Linking administrative data with georeferenced location information
Andreas Schneider* 1
Abstract
A current project of the Institute for Employment Research (IAB) deals with the linkage of social data of the Federal Employment Agency and georeferenced location data. Aspects of this project should be presented at the Conference on Anonymization of Integrated and Georeferenced Data, as they can provide valuable impulses for enhancing administrative data with geo information in economic and social research.
Research with georeferenced data and the stratification of research results organized by geogrids is attracting increasing interest in economic and social research. However, linking social data to georeferenced location information remains critical from a data security point of view. As a research institute of the Federal Employment Agency, the IAB makes extensive use of social data generated during the official processing of clients during their job search, benefit granting and participation in programs, e.g. skill enhancement. Additionally, data on employment contracts is also available to research at the IAB.
It is crucial to protect individual administrative data from deanonymization while simultaneously allow extended research in the meantime. This is because described data provides very detailed information about personal characteristics but individuals often have no opportunity to refuse data collection.
Researchers at the IAB are showing increased interest in working with georeferenced data or in georeferencing existing data. Labor market research improves by georeferenced data, e.g. in order to better delimit regional labor markets or to investigate regional differences in the granting of benefits.
As a consequence, linking social data to geodata is subject to strict legal requirements. Knowing a person's place of residence entails the risk of de-anonymizing their identity, especially if other personal characteristics are also available through the linkage. The fact that research projects are not deliberately aimed at deanonymization is irrelevant in this context.
All of these aspects were discussed in the IAB project mentioned above. The project aims to create a general database combining social data and geodata on a frequent basis at IAB. In addition, we will discuss the IAB's experience in linking administrative data with geo-referenced location information and the challenges that arise in cooperation with those responsible for legal issues. A particular aspect in this context is the question of anonymization by using larger geographical units.
*: Speaker
1: Institut für Arbeitsmarkt- und Berufsforschung