Research -- Conflict
How do counterinsurgency policies affect rebel and civilian behavior? How do transnational dimensions of conflict impact the micro-dynamics of civil wars? My research in this area explores the consequences of investments in state capacity, the efficacy of combatants' efforts to build civilian support, and the resilience of militant groups in the face of repression and coercion.
Refugee returnees in Afghanistan (Source: UN)
Militant Alliances
Militant groups frequently rely on relationships with other violent, non-state actors. The density and durability of transnational militant networks has grown over time. A central contribution of my research on conflict is an original dataset, the Militant Group Alliances and Relationships (MGAR) dataset, which documents 7,409 dyad-year relationships between 2,613 militant groups around the world from 1950 to 2016. I introduce and study these data in papers with Erica Chenoweth, Michael C. Horowitz, Evan Perkoski, and Philip B.K. Potter (IO 2022), Michael C. Horowitz and Philip B.K. Potter (JOP 2022), and Philip B.K. Potter (JOGSS 2023).
How do militant groups sustain cooperation under the shadow of repression? Using original data from MGAR, we study descriptive trends in militant cooperation, and develop and test a theory of alliance durability (IO 2022). State violence represents a severe deterrent to cooperation between violent, non-state actors. Measures organizations take to monitor and enforce inter-group bargains and agreements expose them to greater security risks. Understanding cooperation between militant groups thus requires explaining how groups commit themselves in the face of repression. One established answer is that state sponsors can enforce cooperation between their clients. But shared sponsorship is far rarer than militant cooperation, suggesting other commitment devices must exist. We argue and find evidence that shared ideology, and particularly shared religion, helps militant organizations commit to cooperation under repression. Ideology facilitates commitment by lengthening the shadow of the future, leveraging nonmilitant co-ideologues to detect and sanction defection, providing access to common authority and social structures that reduce uncertainty, and enhancing mutual trust. The value of shared ideology is particularly important for the durability of material alliances, in which groups exchange key materiel. Degrading cooperation between militants may be essential to reduce their capacity. But our findings suggest that increasing repression is not the most effective way to disrupt cooperation, at least when groups share an ideology. Rather paradoxically, deterring and weakening cooperation between violent nonstate groups will require governments to engage in nonviolent efforts to build ties with potential sympathizers and authority figures of the ideologies to which militant groups adhere.
Does leadership decapitation affect relationships between targeted groups and their allies? Cooperation between militant groups is central to the threat these groups pose. Through alliances, militant organizations can bolster their capabilities, and enhance their lethality, reach, and resilience. But comparatively little is known about how governments attempt to degrade militant networks. We examine whether leadership targeting strategies undermine cooperation between violent, non-state actors (JOP 2022). Using original data from MGAR, we demonstrate an association between leadership decapitation and alliance termination. This association is strongest when a targeted group’s founder is killed. To support a causal interpretation, we show that unexpected leadership turnover through leader deaths by natural causes have similar effects. Statistical tests suggest that decapitation can trigger alliance termination by inducing inter-organizational splits even when targeted groups are not destroyed outright. We use process tracing in more than 70 cases to understand the mechanisms underpinning our quantitative results. We identify four specific mechanisms through which leadership targeting can trigger militant alliance termination: (1) weakening or collapsing targeted groups, making cooperation impossible; (2) raising fears about operational security, making cooperation costlier; (3) eliminating personal ties between leaders, making cooperation more difficult to sustain; and (4) inducing command-and-control problems, which drive preference divergence between targeted groups and allies over strategy and tactics, making cooperation less appealing.
What explains the creation of large, transnational alliance networks by major militant groups like al Qaeda and the Islamic State? Using original data from MGAR, we study patterns of cooperation within the alliance networks that emerged around al Qaeda and the Islamic State. Using data-driven case studies, we develop and test a theory of large alliance networks (JOGSS 2023). We demonstrate that militant groups can leverage large alliance networks to bolster their ideological and operational reputations. Organizations can draw on operational capabilities and successes to build international networks that bolster their ideological credibility. Conversely, organizations with reputations for ideological authority can lend it to affiliates, who offer reach into active conflicts, bolstering claims to operational capacity. This logic of comparative advantage suggests that militant alliances can be a strategic response to underlying material or ideological deficits.
