Why do people fight? This paper synthesizes five decades of conflict theory into a unified causal architecture and introduces a formal game-theoretic model that gives that architecture mathematical precision.
We begin by surveying the canonical theoretical traditions, including Gurr's relative deprivation, Collier and Hoeffler's greed-versus-grievance framework, Fearon's bargaining failure, Hirshleifer's paradox of power, and Fearon and Laitin's state capacity argument. We argue that pitting these explanations against one another in horse races obscures a more important insight: they operate simultaneously on different layers of the same causal system. Conflict emerges when the returns to fighting exceed the returns to producing in an environment where the state is too weak to enforce bargained settlements. Grievance provides the fuel; feasibility determines whether rebellion can be organized; economic structure determines which shocks push toward violence; state capacity determines whether rebellion can be deterred.
We then develop a formal two-player dynamic contest model with a cumulative Tullock contest success function and a Stackelberg decision structure. We characterize optimal fighting effort for both a rebel group and a government and derive the equilibrium conflict date via the envelope theorem. The model formalizes the shadow-of-conflict insight from Garfinkel and Skaperdas: armed peace is an equilibrium, but a fragile one. Five policy-relevant insights follow regarding the timing of military investment, the primacy of effectiveness over raw spending, and the dangers of unilateral disarmament.
Finally, we situate climate stress within this framework. Climate shocks enter the system at every level simultaneously, degrading agricultural productivity, weakening state capacity, and narrowing the bargaining space, making them a uniquely destabilizing class of perturbation.