Author's School

Graduate School of Arts & Sciences

Author's Department/Program



English (en)

Date of Award

Spring 4-26-2014

Degree Type


Degree Name

Doctor of Philosophy (PhD)

Chair and Committee

David K Levine


This dissertation focuses on two issues in Public Economics: public goods provision and

voting theory. It consists of three independent papers, with two using experimental

methodology and one using empirical data to examine the effectiveness of the models. In the following I briefly summarize each of the papers.

Economists have long understood the challenges of providing proper incentives to groups, such as divisions or teams of a firm, that produce a joint product. The first paper (chapter 1) proposes a mechanism in which a firm creates a competitive environment for its two teams by awarding prizes based on aggregate outputs produced by these two teams, and uses laboratory experiments to examine how effectively it induces team members to contribute. The experimental results verify the prediction that the proposed mechanism encourages a greater number of participants to make contributions, compared to a simple profit-sharing scheme. I also find that participants contributed significantly more when they believed that their team had lower output, which can be well explained by a model that incorporates the effect of envy at the group level.

The second paper (chapter 2) is an experimental examination of information revelation in a voting model. Typically parties can conduct public events, such as rallies or demonstrations that reveal their level of support in hopes this might influence voter turnout and the outcome of the election. How effective is this? I compare two information-revealing mechanisms in the Palfrey-Rosenthal pivotal voter model: one through which active supporters show their support without paying costs ("polls"), which can be viewed as cheap talk; the other where active supporters have to pay their time (active participation in "campaigns") or money (e.g., contributing to super PACs) to support their preferred candidates, thus providing more certainty about the actual level of support. To capture the difference between the two mechanisms, I assume that polls reveal the distribution of active supporters of a party, while the campaigns provide the actual numbers of the active supporters of that party. There are two main experimental findings. (a) In most of the situations, subjects followed the main ideas of the Palfrey-Rosenthal pivotal voter model, with appropriately responding to the cost of voting and the belief of being pivotal. (b) However, when subjects are informed of being in an advantageous position by campaigns, their turnout becomes significantly higher than the best response to their pivotality belief. This can be attributed to that leading in an interim stage has a positive psychological impact on performance in tournaments.

The third paper (chapter 3) investigates candidates' rallying strategies in two-party races. Like chapter 2, it views campaign rallies as an information-revealing mechanism that allows candidates to project images of strong current support among voters. By incorporating this mechanism into the Palfrey-Rosenthal pivotal voter model, this study can explain under what circumstances a candidate should hold a rally and how that rally affects voters' decisions regarding whether or not to vote. The investigation hypothesizes that (a) when two parties are different in size but have the same chances of strong base support, the larger party is more likely to hold a rally, while (b) when the sizes of the two parties are equal but base support is unequal, the party with a smaller probability of strong base support is more likely to hold a rally. These two hypotheses are supported by the empirical analysis of the 1988, 1992, and 1996 U.S. Presidential elections.

To summarize, this thesis contributes to the literature on group incentive mechanisms, with chapter 1 being the first experimental study on inter-team competition with an endogenously determined prize level in a stage game. This thesis also contributes to the literature on voting behavior and campaign strategy. Chapter 2 is the first experimental study on the effect of information revealed through campaigns on voting behavior in the Palfrey-Rosenthal model. Chapter 3 supplements political science literature on campaigns by using pivotal-voter theory to analyze the effect of holding rallies and how that effect influences candidates' rallying strategy.


This work is not available online per the author’s request. For access information, please contact or visit

Permanent URL: