Banking on African-American Businesses
Between 1912 and 1937, 109 banks owned and managed by African-Americans (African-American banks) operated in the US. This project collects a new data source that reveals twenty additional African-American banks not previously known in the literature. I use this new data set to determine the effect of African-American banks on the African-American community. Panel fixed effects analysis shows that an additional African-American bank per thousand African-American adults in a county doubles the share of African Americans that report employing one or more people, a proxy for business ownership.
Too Small to Not Fail: Virginia African-American Banks, 1915-28
This project looks at African-American banks in extensive. Creating data set of Virginia state banking records, this project contrasts African-American banks with nearby non-African-American banks, and finds that African-American banks were smaller, less profitable, and received more negative comments from examiners (Work in progress, paper coming soon).
The Disappearing Gay Income Penalty
Economics Letters, 2013 (with Purvi Sevak)
Since 1995, labor economists have reported on the income disparities between individuals who engage in same-sex behavior and those that do not. Many of these papers report a significant wage penalty, while others find no effect, but few look at the trend over time. We find, using National Health and Examination Survey (NHANES) data from 1988 to 2007, that the income gap has reversed over time from a penalty to a premium.
Other research projects
Newfoundland’s Two Transitions
This survey uses primary and secondary historical sources to summarize the extraordinary events in Newfoundland in the early twentieth century, as the self-governing dominion becomes the only nation to vote to end representative government, and later joins the Canadian federation.
Over-Investment and Over-Agreement in the Shadow of Conflict (with Barry Sopher)
This paper reports the results of a bargaining model that provides agents with a choice between production and protection in a setting with a winner-take-all disagreement point, and then examines this model by conducting a laboratory experiment. Using different endowments and different contest success functions, we find that, while the risk-neutral mutual best response predicts agent behavior well, most agents over-invest in preparing for disagreement. Additionally, we find that many agents compromise rather than engage in conflict, even when it is more profitable to do so. In the experiment, subjects agree approximately 40% of the time, much higher than the predicted rate of 0%.