77 – Using Multiple Sources of Data
Re-Contact Within the Justice System: Integrating Multiple Data Sources Through Record Linkage
Alexander Reicker
Statistics Canada
The size and scope of linkage projects are increasing as probabilistic record linkage becomes a standard method to integrate administrative and survey files based on personal identifiers. Thanks to recent improvements in computers and software, Statistics Canada is currently undertaking multiple, large record linkage projects, including the Justice Re-contact Project, which follows pathways through the justice system and measures the time between police contacts. In this paper, we give an overview of the Re-contact project which involves a series of internal and external record linkages using data from police services, provincial courts, and provincial and federal correctional services. The files are linked probabilistically using the G-Link software package developed at Statistics Canada. Our results highlight the difficulties in integrating many administrative datasets in one common environment due to the variations in the data availability and data quality, as well as the technical challenges of managing a system where the number of records grows rapidly as new datasets are added.