Surveying Residential Burglaries: A Case Study of Local Crime Measurement

TL;DR

This paper combines police reports and victimization survey data to estimate residential burglary rates in North Carolina's largest cities, highlighting challenges in comparing crime data over time and across locations.

Contribution

It introduces a method for integrating police and survey data to improve local burglary rate estimates and discusses the limitations of such comparisons.

Findings

Police and survey data integration improves burglary estimates.

Between-city and over-time comparisons are fragile and uncertain.

Method highlights data limitations in crime rate analysis.

Abstract

We consider the problem of estimating the incidence of residential burglaries that occur over a well-defined period of time within the 10 most populous cities in North Carolina. Our analysis typifies some of the general issues that arise in estimating and comparing local crime rates over time and for different cities. Typically, the only information we have about crime incidence within any particular city is what that city's police department tells us, and the police can only count and describe the crimes that come to their attention. To address this, our study combines information from police-based residential burglary counts and the National Crime Victimization Survey to obtain interval estimates of residential burglary incidence at the local level. We use those estimates as a basis for commenting on the fragility of between-city and over-time comparisons that often appear in both public discourse about crime patterns.