Ownership and the multinational wage premium

ABSTRACT

Combining administrative data on German workers at the plant level with detailed global ownership information, we provide evidence that the multinational and hierarchical structure of networks matters for wages paid by plants. We find that both network characteristics are equally important for the wage premium paid by multinationals. Moreover, the hierarchical distance is also important for wages paid by plants in national networks.

KEYWORDS:

• Multinational wage premium
• hierarchical network
• ownership

JEL CLASSIFICATION:

• F14
• F23

Acknowledments

The authors would like to thank the reviewer for very helpful comments and suggestions. This work was supported by the German Research Foundation for the project “Theory and Evidence on the Multinational Wage Premium”.

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

¹ For details about the BHP see Schmucker et al. (2018).

² Note that firms identified in Orbis can own more than one plant. Details of the dataset, further summary statistics, and a discussion of control variables used can be found in Egger, Jahn, and Kreickemeier (2018).

³ In line with theoretical arguments and empirical evidence we find a significant positive effect of plant size on wages in all our estimations. Whereas our data do not directly contain information on research and innovation, as a proxy we control for the share of workers being engineers. As expected, this variable has a positive and significant coefficient.

⁴ The worker fixed effects are provided by the IAB and estimated through an Abowd-Kramarz-Margolis (AKM) decomposition (Abowd, Kramarz, and Margolis 1999). The IAB also provides AKM plant fixed effects, which are estimated as seven-year averages. However, including plant fixed effects would pick up the multinational wage premium for plants that have been in foreign ownership over (part of) the estimation period. This would deflate the estimates of the multinational wage premium.

Additional information

Funding

This work was supported by the Deutsche Forschungsgemeinschaft [322072184].