UvA

The study of police corruption faces seemingly insurmountable sampling and validity problems. This paper outlines an experimental technique for overcoming these issues by challenging the assumption that corruption is not prone to systematic observational research. Incident reporting combines a ‘combing’ procedure for sampling observable incidents of corruption in public spaces with a benchmarking procedure for coding observations of misconduct. After introducing the new forms of analysis that this technique makes feasible the paper reviews unique values and limitations of incident reporting. The main contention is that the technique constitutes a valuable tool for improving the study of everyday police corruption.