Friday, September 14, 2018
Justus Haucap, Wolfgang Kerber, Heike Schweitzer and Robert Welker have just finished a study on a potential reform of the German law on abuse of market power in digital times. It has received some public attention by the press in Germany and beyond, also because the minister for economic affairs has publicly endorsed some of the recommendations. There is no English version of it - but it is being prepared. In the meantime, this is an English language official summary.
Rupprecht Podszun has written about the report on his blog.
here was an email sent out to an ABA listserve by Thomas Funke describing the report. Based on that description, Heike Schweitzer wrote me to offer some clarifying comments which I reproduce with her permission below:
According to Thomas Funke, our report suggests that data could be
treated like an essential facility or SEP that should be made available
to third parties where this is necessary to prevent market foreclosure.
This is not what we suggest. The report does not discuss "data as a SEP".
We do discuss the broader issue of access to data and what the role of
competition law can or should be. We make a number of points here - in
an attempt to insert a greater degree of differentiation into a debate
which is characterized by over-generalizations and overly sweeping
According to our view:
(1) it doesn't make much sense to talk about "data" -- there are so many
different types of data, data in the production of which you need to
invest, data which is produced as a mere by-product of use, personal/non
personal data, "raw" data and processed data etc. There are also so many
different uses to which data can be put. For some uses, there may be
substitutes. For others not. If we want to apply competition law to data
access issues, we need to look at these issues case by case and in context.
(2) we need to distinguish between different settings when we talk about
access to data: (a) access to huge "data reservoirs" - that may
sometimes be needed to compete effectively, in particular in areas where
competition is, to a large extent, competition about the best algorithms
and the relevant algorithms need to be trained on large data sets; (b)
access to data in vertical relationships - e.g. machine producer has
exclusive control over use data / machine user wants access to that
data; also: digital platform that has exclusive control over all the
data about platform use / businesses presenting offers on the platform
(but in this latter case, we have to be very careful about data access,
as competition-sensitive data may be at stake - granting access to such
data could result in a publicly orchestrated cartel!) (c) access to data
by third parties in value creation networks where the "monopolisation"
of specific types of data will imply reserving markets for value added
(3) the essential facilities doctrine (EFD) may play a role with regard
to (a) - if all the preconditions of the EFD are shown; in particular:
the essentiality of data access for effective competition/ lack of
substitutability for a given data use. Nonetheless: If this data is
personal data - difficult: the GDPR will apply. If the data sets can be
anonymized, and if the relevant data is of a kind that is produced as a
byproduct without much investment needed, and if the innovation effects
of granting access are strong - under these narrow conditions it may
make sense to apply the "essential facilities"-doctrine. We doubt,
though, that the EFD will, in the longer run, be a sufficient legal
basis to handle the data access issues that will arise efficiently. We
believe that, rather, sector-specific regulation will emerge (and is
already emerging). There is a debate in Germany about an "data for
all"-law, though - which takes up some of the ideas suggested by Victor
As regards the setting (b): that's probably best solved by contract law.
Setting (c) might again be a competition law problem that can already be
addressed under German competition law. It may also be the factual
setting that needs a much closer look and economic/legal analysis in the