Graduation Year


Document Type




Degree Name

Doctor of Philosophy (Ph.D.)

Degree Granting Department


Major Professor

Andrei Barbos, Ph.D.

Committee Member

Gabriel Picone, Ph.D.

Committee Member

Joshua Wilde, Ph.D.

Committee Member

Benjamin Craig, Ph.D.


reputational effects, credence goods, rationality, real-world outcomes


In the first chapter, I investigate reputational effects of the disclosure of negative information in a market affected by adverse selection. A series of recent discoveries has increased consumer concern over the presence of counterfeits in the market for fine and rare wine. For the thousands of bottles sold at auction each year, house reputation is used as a quality assurance mechanism to signal product authenticity. Using sales data from 2005-2015 for the ten largest auction houses, I study consumer reaction following two recent disclosures of an auction house having offered or sold counterfeit wine. My identification strategy to examine reputation involves a series of triple difference regressions analyzing equilibrium prices and quantities. I discover one house experienced no losses following a 2008 incident involving 107 counterfeit bottles. However, three houses associated with a 2012 incident involving thousands of bottles were found to have suffered significant reputation losses following the incident. These losses are demonstrated by a 3-8% decrease in equilibrium sales prices and a 6-9% decrease in sales quantities in the year following the disclosure.

The second chapter of my dissertation involves the transitivity of stated preferences. Revealed preference theory states that, in order for an individual’s preferences to be consistent with utility maximization, they must satisfy the principle of transitivity. Any deviations from this principle result in a logically inconsistent response pattern. I develop a new framework to study the rationality of stated preferences, accounting for both the number and severity of non-transitive responses an individual makes. I implement this method using a nationally representative survey of 3,234 respondents from the U.S. general population and discover that more than 52% of the population exhibit non-transitive preferences. In addition to measuring the number and severity of non-transitive preferences exhibited by each respondent, another aim of this manuscript is to evaluate the relationship between response transitivity and the individual outcomes of each respondent under the premise that high quality decisions are the result of greater decision-making ability. After controlling for demographic characteristics including age, education, race, gender, ethnicity, and work status, non-transitive patterns are correlated with lower incomes and poorer health.

Included in

Economics Commons