##### October 21, 2011

Two SFI scientists say in a yet-unpublished paper that "hyperbolic discounting," a mathematical method for valuing future events that has been largely rejected by economists, can be more rational than economists' traditional methods.

A lack of concern for the future is built mathematically into traditional economic theory. Economists' models "discount" future events: a resource might be valued at $100 if it's immediately available, but only$97 if it is only available in a year's time. Value goes down consistently with time. As a result, the value of assets decreases exponentially, and is effectively zero within decades or centuries. Thus, economics effectively ignores the value of far-future events through "exponential discounting."

But real humans behave as if resources do still have a value significantly greater than zero in the distant future, a tendency called hyperbolic discounting that economists say is "irrational" because it is not "time consistent." Under hyperbolic discounting, a person has a strong preference for getting something today rather than tomorrow, but only a weak preference for getting it on day 100 rather than day 101; yet when day 100 arrives, they will strongly prefer to get the resource immediately.

In their paper describing the results of new modeling, SFI External Professor John Geanakoplos and Professor J. Doyne Farmer show that when you don't know how the economy is going to change, hyperbolic discounting makes better mathematical sense than exponential discounting, particularly for far-future events. The work could have significant implications for public policy, such as decisions to make current-day investments to mitigate far-future climate change.

Geanakoplos says the important thing is to force economists to rethink their assumptions. "If you change the intellectual climate, you're going to change the political climate," he says.