The dissertation addresses a number of core questions about terrorism: why do some terrorist organizations, and not others, adopt the devastating tactic of suicide bombing? What contextual factors predict self-starter terrorism, meaning political violence that occurs without the assistance or direction of a pre-existing terrorist organization (e.g. the Boston Marathon bombers)? How are one's social networks affected by one's position in an affiliation ecology (McPherson 1983), and how do positions in affiliation ecologies affect recruitment into terrorist organizations? To address the first question, I apply event history models to a global dataset that includes all recorded suicide bombing attacks from 19812006. I show that organizations embedded in collectivist cultural ecologies are far more likely to use suicide bombings than organizations embedded in individualist cultural ecologies, conditioning on factors suggested by alternative explanations. To assess selfstarter terrorism, I develop an agent-based model, grounded in empirical data that identifies the mechanisms by which the network ecology promotes and constrains terrorist mobilization. The model is also able to distinguish the manner in which self-starter terrorism is carried out - whether by lone wolves or by small groups. To assess the relationship between affiliation ecologies and terrorist recruitment, I first generate a theoretical framework I term "Blau Status Analysis", which extends McPherson's (1983) approach to individual-level actors. Using the Add Health dataset, I validate the framework by showing that Blau statuses are associated with a variety of network properties. I then apply the framework to show how affiliation ecology affects recruitment into covert organizations. I conclude by discussing additional research veins in political violence that can be exploited using the ecological perspective as well as the theoretical and methodological contributions the dissertation made to questions of general interest to sociology.