The German generation internship and the minimum wage introduction: evidence from big data

ABSTRACT

The new German minimum wage applies a specific exemption clause for internships, where internships that last up to three months are exempted while internships that exceed three months are due to the minimum wage. Negative minimum wage effects on internships are heavily debated as internships are mostly non-productive. Difference-in-difference analyses that exploit establishment and regional variation in the bite of the minimum wage do not show a reduction in the number of internships. In addition, we pursue an innovative approach by using Google search data to analyse how the search intensity for internships changed in course of the minimum wage introduction. Difference-in-difference comparisons with other countries in Europe do not reveal an effect on the search for internship positions in general, but we observe a significant reduction in Google search for ‘generation internship’. This suggests that the underlying societal phenomenon of a generation entering internships without a perspective for regular jobs has lost in relevance.

KEYWORDS:

• minimum wage
• internships
• big data
• difference-in-differences
• Google
• Germany

JEL CLASSIFICATION:

• J38
• J29

Acknowledgments

We thank the anonymous reviewer and the editor for very helpful comments. We also thank Lisa Feist, Clemens Hetschko and Simeon Schächtele, as well as the IAB working group ‘minimum wages’ and participants of the “Workshop on minimum wages: Lessons from recent experiences and European perspectives” in Paris and the “Workshop on microeconomics” in Lueneburg for helpful comments and suggestions.

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

¹ Since we look at the search for internship positions in our analysis, this captures the effects on the hiring process for posted internship positions but not on the length of internships.

² At the establishment level, we use the absolute number of internships. Unlike the regional level, we observe many zeros at establishment observations making a regression in logs infeasible.

³ We try to verify the conjecture that the Google search data are relevant for actual economic behaviour by a comparison with other data sources in section 4.

⁴ We are not confident to extend this time series as the search frequency could potentially be influenced by our own research, which we started in 2017.

⁵ Kernel densities of the outcome variables of interest are presented in Figure A1 of Appendix A.

⁶ The Hawthorne effect describes biases that are due to changing behaviour when individuals know that they are part of a survey study.

⁷ “Mini-jobs“, which are characterized by a remuneration below € 450 per month, are exempted from obligatory social security contributions.

⁸ The country-specific trend coefficients are jointly significant in an F-test which shows a test statistic of 8.4 and a p-value smaller than 0.1 per cent. This supports the intuition of underlying country-specific trends that are irrespective of the minimum wage introduction.

⁹ Again, the F-test statistic for the relevance of the country-specific trend coefficients is 5.89 (monthly data) and 5.10 (weekly data) implying that the irrelevance of country trends is clearly rejected, supporting this latter specification with country trends.