The Harmonised BIBB/BAuA Employment Survey 2006, 2012 and 2018 (H-ETB)

1 Introduction

The analysis of changes in the employment conditions and workplaces in Germany is an ongoing task within the social and economic sciences. The harmonised BIBB/BAuA Employment Survey (H-ETB) 2006, 2012 and 2018 (https://doi.org/10.7803/501.061218.1.1.10) is a valuable source of information on the changing nature of employment in Germany. The dataset combines three waves of the BIBB/BAuA Employment Survey of the Working Population on Qualification and Working Conditions in Germany (short BIBB/BAuA Employment Survey).^[1] The pooled dataset simplifies the use of the BIBB/BAuA Employment Survey for analyses over time, thereby facilitating insights into long-term trends and dynamics of the German labour force.

The BIBB/BAuA Employment Survey is a representative survey of employees in Germany, aged 15 years or above and working at least 10 h per week. The surveys were conducted by the Federal Institute for Vocational Education and Training (BIBB) in collaboration with the Federal Institute for Occupational Safety and Health (BAuA). The data covers information about employment, the workplace and socio-demographic data.

The harmonised dataset of the BIBB/BAuA Employment Surveys 2006, 2012 and 2018 is useful for two purposes. Firstly, the data facilitates time comparisons. Secondly, analysing the pooled data allows more specific analysis for smaller groups, such as certain occupations, due to the larger number of cases.

A number of publications have already used the BIBB/BAuA Employment Surveys to conduct over time analysis. For instance, Bauknecht and Wesselborg (2022) have studied the emotional exhaustion in different occupations between 2006 and 2018. Further, Hall (2021) has used the surveys from 2006 to 2018 to analyse the effects of the Bologna reform on returns on education in Germany. Meyer, Tisch and Hünefeld (2019) have conducted regression analyses to examine the relationship between the implementation of new technologies and work demands.

The pooled data has already been the basis of several publications on smaller workforce groups. For instance, Rohrbach-Schmidt (2020) has used the pooled data from 2006, 2012 and 2018 to analyse whether foreign skilled workers have similar access to licensed and more credentialed occupations, and whether they profit from these regulations in terms of similar wages in these occupations to comparable domestic skilled workers. Storm (2023) demonstrates the presence of omitted-variable bias in conventional task data derived from expert assessment by using the BIBB/BAuA Employment Survey from 2012 to 2018 at the occupation-level.

The H-ETB simplifies various further use cases, for instance, the construction of indicators at occupation level for all three years, which is more straightforward for researchers. The information can then be linked to longitudinal data. Additionally, the H-ETB facilitates the application of complex statistical methods, including those in the field of causal inference, due to the large number of cases available. Matching methods can be used more effectively with a greater number of cases. Therefore, we are convinced that the research potential of time analyses and analysis of pooled data has not yet been fully exhausted and can be exploited with the new dataset.

The harmonised BIBB/BAuA Employment Survey comprises 596 variables surveyed in each of the three survey waves and all 60,048 cases. The principal advantage of this dataset over using each single wave is that all the harmonisation over time has already been done. The dataset only contains variables that can be analysed in all three survey years and these variables all have the same names, labels, categories and filters. Further, the dataset facilitates the easy assessment of which time or cohort comparisons can be processed with the BIBB/BAuA Employment Survey between 2006 and 2018, as it brings together the necessary information. This combination eliminates the need to check the questionnaire for each individual wave of the survey. Instead, researchers can use the codebook that comes with the data. Taken together, using the harmonised BIBB/BAuA Employment Survey for temporal analyses reduces the effort required for data preparation and saves time.

2 Subject Areas and Survey Instruments of the BIBB/BAuA Employment Surveys

The BIBB/BAuA Employment Surveys cover a wide range of topics, enabling research across different fields. First, the BIBB/BAuA Employment Surveys concentrate on the employment and the workplace context at the time of the survey, encompassing job tasks and required knowledge. Second, they focus on information about the job holders, such as job satisfaction, workload and health. Third, the surveys collect longitudinal information on the educational and career history of the respondents. Forth, the data includes sociodemographic and company data. Table 1 provides an overview of all covered topics.