Refugees and Conflict
How does refugee return affect conflict in origin communities? Answering this question is difficult because when and where refugee returnees move is directly linked to security conditions in prospective destination communities. A number of recent policies have aimed at facilitating repatriation by providing returnees with cash transfers to ease reintegration. In a paper with Austin Wright (R&R at AER), we estimate the causal effect of a large cash assistance program for refugee returnees on conflict in Afghanistan. The program led to a significant increase in repatriation. Leveraging historical returnee settlement patterns and previously unreleased combat records, we find that policy-induced refugee return had cross-cutting effects, reducing insurgent violence, but increasing social conflict. The program’s cash benefits were substantial and may have raised reservation wages in communities where returnees repatriated. Consistent with this hypothesis, returnee encashment had heterogeneous effects on insurgent violence, decreasing use of labor-intensive combat, while also reducing the effectiveness of counterinsurgent bomb neutralization missions. Additionally, kinship ties and access to informal institutions for dispute resolution significantly offset the risks of refugee return for communal violence. These results highlight unintended consequences of repatriation aid and clarify the conditions under which refugee return affects conflict.
Counterinsurgency Policies
Can state forces use reparations to mitigate civilian backlash to harm perpetrated during counterinsurgency operations? Counterinsurgents frequently rely on mass arrests to impede rebel operations. In so doing, they risk detaining innocent civilians. Wrongful detention can backfire, fueling insurgent violence by alienating detainees and their kin. I examine whether counterinsurgents can mitigate wrongful detention through targeted compensation (JCR 2022). Using declassified micro-data on US payments to individuals deemed innocent and released from Coalition custody in Iraq between 2004 and 2008, I find that detainee release payments are robustly negatively associated with insurgent violence. The violence-reducing effects of detainee release payments were greatest in mixed and Sunni areas; for the types of insurgent attacks most prone to civilian informing; and when detainee release was complemented by other population-centric reforms to detention. These results suggest that post-harm mitigation helps shift civilian perceptions, inducing civilians to share more information with counterinsurgent forces.
Militant groups frequently rely on relationships with other violent, non-state actors. The density and durability of transnational militant networks has grown over time. A central contribution of my research on conflict is an original dataset, the Militant Group Alliances and Relationships (MGAR) dataset, which documents 7,409 dyad-year relationships between 2,613 militant groups around the world from 1950 to 2016. I introduce and study these data in papers with Erica Chenoweth, Michael C. Horowitz, Evan Perkoski, and Philip B.K. Potter (IO 2022), Michael C. Horowitz and Philip B.K. Potter (JOP 2022), and Philip B.K. Potter (JOGSS 2023).
How do militant groups sustain cooperation under the shadow of repression? Using original data from MGAR, we study descriptive trends in militant cooperation, and develop and test a theory of alliance durability (IO 2022). State violence represents a severe deterrent to cooperation between violent, non-state actors. Measures organizations take to monitor and enforce inter-group bargains and agreements expose them to greater security risks. Understanding cooperation between militant groups thus requires explaining how groups commit themselves in the face of repression. One established answer is that state sponsors can enforce cooperation between their clients. But shared sponsorship is far rarer than militant cooperation, suggesting other commitment devices must exist. We argue and find evidence that shared ideology, and particularly shared religion, helps militant organizations commit to cooperation under repression. Ideology facilitates commitment by lengthening the shadow of the future, leveraging nonmilitant co-ideologues to detect and sanction defection, providing access to common authority and social structures that reduce uncertainty, and enhancing mutual trust. The value of shared ideology is particularly important for the durability of material alliances, in which groups exchange key materiel. Degrading cooperation between militants may be essential to reduce their capacity. But our findings suggest that increasing repression is not the most effective way to disrupt cooperation, at least when groups share an ideology. Rather paradoxically, deterring and weakening cooperation between violent nonstate groups will require governments to engage in nonviolent efforts to build ties with potential sympathizers and authority figures of the ideologies to which militant groups adhere.
