ESRA 2027 sessions by theme
Back to Overview of Sessions
Linking Survey and Administrative Data to Investigate and Reduce Total Survey Error: Benefits, Challenges, and Opportunities |
| Coordinator 1 | Mr Paul Cochet (INED) |
| Coordinator 2 | Mrs Claire Darriagde (INED) |
| Coordinator 3 | Mr Julio Ricardo Davalos (INED) |
| Coordinator 4 | Dr Tom Emery (Erasmus University Rotterdam) |
| Coordinator 5 | Dr P. Linh Nguyen (INED, London School of Economics and Political Science) |
| Coordinator 6 | Professor Joe Sakshaug (IAB & LMU Munich) |
Session Details
Survey methodologists and topic researchers alike have long been interested in embracing the new opportunities resulting from linking survey and administrative data. With many European statistical institutes committed to provide increasing (secure) access to different types of administrative data (ranging from population registers; income, pension, and taxation data; welfare/benefit records; health and healthcare records; employment and labour-market data; education records to environmental and geographic data). This session asks the overarching question of how linking administrative and survey data can help understanding, measuring, and reducing different sources of total survey error?
For example, how can linking survey to administrative data assess or reduce coverage error to increase the inclusiveness in the data collection process; produce more efficient sampling strategies; identify mechanisms of nonresponse; tackle measurement error by validating survey response against external records; and/or improve weighting and calibration procedures?
While administrative data offer opportunities to investigate and reduce some components of the total survey error, the process of integrating two data sources also introduces new methodological, practical, or ethical challenges. Different types of administrative data come with their specific errors that might be exacerbated when linking to survey data. Therefore, we specifically welcome contributions evaluating linkage error and/or formulate linkage quality (measurement) standards, maybe even specific to the type of administrative data used. In addition, we encourage contributions from different substantive and/or interdisciplinary perspectives and from both national or cross-country contexts.
By bringing together these perspectives, the session aims to advance understanding of when and how linked data can improve survey quality, while identifying the limitations and risks associated with their use.
