when you paid more than the market price and one risked being labeled a chump and evoking embarrassment and a negative reaction from the beneficiary. The beneficiary can more easily accept a gift via a lower than market price when buying by chalking it up to clever negotiation on their part, but a gift resulting from someone paying the beneficiary more than the market price stands out like a sore thumb. Camerer and Thaler use the term \"insulting\" to describe an unfair proposal in an ultimatum game. Likewise, a gift that is too obvious may spoil the relationship.
Both sets of games exhibit an asymmetry which could be caused by the difference in the actual or expected behavior of the other party. This behavior is not necessarily incompatible with the concept of manners. An unfair proposer in an ultimatum game is both rude and undeserving of the recipient's care and benevolence (exempts the individual proposer from the general feeling recipients have toward their fellow humanity that is held until proven otherwise). The unfair proposer is no longer the average neutral person who the dictator had agreed to share with. Rudeness (lack of manners) could be the term for anyone who indicates they have little affinity for the actor signaled by unfair (or embarrassing) proposals.
There may be other reasons to prefer an explanation of manners to altruism. Learning of manners is emphasized by Camerer and Thaler. Young children are found to accept minimal offers in ultimatum games that older children and adults reject. This can be attributed to experience in repeated games that unfair offers to others result in your forgoing gainful opportunity. This seems to root the experimental data in more familiar selfish rationality. The authors speak of learning that \"long-run concerns outweigh the short-run costs.\" p.218. But then this can't explain the benevolent behavior in one-shot games or leaving tips in restaurants you never expect to visit again. The authors \"prefer to think that people have simply adopted rules of behavior they think apply to themselves and others, regardless of the situation.\" p. 218. Why can't this be applied to altruism as well as manners? Surely behavior can be reinforced by more than monetary payoffs. From the responses of other people, we can learn to love other people as well as to love money.
The concept of social capital may provide a better explanation of the evidence than either manners or altruism. Altruism is often used in the economics literature to refer to a taste for giving, which as the authors' note in the introduction of their article is often assumed to be stable. However, the authors clearly believe that people learn and preferences change. Their reference to hotel room cleaners who leave their names in the room is an example of an attempt to change the guests's preferences as applied to a specific individual. Social capital is a coefficient on own and other's welfare in one's utility function which is an evolving function of the relationship and affinity between the parties. The authors illustrate it when they observe that allocators in dictatorship games offer more of the pot to friends than strangers. The value of other's income in the allocator's utility is a function of affinity. And, if the other person signals their lack of affinity for the allocator or the responder, the behavior is different. People learn and are reinforced within the transaction as well as outside.
References:
Camerer, Colin, and Richard H.. Thaler, \"Anomalies: Ultimatums, Dictators and Manners,\" Journal of Economic Perspectives, 1995, 9:2, 109-220.
Schmid, A. Allan, and Lindon J. Robison, \"Applications of Social Capital Theory,\" Journal of Agricultural and Applied Economics, 1995, 27:1, 59-66.
Robison, Lindon J., and A. Allan Schmid, Interpersonal Relationships and Preferences,\" in Roger Frantz, et.al eds., Handbook of Behavioral Economics, Vol. 2B, 347-360. Greenwich: JAI Press, 1991.
Staff Paper No. 95-59
Dept. Agricultural Economics
Michigan State University
Sept. 1995 |
|
发表于 2005-8-24 16:11:07