The BIBB/BAuA Employment Surveys have several strengths: a wide range of indicators reflecting the value of vocational qualification, detailed information of all vocational qualifications, well-measured and in a multitude of ways classified information on current, first and training occupations, a high sample size of about 20,000 respondents per survey, and comparability of surveys over time.

3 Study Design and Data Collection

4 Overtime Harmonisation

The H-ETB incorporates variables from the SUFs of individual waves of the BIBB/BAuA Employment Survey, which are comparable across 2006, 2012 and 2018. All variables only asked in one or two waves were excluded from the harmonised dataset unless they are the basis for the filter. Currently, the variable names are not consistently used across the three SUF, which implies additional effort for the data preparation for over time analyses. As part of the harmonisation process, we standardised all the variable names, thereby reducing the number of necessary data preparation steps.^[3] The same is true for the variable and value labels of the H-ETB. Again, differences between the waves exist which are not driven by changes in the data collection. We standardised all variable and value labels based on the 2018 survey.

Although the majority of the questions remained unchanged over time, some modifications were necessary. The date of the H-ETB indicates by the variable label which questions were affected and in which survey year the question changed.

For some variables, further harmonisation steps were necessary, such as merging categories or adjusting filter guides so that they contain comparable information across all three surveys. To ensure transparency, in these cases, a copy of the original variables for each survey year was created with the endings “o06”, “o12” or “o18”. The missing values from the other survey years of these variables were coded with −9006, −9012 and −9018. With these variables at hand, users can decide whether they want to follow our harmonisation or whether, for their research question, another kind of harmonisation would be more appropriate.^[4]

As previously mentioned, an advantage of the data are the occupational codes. However, the coding scheme changed over time. In order to facilitate the data preparation process, the number of occupational classifications in the H-ETB has been reduced. This allows for the classifications to be used in a comparable manner across all three waves. Table 3 shows which occupational classifications are included for each type of occupation.

5 Data Access and Further Documentation

The harmonised BIBB/BAuA Employment Survey 2006, 2012 and 2018 is available as a Scientific Use File (SUF) at the Research Data Centre of the Federal Institute of Vocational Education and Training (BIBB-FDZ). The use of the data is subject to guidelines and requires application (see https://www.bibb.de/en/1400.php). The application could be submitted via e-mail to fdz@bibb.de, via fax to +49 – (0)228-1072020 or via mail to BIBB – Federal Institute for Vocational Education and Training research data centre P.O. Box 201264 53142 Bonn, Germany. After receiving the application, the BIBB-FDZ provides the data as download.

In accordance with data protection regulations, the detailed occupational codes are anonymised and excluded from the SUF. However, a separate dataset for special variables includes those variables and users can apply for it (https://www.bibb.de/dokumente/pdf/BIBB_FDZ_application_ZSV.pdf).

In addition to the data, the BIBB-FDZ provides a codebook in both German and English, which includes all variables their variable and value labels, the corresponding question, and, if applicable, the filter. The data itself is labelled in both German and English.^[5]

For university lectures and students who have not yet attained a master’s degree, the BIBB-FDZ provides a Campus-File (CF; https://doi.org/10.7803/501.061218.1.8.10), which is a more anonymized version of the Scientific Use File. The CF generates identical and comparable results.

6 Outlook

The latest wave of the BIBB/BAuA Employment Surveys just has taken place. Together with the SUF for the 2024 data, just like with every new wave, also the H-ETB will be update.

In addition to potential new waves, there are also those surveys from before 2006. Harmonising those waves as well would be also a great value for the scientific community. However, these datasets are less comparable with those after 2006. Hence, more harmonisation steps are necessary which will be a future project.