Abstract

"We describe a method to calculate individual pension entitlements (earning points, Entgeltpunkte) based on information available in conventional employment history data. Pension entitlements can be seen as an important driver of retirement decisions and of the labour market attachment of older people. Therefore, the calculation of pension entitlements using labour market history datasets improves the usability of these datasets for a broad range of topics around the decision to retire. In the first part of this report, we use a high-quality administrative biographical dataset with linked information on pension entitlements, covering a sample of almost all employees in Germany (Biographical Data of Selected Social Insurance Agencies in Germany, BASiD). Based on the BASiD, we explain which information is needed for the calculation of earning points and outline potential sources of error in this calculation.<br> In a next step, we implement our method to calculate earning points in a larger administrative dataset with only conventional employment history information (Sample of Integrated Labour Market Biographies, SIAB). We describe our calculations and assess possible sources of error by mimicking the SIAB data structure in the BASiD. The average deviation for annual earning points is only around 1%, and the error for the sum of the earning points is around 7%. Most of these errors can be explained by large observation gaps in individual pension contributions during phases of reduced employment such as parental leave, marginal employment or unemployment. There might be additional small errors for the implementation of public compensation payments during these periods as a result of specific legal rules for certain socio-economic groups and minor rounding errors during all periods. Finally, we calculate the earning points using the SIAB and show descriptive statistics for annual and total pension entitlements for the entire sample and selected subgroups. We conclude that our approach to calculating pension entitlements using conventional employment history data opens various new and important research options at least for employees without large employment gaps." (Author's abstract, IAB-Doku) ((en))