This award funds research in behavioral economics that will analyze the possible relationships between economic behaviors using new data and new modeling methods.
Over the past thirty years, behavioral and experimental economists have made great strides in identifying behaviors that are hard to explain using the classic model of economic decision making: these include violations of expected utility theory such as the Allais paradox, ambiguity aversion, present bias, loss aversion and the endowment effect. All of these behaviors have been shown to affect important economic decisions, and have been the subject of enormous amounts of attention by researchers. However, far less attention has been paid to the links between these behaviors. This project examines these links both empirically and theoretically.
The first part of the project employs a laboratory experiment to measure a long list of well-known behavioral phenomena, including the ones listed above, in a single group of subjects, allowing the researchers to estimate the empirical relationship between them. Using these data they are then able to test the predictions of existing models about these relationships, and can use these results to guide their theoretical model. The second project develops part of that model, through a link between two particular behavioral biases: the Allais paradox in choice over risky prospects (i.e. where the probabilities of different outcomes are known) and the Ellsberg paradox in choice over uncertain prospects (i.e. where the probabilities of different outcomes are not known). Both behaviors can be explained by a unique behavioral axiom -- a generalized preference for hedging. This model will potentially be helpful in understanding behavior in environments that have both risk and uncertainty (such as financial decision making) and to capture more precisely the agent's attitude towards each of them. The third project extends the empirical work from project 1 by repeating the same experiment with a different subject pool, focusing on: (1) a representative sample of the US population; (2) subjects who suffer from anxiety and depression; (3) subjects primed into various emotional states.
Because the economic behaviors studied in this project are so ubiquitous, this grant could have a wide impact in many areas of the social sciences. For macroeconomists, who have long understood the impact of these biases on economic performance, understanding the relationship between them is necessary for understanding the distribution of economic outcomes. For policy makers, an understanding of any common cause to these biases could prove to be very helpful for any attempt to rectify them. For neuroeconomists and psychologists, such an understanding could aid the search for the biological bases of economic behavior.
