direct participating interests are to be reported whenever nonresidents hold
more than 50% in a domestic firm and these dependent enterprises themselves
hold 10% or more of the shares or voting rights in other domestic enterprises.
An appealing feature of this data-set is that it comprises the universe of in-
ward FDI (above the reporting threshold) undertaken in Germany. For each
of these enterprises, we have information about the municipality where their
German headquarters are located.10
Municipality-level data on the qualification of employees were compiled
on special request based on the universe of German social-security records of
the German Federal Labor Agency (Bundesagentur für Arbeit). The obser-
vations are the universe of workers registered for unemployment insurance,
representing around 80% of the German workforce.11 Our skill measure is
the share of workers with tertiary school education.
Municipality-level data on business tax rates, population, geographical
area, the independency ratio, the fraction of land area covered with build-
ings and streets (a measure of available infrastructure), and gross investments
are provided by different federal statistical offices of the 16 German states
(Lüander) in the database Statistik Lokal distributed by the German Statis-
tical Office (Statistisches Bundesamt).
As mentioned before, there are two advantages of looking at location
choices at the sub-national level. One is that firms face a much more homo-
geneous institutional setting across municipalities within a country than at
the international level. Another advantage of using sub-national data is that
one may account more accurately for the heterogeneity of locations. Cross-
country data use national averages that might suffer from aggregation bias.12
the micro-level data set MiDi see Lipponer (2008).
10 In fact, the information contained in the data-set does not identify individual sub-
sidiaries but refers to regional headquarters which may report data on several of their
subsidiaries. These headquarters may be responsible for a production facility at the same
location or a set of subsidiaries in Germany. This is not of importance here, since we
focus on regional headquarters location and the number of such entities (rather than em-
ployees, turnover, etc.; the latter would be impossible to assign to headquarters versus
subsidiaries).
11Coverage includes full- and part-time workers of private enterprises, apprentices, and
other trainees, as well as temporarily suspended employment relationships. Civil servants,
student workers, and self-employed individuals are excluded and make up the remaining
20% of the formal-sector labor force.
12 Consider two countries A and B with identical national averages. For instance, country
A might have a skilled labor force but a bad infrastructure in one half of the country