Date created: | Last Updated:

Category: Uncategorized

Description: This study will examine how bureaucratic responsiveness varies across issue sensitivity, citizen identity, and citizen signaling strategies in China’s state-run legal-aid system. I will conduct a two-stage national audit experiment on the Ministry of Justice’s 12348 legal-aid hotlines. In Stage 1, I will randomly assign simulated callers to pose either politically sensitive or non-sensitive legal questions, and to vary caller attributes such as gender, ethnicity, education, and legal experience. This design will allow me to test whether politically sensitive inquiries and marginalized identities reduce substantive responsiveness and shift bureaucratic behavior toward more performative engagement. In Stage 2, within the subset of politically sensitive inquiries, I will randomly assign one of four scripted “public transcript” framings: a collective-action threat, a bureaucratic-oversight cue, a legal-norms appeal, or a social-recognition incentive. These framings will map onto a 2×2 general model of citizen strategies (positive vs. negative incentives; formal vs. informal institutionalization) and will allow me to test which signals most effectively increase responsiveness. The primary outcomes will be composite indices of substantive responsiveness (accuracy of legal advice and absence of deterrence) and performative responsiveness (empathy, friendliness, and patience), measured via a fixed human- and machine-coding protocol from recorded and transcribed calls. Secondary outcomes will include each component score separately. All outcome definitions, coding procedures, and model specifications will be pre-committed before the first call is placed. Expected contributions include: Providing causal evidence on discrimination and selective responsiveness in authoritarian bureaucracies. Identifying which citizen strategies—legalistic, reputational, oversight-based, or mobilizational—are most effective in shaping bureaucratic engagement under political sensitivity. Offering methodological innovations in combining human and AI-based coding for large-scale audit experiments in non-democratic contexts. The findings will have implications for theories of bureaucratic incentives, state-society relations, and the limits of legal mobilization in authoritarian regimes.