A century after Max Weber famously posited that the state is defined by holding a monopoly on the legitimate use of force within a territory, it is abundantly clear that states rarely hold such a monopoly, and they may in fact willingly cede it, both in wartime and in peacetime. Yelena Biberman’s deeply researched book Gambling with Violence takes up the question of when and why, exactly, states engaged in civil wars decide to work with preexisting armed groups, organize civilian militias, or set up squads of ex-insurgents to secure areas or carry out attacks, “outsourcing” violence instead of monopolizing it.

This outsourcing is a gamble: for states, alliances with armed groups may be a force multiplier, provide local intelligence and legitimacy benefits, or offer insulation from accountability for human rights abuses. Yet outsourcing can also backfire: armed groups may turn defect, use state resources for their own ends, or commit counterproductive violence against civilians.

Biberman moves beyond state-centric perspectives, arguing that state–armed group relationships in civil war depend on the “balance-of-interests” between two potential allies (pp. 24–30). When a state has the upper hand in a region, it has little incentive to risk allying with an armed group, and so Biberman predicts that state