Controlling Peripheral Regions with State-Owned Enterprises

How do states extend control over peripheral territories afflicted with armed rebellions? Conventional armies tend to be militarily superior to their domestic challengers but cannot easily collect local information on where the insurgents hide. This paper theorizes and provides empirical evidence for a unique institutional solution that the Chinese state has used to control its peripheral provinces – state-owned enterprises (SOEs) with the hybrid function of military training and land reclamation that are colloquially referred to as the “bingtuan.” Specifically, military training prepared bingtuan employees with the coercive capacity necessary for repressing rebels in an armed confrontation; reclamation of wastelands allowed them to supply their own food for permanent resettlement without intensifying their relationship with the indigenes, which was key to developing intelligence about the local communities. Using a difference-in-differences design and an original dataset based on over 500 historical gazetteers, I show that China increased bingtuan employment by 50% when its control over the Xinjiang province was threatened in 1962. In addition, this employment expansion did not appear to be driven by conventional rationales for SOEs such as the occupation of strategically important sectors and the provision of employment to participants of instability. By elucidating the institutional foundations for China’s strategy to consolidate its historically restive frontiers, the paper provides important insights to a long-standing puzzle for the studies of civil war onset – sustained military quiescence of Xinjiang indigenes for over four decades under communist rule.