Research
I research the economics of peace and prosperity, with a focus on spatial relationships. Methodologically, I combine observational studies, economic theory, and laboratory experiments.
Publications
Journal of Economic Behavior & Organization 235, 107035. 10.1016/j.jebo.2025.107035
I examine the role of comparative advantage in the rise of city-states. After compiling a new dataset on ancient Greece and the natural environment, I provide evidence that the spatial covariance of natural vegetation endowments amongst potential trading partners is important for explaining the development of silver coin money, battles, and city-state formation. I also provide statistical evidence against “key factors” previously emphasized. These findings are consistent with a general model of comparative advantage that I develop, which clarifies how the spatial covariance of factor endowments affects multiple development outcomes.
w/ Patrick Fitzsimmons. Journal of Economic Behavior & Organization 234, 106997. 10.1016/j.jebo.2025.106997
We construct and analyze a database of warfare around the Mediterranean from 600 to 30 BCE. We provide the first in-depth statistical summary of war in this period, examining variation across space, time, and belligerents. Then we evaluate two main explanations for international peace: democracy and hegemony, using both statistical analyses and case study comparisons. We find no democratic peace among Ancient Greek city-states and mixed results, both inside and outside of Greece, about how war relates to state power. Broadly, our results challenge prominent political theories of peace and suggest more exploration of alternatives.
w/ Lucas Rentschler. Economic Inquiry 63(2), 545–567. 10.1111/ecin.13270
We develop a model where the police choose between investigating and patrolling, while civilians choose between producing and stealing. We derive a truth table for the equilibrium numbers of criminals and producers, punished or not, that can holistically evaluate the effects of police performance incentives. To test the model, we conduct an experiment that varies how severely an officer is reprimanded for false punishments. We find that stronger reprimands do not change crime, increase civilian incomes, and decrease false positives. We also find that the clearance rate, a measure of performance used widely in econometric studies, suggests police performance is better when it is unambiguously worse.
w/ Bart J. Wilson. Journal of Economic Behavior & Organization 219, 490–509. 10.1016/j.jebo.2024.01.032
We examine territorial behavior in the ecological conditions that foster conflict. We develop an economic model that isolates the effects of resource skew on territorial ranges, as well as their interactions with unequal appropriation abilities. We then conduct a controlled laboratory experiment to test the predictions of our model and find that observed behavior tends to cluster around the equilibrium predictions and that all comparative statics have the predicted sign. Additionally, we find that equally strong appropriators select more exclusive and less overlapping ranges than what is predicted with symmetric resources, while weaker appropriators choose more engulfed ranges than what is predicted with skewed resources.
w/ Lucas Rentschler. Public Choice 196, 223–227. 10.1007/s11127-023-01089-2
In recent years, we have learned much about how police, defendants, and prosecutors are affected by different policies. At the same time, economic theory is being forgotten or disregarded. Even more so today, economists and political scientists treat “the enforcement apparatus of police, courts, prosecutors, and legislature as a philosopher-king, with imperfect knowledge but only the best of motives” (Friedman 2001). This special issue contains a sampling of papers at the intersection of criminal justice and public choice. Our introduction discusses the recent literature on criminal justice and calls for a multifaceted empirical approach that incorporates the insights of public choice theory.
w/ Erik O. Kimbrough. Journal of Peace Research 60(2). 10.1177/00223433221075006
This article introduces a simple application of contest theory that neatly captures how Boulding’s “loss of strength gradient” determines the geographic extent of territory. We focus on the “supply side” of territorial conflict, showing how the costs of initiating and escalating conflict over spatially dispersed resources shape the nature and scope of territory. We show that economies of scale in the production of violence and varying costs of projecting power at a distance combine to affect the intensive and extensive margins of conflict, and ultimately the geographic distribution of territory. Comparative statics analysis shows how the distribution of conflict and territory change as costs change, helping shed light on, for example, why new transportation technologies have historically led to a redrawing of territorial boundaries.
Journal of Economic Behavior & Organization 190, 838–850. 10.1016/j.jebo.2021.08.009
Cities and marketplaces are central to economic development, but we know little about the mechanisms that cause such agglomerations to form. I theorize that evolutionary forces select for agglomerations when individuals desire to spatially coordinate exchange in complex environments. To test this idea, I perform a laboratory experiment where geographically dispersed individuals bring different goods to a location for trade. Consistent with the theory, I find that individuals spontaneously coalesce to reap the gains from exchange and there is more agglomeration in economies with a larger variety of goods. I also find that agglomerations re-emerge at the same locations after shocks, being land-tied reduces agglomeration but magnifies the effect of variety, individual location choices aggregate to create a Zipf population distribution, and individuals earn more in agglomerations.
Explorations in Economic History 76, 101324. 10.1016/j.eeh.2020.101324
How does a governing coalition’s size affect the extent and type of violence in society? The model developed here predicts that larger coalitions are less likely to fight for private goods (e.g., plunder) than for public goods (e.g., defense), yet this substitution need not reduce the overall scale of fighting. That prediction is tested by investigating how Rome’s transition from Republic to Empire affected military patterns. The raw data and three empirical tests suggest that the Republic engaged in more battles overall and that Republican battles had more of a public goods component. This study furthers our empirical knowledge about the ancient world while bringing data to bear on contemporary debates about the causes of peace and war.
