This award funds research on two distinct topics in economic theory. The first project develops new approaches to the modeling of decision making subject to temptation. The idea is that a person views his/her own future self as driven by currently unknown 'desires' which may conflict with the person's current view of what he wishes to consume. Surprisingly, this approach turns out to be mathematically equivalent to alternative recent approaches where the person expects his future self to choose what to consume to maximize his current utility subject to certain 'self control' costs. The second project studies implementation with evidence. More precisely, the PI and his collaborator consider the optimal design of social institustions when agents may posess evidence to back up their claims. They show that the use of evidence has dramatic effects on what can be achieved via appropriately designed institutions. Since virtually all work on this topic has ignored the possible availability of evidence, these results have important implications for rethinking the theory of optimal institutional design.

This project is basic research in economic theory. It has broader impacts because it adds to a toolkit that researchers in economics, business, political science, and law can use to build models that incorporate important aspects of human behavior.

Project Report

There are two parts to this project. The first part considers mathematical models of temptation. The traditional approach in economics is to assume that economic agents are fully "rational" in the sense that they effectively pursue their own self-interest. This approach has come under criticism for many years from many different angles. A large part of the research here focuses on one particular way in which real agents are not so "rational," namely, temptation. It is a commonplace experience to be torn between conflicting objectives, one of which is clearly the "right" or appropriate action and one of which is tempting for its short-run advantages, despite its long-run costs. This project considers mathematical formalizations of this experience for use in economic model building. We show that some seemingly very different ideas about temptation only have different implications for observable behavior if one looks at a broad set of behavior. More specifically, if we only consider the kinds of commitments people adopt to cope with anticipated temptation, then "costly self control" and "random temptation without self control" cannot be distinguished. On the other hand, if we observe both commitments and the subsequent behavior under those commitments, then these different notions can be empirically distinguished. The second part of the project takes a theoretical approach to design of institutions when agents may have evidence to prove some or all of their assertions. It is not surprising that the existence of such evidence may make it easier for policymakers to design institutions to achieve desirable outcomes. After all, with evidence, agents can be required to prove their assertions so that policymakers can rely on these assertions. On the other hand, many true claims may be impossible to have evidence for. For example, while someone could prove he is a homeowner by providing a deed to a house with his name on it, how could anyone ever prove he does not own a home anywhere in the world? We show that with the proper design, a surprisingly small amount of evidence is needed to ensure that the policymaker's desired outcomes are achieved, even if the policymaker knows little about the goals of the agents and even if the agents know very little about one another.

Agency
National Science Foundation (NSF)
Institute
Division of Social and Economic Sciences (SES)
Application #
0851590
Program Officer
Michael Reksulak
Project Start
Project End
Budget Start
2009-04-15
Budget End
2013-03-31
Support Year
Fiscal Year
2008
Total Cost
$288,131
Indirect Cost
Name
Boston University
Department
Type
DUNS #
City
Boston
State
MA
Country
United States
Zip Code
02215