Bayesian Estimation of State-Level Crime Victimization Rates

This third-party report describes the feasibility of using model-based techniques to produce estimates of state-level crime victimization rates. Specifically, a Bayesian multivariate lognormal model was applied. Direct estimates from the National Crime Victimization Survey served as model response variables. Covariate information was derived from the FBI’s Uniform Crime Reporting Program. The model incorporates multiple crime types and data for two aggregate time periods. This report presents results from applying the model to produce state-level estimates of crime victimization rates.