Economic sanctions are among the most frequently deployed instruments of statecraft, yet fundamental questions about their politics and effects remain unresolved. This dissertation argues that progress requires solving a prior methodological problem: existing research studies only sanctioned states, obscuring both why some violations are punished and others are not, and whether sanctions actually change target behavior. I address these challenges by introducing the Sanctionable Behavior Dataset (SBD), which tracks behaviors that recurrently trigger sanctions and records, for each state-year between 1989 and 2015, whether a given state engaged in that behavior and whether it was sanctioned by the United Nations, United States, or European Union. I deploy this framework to advance two substantive theories. First, I show that states heavily import-reliant on a dominant partner systematically escape its sanctions. Contrary to conventional wisdom, economic vulnerability reduces likelihood of coercion because hierarchical lock-in makes enforcement costly. Second, I advance research on credible commitment, showing that sanctions compel behavioral change only when a sender has both reputation for resolve and enforcement history specific to the targeted behavior. Together, these findings reframe the politics of sanctions enforcement and provide answers to long-standing puzzles in the literature on initiation and effectiveness of sanctions.