Working Papers and Manuscripts in Progress:
This paper investigates some of the causal factors which increased the incidence
of slave insurrections and conspiracies in the antebellum Southern United States.
The analysis relies on a novel dataset, which is an amalgam of geographic data,
occupational data from the Integrated Public Use Microdata Series (IPUMS),
county-level census data and a compilation of incidents of slave unrest as recorded
by Aptheker (1993). I find that slave insurrections and conspiracies were more
likely in areas which, due to natural geographic factors, enjoyed a greater degree of
productivity in cotton production, and were less likely in areas naturally endowed
with a greater degree of productivity in tobacco. I also find that conspiracies
were more likely in counties geographically situated closer to Southern cities.
To interpret these results, I develop a theoretical model of slave insurrection
which incorporates slaves’ incentives to rebel, as well as slaveowners’ incentives
to surveil. I find that if the strenuousness of the work regime is severe, as was
the case on cotton plantations, and if the costs borne by slaves in the wake of a
rebellion decrease, as was the case in urban areas, then revolt is more likely and
slaveowners are more likely to devote greater resources toward stymieing such
outbreaks.
Ties That Bind: The Political Economy of Coerced Labor
This paper develops a game theoretic model of coerced labor, with particular
applications to serfdom in Late Medieval Europe and slavery in the antebellum
United States. Two of the more prominent explanations for the abrogation of
this institution in Western Europe are critically examined. The model predicts
that reductions in population are generally associated with less coercion, in accordance
with the population-based Malthusian theory. More profitable outside
opportunities for laborers and price inflation, in some cases, decreases coercion as
well, which is interpreted as evidence in favor of the market-based commercialization
theory. The theory also provides explanations for why these same factors did
not bring about the demise of serfdom in Eastern Europe, a puzzle posed in the
famous Brenner debate. Because greater coercion increases output per laborer,
the model also accords with the Fogel and Engerman (1974) finding that slave
labor was productive in the antebellum United States.
Antebellum Slave Revolts and Urbanization in the United States
This paper extends previous research on the causal factors behind antebellum slave revolts. In particular, an influential strand within the economic history literature, which attributes the relative decline of Southern industry and urbanization to the inherent difficulty in supervising slaves in an urban environment, is critically evaluated. I find that the probability of a slave insurrection event is not correlated with the degree of urbanization in a given county, and this is interpreted as evidence against this hypothesis. Moreover, these results are qualitatively identical when an instrumental variable strategy is employed to rule out potential endogeneity.
“A Theory of Love and War: The Vietnam Draft, Assortive Mating, and crime in the 1980’s and 1990’s” with Brian Asquith, Nanneh Chehras, and Jose Luis Luna Alpizar. Abstract forthcoming.
Analysis of COVID-19 Incidence In New York City (Report)
Analysis of COVID-19 Incidence In New York City (Presentation)
The COVID-19 pandemic has had a dramatic impact on the economic, social and
personal lives of Americans across the country, and this effect has been perhaps most
keenly felt in New York City. The ongoing national dialogue regarding when and how
to re-open shuttered businesses underscores the need to understand the types of social
activities and venues most strongly correlated with the initial spread of the virus. To
that end, this report investigates the incidence of COVID-19 cases in New York city by
leveraging Foursquare data on social venues throughout the city. Regression and machine
learning techniques are utilized, with population data from the U.S. Census Bureau as
a control. I find that the number of transportation terminals in a given zip code is a
statistically significant predictor of COVID-19 cases, and this result is robust to a number
of model specifications. I also find that the number of restaurants and recreational venues
is not a positive predictor of COVID-19 cases in the early stages of the outbreak, nor a
positive predictor of the growth rate in cases between 4/1/2020 and 5/12/2020, the time
of writing.