Tuesday, March 4, 2008

Cheering over Harvard Girl!

We here at Capyblanca are cheering over our own Harvard girl; who would imagine?

More seriously, we are celebrating the thesis defense of Mrs Anne Jardim, on the ultimatum bargaining game. Anne is an economist, and she spent the last months completing her research at Harvard Law School. We would never miss the chance to poke some fun at her celebrate her achievements. Here's a peek at its conclusion.

Most of economic theory and the literature on decision-making rests upon the assumptions of rationality and maximization of utility. In this thesis, we have provided a review of the modern research literature concerning the ultimatum bargaining problem.

The ultimatum bargaining problem arises in asymmetric situations in which a known amount will be split between two actors--one of which is a proposer for the split, while the other, the responder, accepts or rejects the offer. While the proposer is in a better strategic situation, the responder has the power to block the deal, to the detriment of both proposer and responder. This is not only a recurring problem in applied game theory and economics, but also a theoretically interesting one.

It is recurring because it models a large type of ultimatum situations. It should arise in domains as diverse as biology to human relationships to economic behavior between firms to international relations. When a male marks its territory, that is a kind of ultimatum; it is up to other males to accept it or reject it by fighting. When companies fight publicly, they usually send ultimatum offers through the press: "Unless Apple is willing to alter pricing behavior, NBC will stay out of iTunes". In fact, in many kinds of conflicting interest scenarios, ultimatums are an important part of the bargaining process. The particular model studied here represents an important set of these situations, and is of great importance in the real world.

Moreover, it is also theoretically interesting, because humans do not respond as economic theory would predict. Quite the contrary: human behavior is enormously far from the expected rational behavior.

This fact has triggered an enormous amount of scientific interest in this game. Many different types of studies are being conducted now. On the table below we present a taxonomy/classification of such studies. This table characterizes our critical review of the literature.

There is not yet a consensus on why people deviate from the expected Nash equilibrium, but these deviations from rationality are informative about human cognition. Current economic theory is based on the normative model of decision-making: decision-making is treated as maximization of utility. However, if that cannot be expected to hold even in very simple scenarios, such as the one studied here, new mathematical models may eventually replace the standard "rational actor" model.

These new models should be as general and applicable as the standard rational actor. But they should also be psychologically plausible. As we have seen, progress in understanding ultimatum bargaining is steady. In the coming decade, as new data and new models are discussed, a consensus may form. As we have seen, ongoing research on ultimatum bargaining, ultimately, may turn out to bring sweeping changes into the nature of economic theory.