NEPS-ADIAB is a cooperation project conducted jointly by the Institute for Employment Research (IAB) and the Leibniz Institute for Educational Trajectories (LIfBi). The new linked data product contains survey data of the German National Educational Panel Study (NEPS) and administrative data from the IAB, the research institute of the German Federal Employment Agency. While the NEPS provides in particular information on educational trajectories and competencies, the administrative data contain detailed information on the employment histories including data about establishments the individuals were employed at. The first NEPS-ADIAB product was released in 2018. So far, five of the six NEPS Starting Cohorts have been linked to the administrative data via record linkage methods (SC1 “Newborns”, SC3 “Grade 5”, SC4 “Grade 9”, SC5 “First-Year-Students” and SC6 “Adults”). The NEPS-ADIAB data are updated regularly: By the end of 2023, the linked Starting Cohorts will cover administrative data from 1975 up to 2020 (SC3) and 2021 (SC1, SC4, SC5 and SC6) in addition to the current NEPS waves. The development of the NEPS-ADIAB data portfolio is an essential step in combining different data sources to enable new research questions and advanced empirical findings. In terms of NEPS-ADIAB research questions relating to the labour market and educational trajectories can be answered. Thereby, the heterogeneity of long-term returns to education or gender differences with regard to occupational careers or impairments of these due to motherhood are of interest. The school cohorts in particular allow for a variety of research questions aimed at different forms of transitions from school, vocational training or university studies to employment. The data access is free for non-commercial research purposes. In addition to a large number of on-site access locations remote data execution is offered.
Since 1961, the UGU project (Utvärdering Genom Uppföljning) has combined questionnaire data from pupils, their guardians and teachers throughout Sweden with school administrative and national register data to learn more about young peoples' well-being levels, and cognitive and educational development in Swedish schools. Identical follow-up surveys are sent to the same individual students every three years in order to map the changes in student health, performance, school criteria and home life over time. UGU is Sweden’s largest longitudinal databases for educational research. It consists of survey data collected from 11 birth cohorts between 1948 and 2010.
In Germany, a general school-leaving qualification is typically followed by a vocational education and training (VET), which is then connected to entry into working life (Weil & Lauterbach, 2011). According to the Vocational Training Report of the Federal Ministry of Education and Research (2021), in 2019, before the coronavirus pandemic, a total of 511,799 people decided to apply for an apprenticeship. About one-third had a secondary school diploma at maximum, and about one-fifth have a migration background. Previous studies have shown that particularly low school qualifications and a migration background can hinder the actual transition to VET, while vocational aspirations differ only slightly between pupils with and without a migration background (Granato, 2013; Kroll & Granato, 2013; Sürig & Wilmes, 2014). However, there are contrary theoretical approaches. In addition to self-selection effects, meritocratic arguments and individualized deficit approaches are based on a lack of interest and commitment. On the other hand, structural approaches focus on a lack of support structures, stereotypes of companies, and information deficits on both sides (Busse, 2021; Granato, 2013; Tjaden, 2017). In this regard, Tjaden (2017) reports from NEPS data that about 40% of ethnic inequalities during the transition to VET can be explained by self-selection. But he argues that these selection effects should only be understood as complementary to alternative explanations such as discrimination at this stage. The aim of this study is therefore to analyze the perceived discrimination of pupils its effects on the transition probabilities as well as the transition duration into VET. On the one hand, the general perception of discrimination due to a foreign-sounding name, a rather foreign appearance, wearing a hijab as well as lower German language skills, which were surveyed in the ninth grade, are used. In addition, the perceived personal discrimination based on a rejection for a training position due to origin is included. A specific focus is on differences between pupils with and without a perception of discrimination with regard to the a participation in programs and payments from the Federal Employment Agency in the context of their search for an apprenticeship. For this purpose, data from the fourth starting cohort of the National Education Panel (NEPS Network, 2023) will be used, which represents a comprehensive documentation of the educational and employment trajectories of students who have been surveyed regularly since attending ninth grade (Blossfeld & Roßbach, 2019). The sample NEPS-SC4-ADIAB also offers the possibility of using required administrative data for the analyses.
