It's Official: Predictors of Self-Reported vs. Officially Recorded Arrests

Abstract

PurposeThe study of the distribution and correlates of arrest is widely recognized as an important topic, for the purposes of contributing to changes in police policy and training, which in turn increase the fairness of U.S. policing. Despite agreement that this area of research is an important one, there remains variation in the way arrest is measured. The current study compares two common measurements of arrest, official records and self-reports, for National Youth Survey Family Study (NYSFS) respondents across four time periods. MethodsThe sample was divided by those who reported severe offending and those who did not. Crosstabs, correlation coefficients and logistic regression models were run, to examine the extent to which self-reported and officially recorded arrests are related, and whether there are commonalities in the predictors of self-reported and officially recorded arrests. ResultsWhile the agreement between the two measurements of arrest is over 80%, the majority of that agreement is comprised of respondents who were not arrested. ConclusionsOverall, there were more instances of a self-reported arrest but no official arrest, than the reverse. There does not appear to be a pattern in frequencies or correlation coefficients based on the severity of reported offending.