Biberman, Yelena (2020) Gambling with Violence: State Outsourcing of War in Pakistan and India. New York: Oxford University Press. 220 pp.

​ISBN: 978-0-19-092997-8

Åshild Kolås

Peace Research Institute Oslo (PRIO)

Read more about this book:

​This book describes histories of outsourcing of state violence to Indian and Pakistani non-state actors in war. The key question raised is why powerful South Asian security establishments choose to arm and deploy civilian citizens against domestic rebels, without formally incorporating them into the state security apparatus. Cases covered are the War of 1971 in East Pakistan (now Bangladesh), India's war in Kashmir during the 1990s, 'tribal' insurgencies in Pakistan (anti-Taliban lashkar militancy in Federally Administered Tribal Areas), Indian counterinsurgency against Maoists in Chhattisgarh, Turkey's war against Kurdish rebels (1984–99) and Russia's wars in Chechnya (1994–96, 1999–2002). Spanning nearly half a century of violence in highly diverse and complex socio-political settings, what brings these cases together is the author's pointed focus on alliances between state and non-state actors as a questionable form of outsourcing of state violence. Biberman describes this phenomenon as a high-stakes gamble, problematic for long-term security and stability, but understandable as a form of 'balance-of-interest'. In short, the gamble is acceptable to the state when valued territory is at stake and rebels have the upper hand, or the local balance of power is roughly equal. On the civilian side, idealistic activists may be convinced to support the state to achieve long-term valued objectives, even when the state appears to be losing. Opportunists may also support the state, when the local balance of power is roughly equal or the state appears to be winning. Logic, interests and opportunities are good places to start. Nevertheless, at least one important question remains partly unanswered: If these states have such powerful militaries, how does the local balance of power become 'roughly equal'?