Regional mobility is considered as an important driver of economic growth and technological development because both strongly depend on workers’ ability and willingness to reallocate to the most innovative and productive sectors and labor markets. At the micro level, regional mobility is typically seen as a tool to improve individuals’ economic situation and well-being. Because of its importance on both the micro and macro levels, internal migration patterns and trends have drawn an ongoing attention among researchers and policy makers in many countries. The most prominent example is arguably the US, presumably because internal migration is fundamental to the American narrative (Zimran 2022). On the other extreme, so far, surprisingly little is known about this phenomenon in Europe’s largest economy – Germany. This paper fills the gap in the literature by presenting a comprehensive and detailed analysis of regional mobility patterns over the entire life-cycle in Germany. For this purpose, we draw on the data from the National Education Panel Study: Starting Cohort Adults (NEPS-SC6). The unique feature of the data is that for a large representative sample of individuals born between 1944 and 1986, it provides detailed information on their location nearly throughout their entire life cycle. We focus on approximately 15,300 individuals who were born in Germany. For them, we construct a balanced panel that allows us to track their geographic mobility in monthly intervals starting from birth, throughout their entire educational career, and until the age they reach in 2020 (i.e., 34 to 76 at maximum depending on the birth cohort). For different groups (e.g., by birth cohorts, gender, education), we document internal mobility patterns defined as cross-state, cross-county, and cross-municipal location changes. Finally, later using geocodes and administrative data from the ADIAB, we also calculate actual distance measures and socio-economic intentions. Contrary to the common conjecture that regional mobility in Germany is generally low, we find important differences across the life course with major location changes occurring in early childhood (i.e., before school enrollment) and early in adulthood, which coincides with post-secondary educational choices. Starting from the age of 20-25, individuals’ locational choices remain relatively stable. Nevertheless, there are substantial differences in age-mobility profiles across educational groups, where higher education is related to generally higher mobility rates during prime ages. We also document striking differences across federal states, not only along the former East-West-German border. Interestingly, we do not find notable differences by gender, which suggests that German women are not significantly less mobile than men.
Participation in further vocational training (FT) aims to ensure employability. This particularly holds in times of crises, e.g., recessions or pandemics, which heavily affect the labour and training market, e.g. by rising unemployment or decreasing human capital investments. Yet, research on the interrelation of the business cycle (BC) with FT is limited, primarily because an encompassing data-base is lacking. Existing research mainly focuses on the supply-side and is restricted to isolated exogenous shocks such as the financial crisis of 2008/09 or the COVID-19 pandemic. Consequently, our project focuses on how the BC affects individuals’ FT decisions across longer periods of the BC. Moreover, we ask whether reduced firm investments in FT are substituted by higher individual-level investments. We first give an overview on how BCs affect individuals’ FT (1). We then examine (2) the mediating influence of the technological change and (3) individuals’ risk preferences as potential mediators/moderators of the BC-FT relation which has been neglected hitherto. For (1) we use German Microcensus data (2005-2020), for (2) and (3) we use the longitudinal data from the adult cohort of the National Educational Panel Study (NEPS; 2010-2020 + Covid-19 survey). Both individual-level data is enriched by administrative data comprising various indicators of the BC (e.g. unemployment rates; GDP), and aggregated data on technological change (e.g. Mannheim Innovation Panel, EU KLEMS). The BC indicators are matched to NEPS data three months prior to the respondents' FT, or in case of no FT participation 15 months prior to the interview. The administrative data is linked by using regional, year and/or sectoral information. Since individuals’ FT starting months are mostly unknown within NEPS, we impute them from GSOEP (German Socio-Economic Panel) data using statistical matching techniques (Alpman 2016). For our analyses we will make use of several fixed-effects estimators which address different methodical challenges such as unobserved heterogeneity, asymmetrical BC effects (Allison 2019), and reverse causality (Ludwig & Brüderl 2018). Our presentation will both elucidate how we generated our enriched data-sets – which helps understanding individuals’ investments in FT along the BC – and summarize our first results.