Does leadership decapitation affect relationships between targeted groups and their allies? Cooperation between militant groups is central to the threat these groups pose. Through alliances, militant organizations can bolster their capabilities, and enhance their lethality, reach, and resilience. But comparatively little is known about how governments attempt to degrade militant networks. We examine whether leadership targeting strategies undermine cooperation between violent, non-state actors (JOP 2022). Using original data from MGAR, we demonstrate an association between leadership decapitation and alliance termination. This association is strongest when a targeted group’s founder is killed. To support a causal interpretation, we show that unexpected leadership turnover through leader deaths by natural causes have similar effects. Statistical tests suggest that decapitation can trigger alliance termination by inducing inter-organizational splits even when targeted groups are not destroyed outright. We use process tracing in more than 70 cases to understand the mechanisms underpinning our quantitative results. We identify four specific mechanisms through which leadership targeting can trigger militant alliance termination: (1) weakening or collapsing targeted groups, making cooperation impossible; (2) raising fears about operational security, making cooperation costlier; (3) eliminating personal ties between leaders, making cooperation more difficult to sustain; and (4) inducing command-and-control problems, which drive preference divergence between targeted groups and allies over strategy and tactics, making cooperation less appealing.
What explains the creation of large, transnational alliance networks by major militant groups like al Qaeda and the Islamic State? Using original data from MGAR, we study patterns of cooperation within the alliance networks that emerged around al Qaeda and the Islamic State. Using data-driven case studies, we develop and test a theory of large alliance networks (JOGSS 2023). We demonstrate that militant groups can leverage large alliance networks to bolster their ideological and operational reputations. Organizations can draw on operational capabilities and successes to build international networks that bolster their ideological credibility. Conversely, organizations with reputations for ideological authority can lend it to affiliates, who offer reach into active conflicts, bolstering claims to operational capacity. This logic of comparative advantage suggests that militant alliances can be a strategic response to underlying material or ideological deficits.
Refugees and Conflict
How does refugee return affect conflict in origin communities? Answering this question is difficult because when and where refugee returnees move is directly linked to security conditions in prospective destination communities. A number of recent policies have aimed at facilitating repatriation by providing returnees with cash transfers to ease reintegration. In a paper with Austin Wright (R&R at AER), we estimate the causal effect of a large cash assistance program for refugee returnees on conflict in Afghanistan. The program led to a significant increase in repatriation. Leveraging historical returnee settlement patterns and previously unreleased combat records, we find that policy-induced refugee return had cross-cutting effects, reducing insurgent violence, but increasing social conflict. The program’s cash benefits were substantial and may have raised reservation wages in communities where returnees repatriated. Consistent with this hypothesis, returnee encashment had heterogeneous effects on insurgent violence, decreasing use of labor-intensive combat, while also reducing the effectiveness of counterinsurgent bomb neutralization missions. Additionally, kinship ties and access to informal institutions for dispute resolution significantly offset the risks of refugee return for communal violence. These results highlight unintended consequences of repatriation aid and clarify the conditions under which refugee return affects conflict.
Counterinsurgency Policies
Can state forces use reparations to mitigate civilian backlash to harm perpetrated during counterinsurgency operations? Counterinsurgents frequently rely on mass arrests to impede rebel operations. In so doing, they risk detaining innocent civilians. Wrongful detention can backfire, fueling insurgent violence by alienating detainees and their kin. I examine whether counterinsurgents can mitigate wrongful detention through targeted compensation (JCR 2022). Using declassified micro-data on US payments to individuals deemed innocent and released from Coalition custody in Iraq between 2004 and 2008, I find that detainee release payments are robustly negatively associated with insurgent violence. The violence-reducing effects of detainee release payments were greatest in mixed and Sunni areas; for the types of insurgent attacks most prone to civilian informing; and when detainee release was complemented by other population-centric reforms to detention. These results suggest that post-harm mitigation helps shift civilian perceptions, inducing civilians to share more information with counterinsurgent forces.
https://orcid.org/0000-0001-5774-5976