What is the role of political behavior in shaping the dynamics of conflict? World Politics models of civil conflict often overlook political behavior measures and mechanisms. In this work, I draw on American Politics research on domestic politics, trust and information sharing, and political psychology to bridge existing theoretical gaps in World Politics. I do this in three papers: one on audience costs in the U.S. Congress, one on information-sharing during counterinsurgency, and one on a theory of non-material insurgent support through social goods. In the first, ``Where are the Costs?", theories of audience costs assume that costs in democratic societies are distributed across a wide support coalition in the public. I measure costs among a powerful group well-equipped to punish the president, the American Congress, and find evidence of punishment through negative messaging in getting ``on the record" in public hearings, and some evidence of changes in roll-call voting behavior. In the second, ``Troops or Tanks?", Ryan Van Wie and I address an existing counterinsurgency literature that draws a stark binary between dismounted troops that interface with civilians to build trust, and those that stay in vehicles to protect themselves. We argue that this is a dynamic trade-off, one that changes in relation to the severity of violence, the quality of non-human sources of intelligence, and the force employment strategy of a military unit. In the third, ``Winning Legitimacy", I address a gap in existing literature, explaining how insurgencies can retain a resilient base of support when their capacity for violence and material aid is overpowered by a counterinsurgent. I argue that basic psychological needs for group identity and self-esteem power a lasting urge for political association with an insurgency. Insurgencies are often organic growths designed to fill a vacuum in social institutions, and often share more in common with a local population than the counterinsurgent, giving them an asymmetric advantage in winning hearts and minds. I bring in new literature, data, and case studies to enrich and resolve current impasses in this work. In ``Where are the Costs?", I introduce a line-by-line sentiment analysis of the speech of members of Congress in 52 Congressional hearings. In ``Troops or Tanks?", we employ a mixed-methods approach, creating a new dataset of unit-level mechanization at district-week granularity in Operation Iraqi Freedom, contextualizing our results with two unit-based case studies in Ramadi and Basra. In ``Winning Legitimacy", two case studies on British Palestine and Dhofar support a nuanced theory of the political choices faced by individual civilians, and the effect of that behavior on insurgent conflict outcomes. This has implications for threat making, fighting, and goal setting in armed interventions. In the American system, long-run presidential punishment may manifest as a tarnished historical legacy and lasting media tropes. During counterinsurgency, the capacity of non-human intelligence and the dynamic use of protective equipment matters more to resolving insurgent violence than a stark choice to dismount or protect troops. In civil conflict, insurgencies are resilient political organizations that provide value through identity, association, and worldview that outlasts their capacity to pay or inflict violence.