This paper is part of a larger project that addresses a set of fundamental issues linking security and development: Does foreign assistance affect resilience to intrastate armed conflict—and if so, where, when, and how? To evaluate the association between development aid and the likelihood, escalation, severity, spread, duration, and recurrence of violence—spanning the phases before, during, and after conflict—we develop an agent-based computational model. Disaggregated empirical data on conflict, aid and other covariates is used to seed, optimize, and validate the model in an effort to identify parameter constellations that best represent empirical outcomes during various phases of violence. In particular, we examine a range of alternative causal mechanisms that link the effect of foreign assistance on conflict—increase, no effect, decrease—to variations in the host environment, including the fine-grained micro-characteristics of population centers, the relevant actors and their behavior. The model is then used to probe “alternative futures” that explore the impact of the size, timing and location of aid provisions in F/CA countries.
