One of the major labour market policies of the previous German government was the introduction of a national minimum wage that was initially set at €8.50 per hour and came into effect on 1 January 2015. New research evaluates this policy with the aim of estimating the range within which a minimum wage can be set without leading to adverse employment effects.
The study finds that Germany’s national minimum wage has affected regions differently. Specifically, while below-minimum wage workers were not adversely affected in more productive regions, the evidence suggests that there were employment losses in less productive regions. The results suggest that the employment effect turned negative in regions in which the value of the minimum wage was approximately 80% of the regional median wage.
The starting point for the analysis is the notion that despite being set at a common level, a national minimum wage can have regionally different implications. This is the case because in poorer regions it will be relatively high compared to the regional average or median wage, but relatively low in richer regions.
Starting from the assumption that regions are composed of differently productive firms which each have a certain degree of market power, the study develops theoretical analysis that relates the effect that the minimum wage has on employment in a specific region to that region’s level of productivity.
The analysis predicts that the minimum wage will lead to employment losses in the least productive regions. But employment losses become smaller and eventually turn into employment gains as regional productivity increases because sufficiently productive firms in those regions respond to the introduction of the minimum wage by raising employment.
This effect starts to subside as regional productivity becomes even larger because those regions will be mainly composed of firms that already pay above the minimum wage and will therefore be unaffected by its introduction.
The authors test this analysis empirically using administrative data from Germany. Given the richness of the data set, they are able to estimate the employment effect of the minimum wage separately for about 4,500 regions.
Consistent with the theoretical predictions, they find that in regions with lower median wages, the probability of being employed falls after the introduction of the minimum wage for those workers who initially earned below the minimum wage level. Regions with higher median wages, by contrast, experience an increase in the employment probabilities of below-minimum wage workers, while the beneficial effect flattens out in even richer regions.
The first implication of this finding is that Germany’s national minimum wage has affected regions differently. Specifically, while below-minimum wage workers were not adversely affected in more productive regions, the evidence suggests that there were employment losses in less productive regions.
The second implication concerns the range of the minimum wage. The results suggest that the employment effect turned negative in regions in which the value of the minimum wage was approximately 80% of the regional median wage. By contrast, the largest employment effect is predicted for regions in which the minimum wage accounted for 48% of the regional median wage.
The authors are currently working on a tractable general equilibrium model that would allow them to address other questions that are of interest to policy-makers:
- How large are the employment losses in negatively affected regions and by how much did employment grow in other regions?
- What would have happened, had the minimum wage been set at a considerably higher level of €12 per hour, as has been proposed by political parties?
- What is the impact of the minimum wage on regional employment if labour markets are no longer analysed in isolation, but if indirect effects working through adjustments on goods and housing markets as well as impacts on a worker’s decision to live or work in a different region are also taken into consideration?
Author information
Gabriel Ahlfeldt, London School of Economics and Political Science, g.ahlfeldt@lse.ac.uk
Duncan Roth, Institute for Employment Research, duncan.roth@iab.de
Tobias Seidel, University of Duisburg-Essen, tobias.seidel@uni-due.de