My research focuses on international conflict and cooperation, with a heavy emphasis on statistical analysis and formal modeling. In particular I am interested in the political economy of military conflict; the causes and effects of conflict management and other types of interventions in wars and crises; bargaining theory and strategic communication; contract theory and its application to rebel organizations; the integration of micro-level data in international security studies; and the development of quantitative methods in the social sciences. Please see my research statement for details on three current projects (including my dissertation) and three future projects that speak to this broad research agenda.

Dissertation

International Mediation of Wars: Causes and Consequences

My dissertation asks two related questions: First, what explains the number and type of international mediation attempts that wars experience? Second, what kinds of mediation are more likely to lead to a settlement?

The theoretical and statistical literature on mediation rarely addresses these questions jointly, despite the error this could introduce in our understanding of effective mediation, if different types of mediators select into different types of conflicts. I provide a theoretical account of such a selection process by developing a game-theoretic model of both mediation initiation and implementation. I then evaluate hypotheses derived from the model using an original dataset of both mediation onset and its implementation in wars, constructed from more than 15,000 newspaper and wire reports published between 1990 and 2005.

In my dissertation's empirical work, I propose that we can causally identify and adjudicate between competing claims about the efficacy of mediation by using event timing as an instrument for whether talks were mediated or not: Mediation activity peaks in early summer when many legislatures are in recess and drops in the weeks prior to and during the UN General Assembly as well as at the end of the business year.

Among other results, I find that mediated talks are significantly more likely to lead to a settlement than non-mediated talks, but only if we account for strategic selection. Conversely, multi-party mediation appears to be more effective than single-party mediation in a naive regression, but not so when we adjust for endogeneity and selection.

Papers

"Summer Holidays and Conflict Resolution: Event Timing as an Instrument for the International Mediation of Wars". For an abbreviated version, see the poster. Additional information about the original mediation dataset I created, including coding procedures, can be found in the codebook.

"A Formal Theory of the Initiation and Implementation of Mediation in Wars". See also this presentation.

"What the Numbers Say: A Digit-Based Test for Election Fraud Using New Data from Nigeria" (with Alexandra Scacco), under review. We leverage psychological biases to detect election fraud by only looking at digit patterns on return sheets: People favor small numbers and avoid repetition when manipulating numbers. Link from Freakonomics, via Chris Blattman's blog. We presented some new results using data from Chicago in the 1920s (generously shared by Kevin Corder) here.

"Rebels without a Cause: The Use of Coercion and Children in Guerrilla Warfare" (with Chris Blattman). We argue that adolescent soldiers are targeted for recruitment because younger, less productive children contribute suboptimally to rebel activities while older, highly productive individuals are difficult to retain. An extended version of the formal model only is available here. The paper uses data from the Survey of War Affected Youth (SWAY).

Selected Presentations

International Studies Association, New York, February 2009
Presentation, "International Mediation of Wars: A New Dataset and Instrumental-Variable Approach"

American Political Science Association, Boston, August 2008
Presentation, "What the Numbers Say: A Digit-Based Test for Election Fraud Using New Data from Nigeria"

American Political Science Association, Boston, August 2008
Presentation, "Rebels without a Cause: The Use of Coercion and Children in Guerrilla Warfare"

Society for Political Methodology, Ann Arbor, July 2008
Poster, "Divide and Rule: Improving the Use of Ratio Variables"

Conference on Participation in Civil War, Stanford University, June 2008
Presentation, "Forcible Recruitment in Northern Uganda"

International Politics Seminar (CUIPS), Columbia University, May 2008
Presentation, "International Mediation of Militarized Conflicts: When is it Initiated and When is it Effective?"

Society for Political Methodology, State College, July 2007
Poster, "Identifying Manipulation in Electoral Results: A Novel Test and New Data from Nigeria"