Appendix 1: Data description
The three datasets used are the Report System [on] Further Education (Berichtssystem Weiterbildung, BSW), the German Socioeconomic Panel (GSOEP), and the Qualification and Careers Survey (Qualifikation und Berufsverlauf, IAB-BIBB).
The BSW is relatively unknown compared to the other datasets. The BSW survey was conducted seven times (1979, 1982, 1985, 1988, 1991, 1994, 1997, 2000, and 2003) by the Federal Ministry for Education and Research (Bundesministerium für Bildung und Forschung); data are provided by the Central Archive for Empirical Social Research, University of Cologne. Each survey year, about 7,000 persons between 19 and 64 years are interviewed orally (this includes employed and non-employed people). The BSW dataset is at present the only regular representative survey containing all kinds of training incidences in Germany.Footnote 20 In contrast to the other datasets, questions on training are the focus of this survey. We take the year 1988 as observations before and 1994, 1997, and 2003 as observations after the reform. Questions on job-related training refer to the last 12 months.
The GSOEP is an individual-level dataset with panel structure. It is the largest representative longitudinal study of private households in Germany. The same private households, persons, and families have been surveyed annually since 1984. In this dataset, we have information on whether a person took part in job-related training in the last 3 years. Observations before the reform refer to 1989 and observations after the reform to the years 2000 and 2004. The GSOEP has been conducted since 1984, but questions on job-related training started in 1989 and were only repeated in 1993, 2000, and 2004. We do not use 1993 because in asking for training during the last 3 years, this wave barely covers the 1992 reform.
The IAB-BIBB data are a representative survey of employed persons, which was conducted in 1985, 1991, and 1998. It focuses on job descriptions and detailed information on qualification profiles and occupational development. Each survey wave consists of more than 34,000 observations; questions on job-related training refer to the last 5 years.
Although there are some questions on job-related training in the German Micro Census (Mikrozensus, MZ), this dataset is not suitable for this analysis, because training participation is underrepresented there.Footnote 21 As pointed out by Wohn (2007), there are several reasons why training participation in the MZ is underrepresented compared to the BSW training participation. Since the other two datasets (GSOEP and IAB-BIBB) have comparable training incidences to the BSW, we focus on these three datasets in the regression analyses and use the Micro Census data only for descriptive analyses (see Fig. 2a–c).
The choice of datasets is driven by information on job-related training at the individual level both before and after the parental leave extension of 1992. Because the treatment group comprises all women of childbearing age, actual information on parental leaves of the mother was not required for a dataset to be used here. Nevertheless, problems do arise in the dataset comparison. First, all three datasets measure the outcome variable, job-related training, for a different period of time: the last 5 years in the IAB-BIBB data, the last 3 years in the GSOEP, and the last 12 months in the BSW. The second difficulty stems from the needs of our difference-in-differences analysis. Not only does it require training incidence observations before and after the parental leave extension, but these can only be done properly by focusing on the most drastic reform, that which lengthens parental leave from 18 to 36 months. However, the post-1992 reform surveys differ enormously in timing: 1994 for the BSW, 1998 for the IAB-BIBB, and 2000 for the GSOEP. Variation also exists in the timing of the pre-1992 reform surveys, which refer to the following years: BSW, 1988; GSOEP, 1989; and IAB-BIBB, 1991. Obviously, these differences must be taken into account. For example, by asking for training in the 5 years previous to 1991, the pre-reform survey refers to a period during which three smaller extensions of parental leave benefits occurred (see the gray-shaded boxes in Fig. 2a–c). The surveys also differ somewhat in their sample sizes, with the largest, the IAB-BIBB, containing more than 16,000 observations per wave. GSOEP and BSW are smaller, the former with over 2,700 observations in 1989 but more than 5,000 in 2000 because of refreshment samples, and the latter with more than 3,000 and 2,000 observations before and after the reform, respectively.