Saturday, September 25, 2010


Ben Bernanke looks at what the financial crisis means for economics. In the middle of this there is an overview of information and expectations under uncertainty.

Most fundamentally, and perhaps most challenging for researchers, the crisis should motivate economists to think further about their modeling of human behavior. Most economic researchers continue to work within the classical paradigm that assumes rational, self-interested behavior and the maximization of "expected utility"--a framework based on a formal description of risky situations and a theory of individual choice that has been very useful through its integration of economics, statistics, and decision theory.9 An important assumption of that framework is that, in making decisions under uncertainty, economic agents can assign meaningful probabilities to alternative outcomes. However, during the worst phase of the financial crisis, many economic actors--including investors, employers, and consumers--metaphorically threw up their hands and admitted that, given the extreme and, in some ways, unprecedented nature of the crisis, they did not know what they did not know. Or, as Donald Rumsfeld might have put it, there were too many "unknown unknowns." The profound uncertainty associated with the "unknown unknowns" during the crisis resulted in panicky selling by investors, sharp cuts in payrolls by employers, and significant increases in households' precautionary saving.

The idea that, at certain times, decisionmakers simply cannot assign meaningful probabilities to alternative outcomes--indeed, cannot even think of all the possible outcomes--is known as Knightian uncertainty, after the economist Frank Knight who discussed the idea in the 1920s. Although economists and psychologists have long recognized the challenges such ambiguity presents and have analyzed the distinction between risk aversion and ambiguity aversion, much of this work has been abstract and relatively little progress has been made in describing and predicting the behavior of human beings under circumstances in which their knowledge and experience provide little useful information.10 Research in this area could aid our understanding of crises and other extreme situations. I suspect that progress will require careful empirical research with attention to psychological as well as economic factors.

This can be tied to the modelling of expectations in the UIP model. There may be a non-linear model that is distributed with a negative skew and kurtosis in regular times and collapses into something binary in a crisis.

No comments: