Evaluating the causal mechanisms to police legitimacy and compliance in the United States: An in-depth mixed-methods approach

Procedural justice theory states that when police treat people in a fair, respectful, and neutral manner, individuals are more likely to perceive the police as legitimate and obey the law. To test this perspective, researchers often use experimental vignettes that depict police-citizen interactions and measure subsequent attitudes. However, it is not straightforward to determine causal effects from these designs, as one must carefully assess whether the effect runs through the intended theoretical pathways in order to properly test the theory. This paper advances beyond ‘traditional’ experimental designs to evaluate the pathways linking police-citizen interactions, attitudes, and behavioral intentions. The data consist of a representative sample of 2,003 adults residing in the United States using an online panel. We use a mixed methods approach combining a pre-registered 3x2x2 experimental vignette depicting a traffic stop by police with follow-up closed- and open-ended questions, allowing respondents to explain their reasoning. We found that the quality of treatment by police was statistically related to subsequent perceptions of police legitimacy, but not compliance. However, follow-up placebo tests and open-text responses revealed a number of confounding factors limiting the strength of causal claims regarding procedural justice theory. Most notably, a number of respondents expressed that they would comply with the officer in the moment, even if they disagree, out of fear of escalation and their own personal safety. Research using (quasi-)experimental designs should evaluate causal assumptions and mechanisms more rigorously when theory testing, and likewise incorporate methods that can better assess these intermediate processes, such as open text responses and process tracing.