Public Choice 186(3), 467–490. 10.1007/s11127-019-00763-8
Is there more violence in areas with many small countries or only a single large one? I build on Bernholz (1985) to create a unifying framework where both internal and external contestants engage in conflict, and then summarize how the spatial configuration of countries affects all types of violence with the Herfindahl–Hirschman Index of state sizes. Empirically, I examine fatalities from the conflict in Africa, where I use the borders set by the colonial powers of Europe to identify the effect of concentration. I find the most fatalities in areas with many small countries, but that violence decreases with concentration at a decreasing rate and eventually increases in areas with only one large country. These findings suggest an important difference between the observed average effect of concentration on violence and the expected marginal effects of further concentration.
Working papers
SSRN 4091833. [Revisions Requested at Oxford Bulletin of Economics and Statistics].
Many scholars study conflict using ethnic groups as the unit of observation, but these groupings may be shaped by the same factors that influence conflict. To better understand this “endogenous areal unit problem”, I create an econometric model of ethnic maps and explore how the resulting areal units affect statistical inferences about conflict. I first estimate several competing models of the Murdock map of African ethnic groups and do not reject the hypothesis that the map is endogenously determined by environmental data. I then find that the Murdock map generates more spurious spatial correlations than completely random maps which, in turn, generate more spurious spatial correlations than equal-area geocells. I also identify a distinct problem when the same variables determine both conflict outcomes and areal units. Provisional recommendations are also provided for researchers who use ethnic maps in regression analysis.
w/ Dan Stephenson. SSRN 4869123. [Revisions Requested at European Economic Review].
Is war good for economic development? Several macrohistorical theories suggest “yes,” yet the supporting evidence relies on war being an exogenous shock. We first examine the raw data from Europe, showing a negative relationship between war and production in both ancient and modern periods. We then create a model to investigate what typical statistical analyses show on data generated with standard conflict mechanisms. We show that typical statistical analyses suggest war positively affects growth in our model economy when it is in fact harmful. We caution against “mostly harmless” regressions that suggest European prosperity comes from its harmful history rather than its peaceful present.
w/ Lucas Rentschler. SSRN 4804107.
We propose a modified Groves–Ledyard (1977) mechanism that estimates the demand for public goods using population samples and selects an approximately optimal quantity. We then test our mechanism in a virtual Phase-0 trial that mirrors existing experiments, with known utility functions over generic commodities, and find our mechanism performs extraordinarily well. Further tests in a Phase-1 experiment, where participants choose the quantity of police in a virtual environment, show that our mechanism provides quantities that are close to theoretically optimal. Our proposed mechanism thus performs well in laboratory experiments and is practical to implement and scale in the real world.
I develop a model of the spatial extent and intensity of war motivated by new stylized facts about the major war in two different eras: the Second Punic War (218–201 BCE) and the Hundred Years War (1337–1453 CE). I then provide analytic results about how battles are generated and observed, how their frequency depends on distances to both bases, and the link between military and production uncertainty. The model generates battle maps and distance relationships that qualitatively match the historical observations, and also suggests inherent limits on the ability to predict the costs or locations of conflict. Still, fitting a simple version of the model to data provides remarkable out-of-sample performance.
w/ Frank Simmen. SSRN 5100513.
We conduct a comprehensive analysis on the long-run relationship between homicide and economic development. We first compile the largest known dataset on English historical homicide rates, which allows us to document clear but complex relationships between homicide, wages, and GDP per capita. We then create a model of our stylized facts, where conflict reduces productivity and this in turn fuels homicide, as well as a state-space extension that allows us to assimilate and reconstruct historical homicide rates. Our model predicts better than leading statistical competitors and also suggests that economic explanations are more important than previously recognized.
w/ Warren Anderson and Lucas Rentschler. SSRN 5381962.
What causes crime to concentrate in a small set of locations? To answer this question, we employ a three-stage methodological approach that moves from observation to theory to experimental validation. We first revisit the micro details of the 1967 riots in Detroit, a major motivator for grievance theories prominent in sociology and psychology, and identify key empirical patterns in the spatial distribution of crime. Based on these observed patterns, we develop an economic model where civilians choose where to commit crimes while officers choose where to patrol, generating testable predictions about the conditions under which crime concentrates. Finally, we conduct a laboratory experiment that tests these theoretical predictions in a controlled environment.
Work in progress
Is war associated with agglomeration? To find out, I develop a nonstationary spatial cross-correlation function (CCF) that captures spatial dependencies which vary across locations. Applying this method to data from Modern Europe and the Ancient Mediterranean, I uncover three general patterns: cross-correlations are globally positive, locally mixed, and weaker than the autocorrelation of population itself. The first two patterns suggest that war happens where people are, broadly, and that war near cities either pushes people inwards or results from larger cities exerting more force in their periphery. The last general pattern suggests that either war impacts urban patterns less directly than intrinsic forces or that war is strategically dispersed. By uncovering these spatial patterns, I provide new insights into the complex and underexplored dynamics between conflict and urban development.