Social identity, prior exposure, and hiring outcomes in police recruitment: A machine learning analysis

Police recruitment is commonly framed as a process of identifying individuals who possess the traits and competencies associated with effective police work. Less attention has been paid to how differential exposure to policing, organisational familiarity, and anticipated identification with the occupation shape progression through recruitment systems. Drawing on data from a large recruitment campaign within an English police constabulary (n = 886), this study employs a machine learning model to examine how demographic characteristics, prior policing exposure, social ties, and perceived identification with policing combine to predict appointment outcomes. The analysis shows that hiring decisions are most strongly associated with sustained pre-entry exposure to policing roles, relational proximity to the organisation, and perceived identification with policing, while demographic variables such as age, gender, and ethnicity exhibit comparatively weaker predictive influence. These findings suggest that recruitment operates as a socially mediated process in which familiarity with organisational norms and expectations reduces friction during selection and increases the likelihood of appointment. Interpreted through the Social Identity Approach, the findings indicate that recruitment outcomes reflect uneven access to preparatory experiences and identity-relevant resources accumulated over time, rather than isolated assessments of merit at the point of application. The study contributes to socio-legal debates on institutional access and organisational reproduction by demonstrating how recruitment systems may indirectly privilege candidates with prior social and cultural proximity to policing.