A predictive model for serious police misconduct by variation of the theory of planned behaviour

Purpose – Misconduct and deviance amongst police officers are substantial issues in policing around the world. This study aims to propose a prediction model for serious police misconduct by variation of the theory of planned behaviour. Design/methodology/approach – Using two data sets, one quantitative and one qualitative, provided by an Australian policing agency, a random forest analysis and a qualitative content analysis was performed. Results were used to inform and extend the framework of the theory of planned behaviour. The traditional and extended theory of planned behaviour models were then tested for predictive utility. Findings – Each model demonstrated noteworthy predictive power, however, the extended model performed particularly well. Prior instances of minor misconduct amongst officers appeared important in this rate of prediction, suggesting that remediation of problematic behaviour was a substantial issue amongst misconduct prone officers. Practical implications – It is an important implication for policing agencies that prior misconduct was predictive of further misconduct. A robust complaint investigation and remediation process are pivotal to anticipating, remediating and limiting police misconduct, however, early intervention models should not be viewed as the panacea for police misconduct. Originality/value – This research constitutes the first behavioural model for police misconduct produced in Australia. This research seeks to contribute to the field of behavioural prediction amongst deviant police officers, and offer an alternative methodology for understanding these behaviours.