Quality and Quantity: Using Administrative Data for Scientific Purposes in Labor Market Research

• Activity Number: 414
• Type: Contributed
• Date/Time: Tuesday, August 3, 2010 : 2:00 PM to 3:50 PM
• Sponsor: Section on Government Statistics
• Author(s): Patrycja Scioch*+
• Companies: IAB
• Keywords: Administrative Data ; Data Quality ; Imputation Rules ; Education ; Record Linkage

Abstract

The study focuses on the education variable in a widely used German administrative data set, the Integrated Employment Biographies (IEB), which is very important for analyzes (e.g. wage inequalities, employment chances). The quality of this variable became extremely worse over the last 10 years (35% without information). Some studies tried to improve the quality by developing imputation rules to fill the gaps in dependence on existing information. I use additional data (German patent data). Approx. 90% of inventors are in the social-security system and education is identifiable via their titles (prof, dr or diploma). Using record linkage they are linked to the IEB to verify the information. The education is compared to the original and the imputed information. Multivariate models explain the mechanism of the measurement errors of the original and the various imputed education variables.