Handcuffs and Backlash: Arrest, Torture, and Accountability

Does the state responding to a peaceful protest event by arresting protesters increase or decrease the likelihood that future protest events occur frequently? Existing research on the protest–repression nexus yields mixed findings, often treating repression as a binary and neglecting contextual variation. We argue that arrests send distinct signals that can either deter or mobilize dissent depending on political conditions. Using event-level data, we examine how arrests at peaceful protests affect the frequency of subsequent protests. Our analyses show that arrests typically generate backlash, increasing protest activity. However, this effect is conditional. In contexts where torture of detainees is systematic, arrests reduce future protests by making dissent prohibitively costly. By contrast, when governments refrain from torture, arrests amplify mobilization. Electoral accountability further moderates these dynamics: arrests spur protest when citizens lack either the capability or opportunity to hold leaders accountable electorally, but their effect weakens as elections approach. These findings highlight how the meaning of repression is context-dependent and demonstrate that arrests can simultaneously function as deterrents or catalysts, depending on broader institutional environments.