BJS currently collects annual aggregate counts of jail populations through the Annual Survey of Jails (ASJ) (a sample of 891 jurisdictions in 2013), as well as through the Death in Custody Reporting Program summary form (2,872 jurisdictions in 2013) and Jail Census (2,872 jurisdictions in 2013). While these collections can provide a one-day snapshot of the population, as well as total counts of admissions and releases, they cannot provide more detailed descriptions of the population, such as sex by race distributions, age or booking offense distributions, moments of distribution for length of stay, or recidivism estimates.
The new Local Jail Reporting Program (LJRP) is aimed to fill this gap through a cooperative agreement with Appriss to obtain individual-level jail inmate administrative data annually or semi-annually. Data collected under the LJRP will include law enforcement operational variables (e.g., arrest and booking timestamps, charge codes, release reasons, and bond amount), sociodemographic characteristics (e.g., race, sex, marital status, and place of birth), physical attributes (e.g., weight, height, and photo), and personal identification information (names, social security numbers, state and federal inmate ID numbers, date of birth, and address of residence). Currently, Appriss can provide data from 69% (2,036) of the jail facilities in the country. Pending a successful pilot study of the data by BJS, Appriss will work with its data providers and jail jurisdictions to expand data coverage and reduce missing data on key variables, including state and national fingerprint identifiers.
To assess the feasibility and benefits of the LJRP, BJS will conduct a pilot study to evaluate the quality of Appriss jail booking data. For this pilot study, Appriss will provide a data set consisting of all individual-level historical and current booking records from three statesâCalifornia, Texas, and Georgiaâthrough the end of 2014. The data set will contain all available variables, including personal identifiers and photos. BJS will assess data quality in following ways:
- assess data completeness;
- check data validity and reliability;
- compare jurisdiction-level inmate counts, admissions and releases to statistics from BJS collections;
- and link jail booking records to BJS administrative datasets to check the quality of personal identifiers.