The project includes two research projects in economic theory. One concerns how to model temptation and the second concerns polarization of beliefs.
Temptation occurs in contexts where an individual understands that he may make future choices that are not in line with his or her long-run objectives. A simple example is food consumption; while someone may decide today to lose weight, actually acheiving this objective means facing daily temptation to eat unhealthy foods. While temptation is a universal experience, the formal economic model of how individuals make decisions does not include the possibility of temptation. Developing richer economic models that includes temptation allows us to better understand a wide range of human behavior that contributes to problems like obesity.
One way that individuals deal with temptations is by restricting the range of their future choices through unbreakable commitments they make today. The principal investigator builds and analyzes a formal mathematical model of how people make such commitments assuming that individuals are fully rational. In the model, there are two costs of temptation for individuals: the internal cost of attempting to resist the temptation and the effect of temptation on actual choice. The investigator analyzes the model under two different assumptions about the size of these costs. He first examines the case where the cost of resisting temptation is so high that the individual always succumbs to temptation. He then considers the case where the cost of temptation is lower, so an individual can resist temptation even without a commitment. In both cases the model allows for uncertainty about the extent and type of temptation that might occur.
The second project is also formal economic theory using mathematical modeling techniques. Here the PI develops and analyzes a model of how media bias may affect viewers' beliefs and decisions. The model starts with the assumption that individual voters are fully rational and seek information. They have a variety of information sources. They know that these sources (eg, the media) are biased, but they do not know the exact bias of a particular source. The project analyzes how the proliferation of media, each with a distinct bias, changes voter beliefs and decision outcomes. The goal is to explore whether in general the social costs of the change in beliefs due to media proliferation are not directly from the decisions made, but through losses in the process: either through wasteful competition for control or less efficient communication.
The project includes completed work on two topics and the initial steps of a third. The first main topic is in the area of decision theory and in particular studying models where individuals suffer from temptation. These individual would naturally want to commit today to avoid succumbing to temptations in the future. For example, locking money in a retirement account avoids splurging, and not buying fattening desserts while at the store avoids breaking one's diet, etc. There are two related but distinct models in the literature for studying these situations. In one the person knows she will give in completely to temptation. In the other the individual suffers a cost from resisting the temptation. These models lead to different predictions regarding what commitments will be made and clearly have different interpretations. We show that when uncertainty about the form temptations will take is added to both these models they become indistinguishable in terms of the types of commitments undertaken. This relationship also suggests new extensions of these models to richer dynamic environments. Another area is that of vote buying. Specifically, what is the effect of allowing share holders to sell their vote but not their share when a raider wants to take over a firm. It has been shown that even when vote trading per se is forbidden, there are trading methods (using derivatives) that are equivalent to trading votes. Moreover, vote renting is common and legal. Finally, the issue of whether votes should be legally trade-able has been discussed by regulatory agencies. Our research characterizes the efficiency costs of allowing vote trading that is separate from shares. It shows when the incumbent and when the raider benefit from allowing such trades. These results are useful for understanding the implications of allowing vote trading in the corporate environment. The final area is new research that just began with the end of this funding. It is well documented that there is a wage gap in favor of males and a job satisfaction gap in favor of females. We provide a model that predicts a wage and opposite job-satisfaction gap (although it does not predict whether this will occur for males or females). The idea is that individuals differ in their suitability for different jobs and their satisfaction from different tasks. When choosing an occupation they often must give up on what they enjoy in order to do what they are better at. However, how they perform this tradeoff will depend on how they expect their spouse to do so, since they know they need at least one high earner who focuses more on job suitability than satisfaction. The model makes novel predictions about various policies, such as the conclusion that changing the tax on high income households (where both men and women choose the job to which they are more suitable) will effect men and women in opposite ways: one gender will increase their focus on job satisfaction while the other on suitability.
