Statistical Matching of Administrative and Survey Data: An Application to Wealth Inequality Analyses

• Activity Number: 115
• Type: Topic Contributed
• Date/Time: Monday, August 1, 2011 : 8:30 AM to 10:20 AM
• Sponsor: Section on Survey Research Methods
• Author(s): Anika Rasner*+ and Frick R. Joachim and Markus M. Grabka
• Companies: German Institute for Economic Research (DIW Berlin) and German Institute for Economic Research (DIW Berlin) and German Institute for Economic Research (DIW Berlin)
• Keywords: statistical matching ; imputation ; inequality ; social security wealth

Abstract

Using population representative survey data from the German Socio-Economic Panel and administrative pension records from the Statutory Pension Insurance, the authors compare three imputation techniques (hotdeck, regression-based, univariate imputation sampling) and Mahalanobis distance matching to complement survey information on net worth with social security wealth information from the administrative records. The unique properties of the linked data allow for a direct control of the quality of matches under each technique. Based on three evaluation criteria (correlation coefficient, average distance, kernel density plots) comparing the observed and matched benefit, we identify Mahalanobis distance matching to perform best. Exploiting the advantages of the newly assembled data, we include social security wealth in a wealth inequality analysis. Despite its quantitative relevance in Germany, social security wealth has been thus far omitted because of the lack of adequate micro data. The inclusion of social security wealth doubles the level of net worth and decreases inequality by almost 25 percent. Moreover, the results reveal striking differences along occupational lines.