@article{NI2026, 
author = {Xing NI and Luye ZHAO},
title = {Punishment and Compliance: A Conjoint Experiment Based on Red Light Running Behaviors},
year = {2026},
journal = {China Public Administration Review},
volume = {8},
number = {2},
pages = {109-146},
keywords = {punishment, compliance, perceived legitimacy, conjoint experiment},
url = {https://www.sciopen.com/article/10.26599/CPAR.2026.9680206},
doi = {10.26599/CPAR.2026.9680206},
abstract = {Punishment serves as a crucial mechanism for maintaining political authority and the effectiveness of governance. In modern governance, punitive power has become increasingly institutionalized and routinized, embedded in standardized regulatory frameworks and everyday enforcement practices. Yet a growing body of research suggests that coercive, top-down sanctions do not automatically translate into higher levels of citizen compliance and may even generate legitimacy crises and unintended counterproductive effects. Understanding how punishment operates, through which mechanisms, and under what conditions it shapes compliance therefore remains a central theoretical and empirical challenge.This study investigates the impact of various forms of punishment on citizens' compliance, along with the underlying mechanisms and heterogeneous effects. Different perspectives of instrumentalist deterrence theory, normative legitimacy perspectives, and recent behavioral insights are integrated. Empirically, the research employs a conjoint experiment based on the context of red light running behaviors, to systematically vary punishment certainty, severity of material punishment (monetary fines), severity of spiritual punishment (educational and shaming-based sanctions), group factors (the probability of others' non-compliance), and natural factors (the likelihood of a safety accident by non-compliant behaviors). This design enables causal identification of the effects of distinct punishment attributes while reducing social desirability bias and approximating real-world decision environments.Based on a nationally distributed sample of 1000 respondents and 6000 experimental observations, the findings reveal that, first, the enhancement of the probability of punishment, the implementation of material penalties, and some spiritual punishment methods can promote citizens' compliance, but the relationships between punishment certainty, severity, and compliance are non-linear. Beyond a certain threshold, harsher sanctions do not yield proportionally stronger compliance effects.Second, perceived legitimacy plays a mediating role in this process. When individuals regard punishment as effective, morally justified, and fair, their compliance rate increases. Excessively harsh or stigmatizing penalties may undermine legitimacy and weaken compliance.Third, punitive effects are heterogeneous and context-dependent. The high level of possibility of red light violations by others will negatively moderates the deterrent impact of fines. Individual characteristics and situational risk perceptions further condition responses to punishment, where cognitive biases, personal experiences, and other demographic characteristics contribute to heterogeneous deterrent outcomes.This paper argues that it is necessary to balance the effectiveness and legitimacy of punishment. Such a balance has important implications for achieving social stability, optimizing state-society relations, and advancing the modernization of the national governance system and governance capacity.}
}