Occupations and survey research: methodological and substantive applications exploiting occupations as social contexts 2 |
|
Convenor | Professor Christian Ebner (University of Cologne, Germany ) |
Coordinator 1 | Dr Daniela Rohrbach-schmidt (Federal Institute for Vocational Education and Training, Bonn, Germany) |
Scholars might be interested in differences of occupation specific wage growth and want thus want to estimate a multi level growth model. But if employees change their occupation, multi level growth models are problematic because the relation of occupations and employees are not strictly hierarchical any more. It is analytically well known that cross classified models are the better choice. But there are no valid simulation studies showing the extend of the problem. Within the talk I will present results of Monte-Carlo-Studies which show how big the problem really is given more or less movement between occupations.
Previous research on wage growth after further education and training yielded mixed results. I argue that these unclear findings result from the neglect of labor market segmentation. This follows from the notion that the mechanisms that link education and wages differ between occupations and industries. In this paper I therefore compare the returns to further training after initial education between different segments of the labor market. I use the NEPS-SC-6 data, a panel study of working-age Germans. Individual level effects obtained through conditional difference-in-difference estimation are compared between occupations and industries using multilevel-methods.
According to the OECD, literacy and numeracy play a crucial role for the participation in economic life. Research demonstrated that these skills add variance to the prediction of wages. We extend the research by studying whether these competencies have differential effects on wages depending on people’s occupation defined by a subset of tasks. We expect competencies to have stronger effects on wages for occupations with complex tasks. Using German PIAAC and BIBB/BAuA data sets we expand the classical Mincer-Regression adding numeracy, tasks and their interactions to predict earnings. Results reveal small interaction effects, not fully supporting hypotheses.
The paper provides an effort to study job changes and their wage consequences in Germany by considering task content and its change at the worker level. Using recent data that provide information on the time spent on job tasks before and after individuals have changed their jobs, we find the following: First, job tasks differ among workers within an occupation and this variation is significantly related to workers socio-demographic characteristics and their human capital, second, individuals move into jobs that require similar task content and, third, job task requirements and task-specific human capital play an important role for