Kara A. Contreary

PhD Candidate, Economics

Home     Research     Teaching     CV     Contact
 

Re-election of Corrupt Politicians (MRes Paper)

I present a two-period political agency model with re-election. There are two parts: a benchmark with only rational voting and an extension with campaign spending. I find that, if electoral results are sensitive to campaign activity, good politicians (those who wish to maximize voter welfare) can be induced to behave corruptly. Rational voters realize this, and are willing to forgive corruption in politicians they believe to be good. A preliminary welfare analysis shows that, in some instances, voters may prefer to leave the option of campaign spending open for the information it provides, and also may prefer that good types act corruptly in the interest of re-election.

Optimal Auditing in Networks

This paper considers optimal auditing strategy for an authority who wishes to incentivize networked agents to behave honestly.  Results turn on a behavioral assumption concerning the reporting behavior of honest agents.  If honest agents can be expected to inform the auditor of dishonest behavior in their neighborhoods, then the auditor should focus most of his auditing effort on more peripheral agents.  If, however, honest agents will keep quiet about dishonest actions they observe, the auditor cannot rely on internal monitoring and must use a more symmetric auditing strategy.

Own-Ability Bias and Group Decision-Making

I analyze a 3-person committee that must decide between two candidate policies. Two members receive an informative signal about the most appropriate policy. The third member is uninformed, but suffers from overconfidence that leads him to think his noisy signal is informative. When he votes according to his signal, adding noise to the outcome, the lesser-informed of the other two agents strategically ignores her signal under majority rule, effectively trading bad information for good. Because of this behavior, majority rule ceases to be the preferred voting rule. Social welfare is instead maximized by selecting a voting rule that requires unanimity, and in some cases requires unanimity for the a priori more likely policy. This is never the case in the absence of overconfidence. Thus, bias in one member's self-assessment leads to a reversal of preferences regarding the constitutional voting rule.

Dissention (in progress)