Using the massive opening of academic track schools throughout the German educational expansion (1955-1980), we analyze monetary returns to high school education across the life cycle. For the analysis, we use linked survey and administrative labor market data from Germany (NEPS-ADIAB) combined with a purpose-built data set on all academic track school openings for cohorts born between 1950 and 1980. We exploit local changes in the geographical access to academic track schools to estimate local average treatment effects (LATE) and marginal treatment effects (MTE). We find sizeable monetary returns to the highest secondary schooling degree with average returns of over 60% (14% per year of additional schooling) for the primary working ages from 25 to 54. By applying the MTE approach, we want to study how the heterogeneity in the returns evolves after the labor market entry of the academic track graduates. Because the essential heterogeneity uncovered by the MTEs eventually correlates strongly with prime-age earnings (a stylized fact in the literature), it provides us with private information about the earnings potential of employees that employers arguably do not have initially, directly after labor market entry. Hence, by studying this evolution, this paper may inform about the relative strengths of human capital and the signaling effects that govern the returns to education.
During the COVID-19 pandemic, the German government minimized public life while preserving essential work activities (EWAs), categorizing workers as essential or ‘non-essential’. While prior research documented substandard recognition and compensation, dissatisfaction, and psychological distress among designated essential workers (EWs), little is known about the subjective group composition and self-perception of designated EWs. This study analyzes why designated EWs did (not) perceive themselves as ‘systemrelevant’ by integrating multiple data sources. The research draws on Social Identity Theory (SIT), which posits that people principally seek group identities to feel distinct, positive, and purposeful. Contextually, group identification is particularly likely when people align with the in-group prototype, the in-group is pertinent, and in-group membership promotes self-esteem. Using a mixed-methods design, the study first explores whether designated EWs are aware of in-group membership and attach value or meaning to it based on interview data (N = 57). This corroborates SIT hypotheses regarding (in-)congruent self-perception. Second, legislative categorizations of essential sectors and occupations are compiled from federal states’ ‘Infection Protection Act’ decrees and linked to the NEPS-SC6 Corona survey (N = 846) that asked respondents to self-categorize as ‘systemrelevant’. The legislative and subjective categorizations are contrasted, and logistic regression is used to test the SIT hypotheses. Qualitative findings suggest that EWs hold incomplete and prototypical notions of EWAs, exhibit mixed feelings towards the designation and derive varying degrees of purpose from in-group membership. Contextually, EWs who deviate from the in-group prototype of the burdened frontliner tend to disregard themselves as ‘systemrelevant’. Furthermore, in-group membership appears to be crucial for work-care reconciliation, while surprisingly perceived illegitimate status differences play a minor role. Quantitative findings show that only 60% of the designated EWs self-categorized as ‘systemrelevant’. Logistic regression largely confirms SIT hypotheses. Congruent self-categorization is strongly related to in-group similarity with on-site EWs being 25% more likely to self-categorize as ‘systemrelevant’ compared to EWs that relocated to home office. Likewise, the pertinence of in-group membership is a major factor; among EWs, parents are up to 31% (toddlers) more likely to self-categorize as ‘systemrelevant’ compared to those without children. Furthermore, EWs in low-prestige occupations, who likely experienced the highest status elevation, are 8 to 15% more likely to self-categorize as ‘systemrelevant’ compared to EWs in higher-prestige occupations. In conclusion, this study cautions research relying solely on the legislative categorization of EWs and challenges the notion of a coherent essential workforce group as frequently portrayed in media.
