Opportunities and Challenges of Multiple Administrative Data at the Intersection of Education, Labour Market, and Pension Research

The advantages of administrative data for social research are obvious: Administrative data are mostly a representative selection limited by administrative tasks but with high numbers of cases. Since losses of answers and failures in one’s memory are impossible, the data quality is quite high – objectivity and reliability is granted. However, systemic missing data in the registration of key variables for social research limit data quality. The limitation is caused by the fact that there is neither a need nor a benefit from correctly reporting or updating this information towards the administrative bodies. Furthermore, the lack of these data on e.g. qualification makes complex procedures of imputations necessary.

By referring to my studies at the intersection of education, labour market, and pension research, I present some of the benefits and challenges associated with the use of multiple administrative data and possible solutions.

To analyse the individual outcome of an increasing educational participation in Germany, I observed four cohorts in the pre- and post-educational era during career start. The indicator for the educational outcome was the entitlements for later pension income. This is innovative and appropriate since pension entitlements in a contribution-based system highly depend on lifetime earnings.

For my multivariate modelling, I used a unique data set called BASiD (biographical data of selected social insurance agencies in Germany), provided by the German National Pension Insurance and the Federal Employment Agency. The linked micro data set includes employment biographies of 568.468 persons with day-to-day information e.g. on pension entitlements. Additionally, firm’s data such as wage level and number of employees are included.

My contributions aims to discuss the applied frame of analysis of BASiD with regard to their opportunities and limitations for research in social inequality.