Pietro Ortoleva, National Science Foundation

Incomplete Preferences, Stochastic Choice, and Time and Risk Preferences

Economists who study how people make decisions have for many years started from the assumption that if an individual is faced with a choice between different alternatives, he or she can order the options from "most preferred" to "least preferred", as well as a number of other assumptions. However, a wide range of evidence now demonstrates that this theory is not able to explain how many people make certain specific kinds of decisions -- for example, the highest price someone is willing to pay for an object is often lower than the lowest price he or she is willing to accept to go without the same object. This award will fund the PI's continued efforts to develop alternative models of behavior based on the idea that individuals may not be able to compare all options; in other words, people have incomplete preferences. The PI plans to consider whether and how this theory can account for a range of document decision making behaviors. He also plans to conduct decision making experiments in a lab environment to test hypotheses drawn from the theory. The results will advance our scientific understanding of this kind of human decision-making, and could be useful to businesses and governments that need to predict how individuals will respond to new products and policies.

The research will focus on developing a model in which subjects' preferences may be incomplete, and using that model to analyze several specific biases in decision making: the gap between willingness to accept and willingness to pay, certainty effects, decision in stochastic environments, and intertemporal choice. The goal is to provide axiomatic characterizations of models in which subjects chose to randomize, including a special case where the desire to randomize comes from the agents' uncertainty about her own utility of consumption. The project will also include work on preferences on lotteries over consumption streams and will introduce an axiom that captures aversion to variability of consumption over time.