Semuhi Sinanoglu

Containing Ethnic Conflict: Repression, Cooptation, and Identity Politics

with Ceren Belge, published on Comparative Politics, December 17.

🚨 NEW PUBLICATION ALERT! @cerenbelge and I've been working on this #civilwar paper for a while. So grateful to @Journal_CompPol, and @laiabalcells & @YegorLazarev for their feedback on earlier drafts. A mini 🧵: #polisci #AcademicTwitter #AcademicChatter #PhdLife #phdchat

— Semuhi SinanoÄźlu (@semuhi) December 20, 2021


Why do states target some civilians with collective punishment while coopting others with material goods during an ethnic civil war? This article examines how the Turkish government calibrated its repression and cooptation policies towards the Kurdish population during the counterinsurgency of the 1990s. In contrast to the situational conflict dynamics emphasized by the civil war literature, we explain the distribution of cooptation and repression with the state’s identity policy: government policies were more punitive in areas that displayed strong Kurdish linguistic/political identity, or high tribal concentration, while they were more cooptative where the government had fostered a Sunni-Muslim Kurdish identity. The study is based on a novel dataset that includes information about displacement, tribal concentration, and violent events from archival sources.


  1. Collective victimization isn’t simply an informational problem. Because of the state’s identity policy, some civilians are more vulnerable to coercion and cooptation. Civil wars aren’t just fought over the logistical base for guerillas, but also an identity the government wants to kill. The distribution of rebel violence does not explain the full patterns of government’s cooptation and coercion decisions.
  2. We analyze cooptation and coercion together as a strategy to reshape identity, in line with an emerging scholarship that brings these concepts together analytically.
  3. Although the master cleavage of a conflict might emphasize a particular binary—such as Kurdish versus Turkish—there is almost always a broader set of sub- and supra- identities that can unite or divide the communities on each side of the master cleavage, based on religious, sectarian, racial, linguistic, tribal, and other divisions. Our paper offers a quantitative analysis of the impact of different identity cleavages.

Dependent variables

Distribution of repression and cooptation
the distribution of repression share of tribal population

Main results