Information technologies gain importance in various professional fields, leading to an increased need for ICT skills (Kiener et al. 2019). Previous studies (e.g. Greenwood et al. 2011; Falck et al. 2016) confirm that ICT skills are relevant for individual labour market outcomes. This paper aims to contribute to this research focusing on the labour market outcomes of ICT skills using data from the Starting Cohort 6 of the German National Educational Panel Study (NEPS) (N=2,808) and the French 1998 Generation survey of the Centre for study and research on qualifications (Céreq) (N=13,342). Germany and France are two major European economies with distinct labour markets, industrial structures, and education systems. The comparative study allows us to analyse how the importance of ICT skills in the workplace and their impact on labour market success differ between these two countries. It answers the question how ICT skills pay off on the labour market in Germany and France. Following human capital theory (Becker 1964; Schultz 1961), we expect that the more ICT skills an individual has, the better their labour market outcome is. Firstly, we calculate linear regression models using log gross wages (Mincer 1974) as dependent variable and individual ICT skill values as independent variable controlling for gender (male versus female), labour market experience, migration status (yes versus no), type of education, and occupation (STEM versus Non-STEM). The results show that ICT skills have a significant positive wage effect in both countries (Germany: B=0.14; France: B=0.07 for ICT skills in email/information seeking and B=0.06 for ICT skills in data analysis/product conception programming). Furthermore, we could identify negative wage effects of being female (Germany: B=-0.507; France:-0.018). We find that working in a STEM-occupation or not only has a significant wage effect in France. The results replicate Mincer’s regression model and confirm the gender pay gap in both countries. Secondly, as the income is hardly the sole indicator of labour market success, we calculate the same models with a labour market success index as an alternative to the log gross wages as the dependent variable (cf. Annen 2018). This enables us to investigate labour market success from a broader perspective beyond pay inequity (Lester 2006), considering employment status, gross monthly wage, hierarchical status, type of contract, weekly working hours, formal qualification level required for the job and job characteristic, which are available in both data sets. The use of this index provides a broader view on labour market outcomes beyond a purely monetary perspective.
Research on survey methods shows that reminders are one of multiple strategies to increase participation rates in cross-sectional as well as panel studies. According to social exchange theory, which was described by Blau (1964) and further developed by Dillman (1978; 2000), the increase occurs as the survey institution has made an effort that should be rewarded by participation (Bethlehem, Cobben, & Schouten, 2011). While the positive effect of reminders on response rates is generally accepted (Becker, 2022), it is rather vague if they have any effects on the sample selectivity and it is discussed if this leads to lower quality of response data. With this background, we want to provide new evidence on whether reminders reduce social selectivity. Further, we investigate if increased participation rates come at the cost of lower data quality, as potentially less motivated invitees might be pushed to participate, leading to less accuracy or higher rates of item non-response. In the NEPS Starting Cohort 1, newborns have been tested and mothers interviewed annually since 2012. Mothers have provided information not only on their child, socio-demographics and social origin, attitudes, feelings and intentions, but also key information about their partners. In 2022, a NEPS pilot study attempted to include partners living in the same household to obtain first-hand information on their socio-demographics (in addition to proxy-information), satisfaction and well-being, partnership and family life, attitudes on value orientation and social trust, as well as their assessment of the child. The partner received an invitation to participate in an online survey after the mother had given her consent. A total of 1,529 invitations were sent in 9 tranches, depending on the mother’s interview date, spanning a time of 16 weeks. Initially, only 279 partners responded to this invitation. A reminder was sent to all of those who had not responded yet, on the same date, resulting in an additional 369 interviews. To provide answers on selectivity in participation, we apply survival and event history analysis, comparing selectivity by socioeconomic status, non-German-language use in the household and migration status, but also selectivity by previous participation behaviour of the family and parental relationship quality. Event history models allow us to compare selectivities after the first and the second invitation. In a second step, we examine data quality and compare item non-response, survey duration and interruptions between those who participated after the first invitation and those who responded after receiving a reminder.
back to top