Imagine a game in which two people–strangers–are told they will be given $100 to share, and that one of them will have the power to decide how much to offer the other.

But the game isn’t quite the take-it-or-leave-it situation it first appears to be. The rules give the recipient the power to reject the offer, in which case neither player will get anything.

Now come the decisions that make it a game of calculated guesses. How generous an offer will it take to secure the recipient’s acceptance? What consequences should the two players weigh, especially if they are told that their decisions will not be known by others, and that their paths will not cross again–in other words, that reputation and fear of retribution are not to be considered? How will they perceive their self-interest, and act accordingly?

Economists will recognize this as the Ultimatum Game, a widely taught experiment that is now also studied by nearly 100 Harvard Law School students in a course titled Rationality, offered jointly in the fall by HLS Professor Christine Jolls ’93 and Faculty of Arts and Sciences Professor Amartya Sen, the 1998 Nobel laureate in economics.

The Ultimatum Game and several others derived from it are thought to shed light on the rationality–and irrationality–of human decision-making. Some economists might say that the most “rational” expectation is that the offeror will tender the smallest possible amount and the recipient will accept it (on the premise that something is better than nothing). Wrong.

The experimental evidence is that the offer is often quite generous–even as much as a 50-50 split. Even more surprisingly, perhaps, the recipient frequently rejects anything less, a response widely interpreted as a willingness to pay a price to punish unfair splits. One lesson: People do not act solely to maximize their economic gain. Selfishness is not always what motivates people in a free-market economy. Students in the class are asked to ponder: What is “rational”? How is self-interest to be defined?

These questions are important for law, say Jolls and Sen, because the design and implementation of legal rules are ultimately about predicting and regulating how people will behave.

“It is common to understand legal rules against a background assumption that people will always act rationally in response to them,” Jolls says. “How and when we regulate can be informed and improved by an understanding of the likely behavior of those who are regulated.”

Furthermore, she says, “we tend to think that the law should protect people from selfish or advantage-taking behavior by others. But one of the things we learn from games such as the Ultimatum Game is that people’s choices are not always well-predicted by traditional economic theory, and the consequences of people’s ‘wrong’ choices are not always bad, and in fact things can turn out quite well.”

“Sometimes, this gets at the fundamental question of whether and when a particular law may really be needed.”