Insurgency and Small Wars: Estimation of Unobserved Coalition Structures
Insurgency and guerrilla warfare impose enormous socio-economic costs and often persist for decades. The opacity of such forms of conflict is often an obstacle to effective international humanitarian intervention and development programs. To shed light on the internal organization of otherwise unknown insurgent groups, this paper proposes two methodologies for the detection of unobserved coalitions of militant factions in conflict areas, and studies their main determinants. Our approach is parsimonious and based on daily geocoded incident-level data on insurgent attacks alone. We provide applications to the Afghan conflict during the 2004-2009 period and to Pakistan during the 2008-2011 period, identifying systematically different coalition structures. Further applications are discussed.
The authors would like to thank Eli Berman, Ethan Bueno de Mesquita, James Fearon, Camilo Garcia-Jimeno, Jason Lyall, Carlos Sanchez-Martinez, Jake Shapiro, Drew Shaver, Austin Wright and seminar participants at UCSD, Penn, Chicago Harris, Columbia, Stanford, Berkeley, Rochester, UQAM, Kobe, Tokyo, and Princeton for useful comments and discussion and the researchers at the Princeton University Empirical Studies of Conflict Project for generously sharing their incident data online. Nathan Canen provided excellent research assistance. We are grateful to the Social Science and Humanities Research Council for financial support. The views expressed herein are those of the authors and do not necessarily reflect the views of the National Bureau of Economic Research.
Francesco Trebbi & Eric Weese, 2019. "Insurgency and Small Wars: Estimation of Unobserved Coalition Structures," Econometrica, Econometric Society, vol. 87(2), pages 463-496, March. citation courtesy of