According to the managerial strategy literature, a, if not the, key to large
profits is the creation of “moats” that protect firms from competition.1 Firms
with market power create moats to maintain that power, and there exist strong
incentives to develop new technologies that allow for broader and deeper
moats. On the other hand, from a broader societal perspective, and particularly
from the perspective of consumers, these moats often are harmful: they
surround customers and deny them the opportunity to purchase from
competitors. As a result, consumers suffer from the high prices and/or low
quality imposed by the incumbent firm, whose incentives to provide the
amount and type of innovation desired by consumers are decreased.
Although the development of the digital sector in the last forty or so years
has brought enormous benefits, it—as is well known and will also be clear
from our analysis below—also creates natural moats and facilitates the creation
of artificial ones. Therefore, all over the world, researchers and policy makers
are discussing possible novel government interventions to promote the
competitiveness of the economy. One of the most prominent recent examples is
the new European regulation, the Digital Markets Act (DMA).
2
It is a complex
piece of legislation that has led to a new set of obligations on a small number of
the largest firms in the digital sector.
The two key and repeatedly stressed concepts underlying the DMA are
those of “fairness” and “contestability.” It is therefore important to explore in
depth the way in which fairness and contestability should be understood in the
implementation of the DMA, particularly from an economic perspective. This
clarification and explanation of the underlying economic foundations will be
helpful in several ways.
First, it will help with the interpretation of the obligations that the
gatekeepers must fulfill, described in Articles 5, 6, and 7.3 The DMA
obligations are not always perfectly well defined; nor does it seem possible to
define them ahead of time with sufficient precision and for every possible
future application. A better understanding of the concepts underpinning DMA
obligations may thus guide the Commission and the gatekeeper firms in their
discussions about how to apply them. This understanding might also help the courts when, at some point or the other, a firm accused by the Commission of
breaching its obligations under the DMA seeks judicial redress.
Second, the DMA provides in Article 8 certain processes designed to
ensure that the gatekeepers’ efforts to comply with their DMA obligations in
fact advance the DMA’s underlying goals. These “further specification”
processes—available with respect to obligations imposed by Articles 6 and 7
but not those imposed by Article 5
4—may be initiated by either the Commission
or by the gatekeeper. In either case, the language describing the specification
process makes clear the goal is to ensure that gatekeeper compliance efforts
match up with the purpose of the obligations, which is to advance fairness and
contestability.5 A thoughtful understanding of fairness and contestability will
therefore be important to the further specification process as well.
Third, the DMA describes when and how the Commission can and should
update, modify, or supplement existing obligations. In all such instances, the
DMA directs that changes to obligations are to be made in service of the
related goals of fairness and contestability. For example, the Commission may
reopen certain specification proceedings if the measures previously specified
are “not effective” in advancing the DMA’s goals (i.e., fairness and
contestability).
6 Additionally, the DMA empowers the Commission to adopt
“delegated acts”—essentially, amendments to nonessential provisions of a
piece of legislation—if a market investigation has identified specific
obligations that ought to be updated to address new gatekeeper practices that
limit contestability or generate unfairness.
7 Even decisions as idiosyncratic and
specific as adding or deleting messaging app functionalities to or from the list
of those covered by Article 7’s interoperability mandate must be made in with
the purpose of advancing the twin goals of contestability and fairness.
8 All of the amendment processes will be more predictable and effective when the basic
concepts on which they rest are better understood.
The analysis we conduct will allow us to develop some policy
prescriptions, which we summarize below. We will briefly discuss them as we
go along and return to a more extended discussion in the conclusion.
Policy Recommendations9
• The DMA should clearly define contestability and fairness.
• The implementation of the DMA should focus on encouraging
competition in the market and not just competition for the market.
• The platform economy leads to “unfair” outcomes when users are not
rewarded for their contribution to the success of the platform.
Regulation should aim at correcting this distortion.
• If well implemented, regulations based on the concepts of fairness and
contestability can be favorable to innovation.
The Article provides more economic background for contestability and
fairness, the two goals of the DMA. Before beginning our discussion, we
should state a few preliminaries.
First, all the authors of this document believe that some form of
regulatory response to the challenges posed by the digital economy is
warranted. They may not, however, be in full agreement on the form that this
response should take. Some of them believe that the creation of the Digital
Markets Unit in the United Kingdom and the new German competition law
provide attractive alternative models. We all believe, however, that a well
implemented DMA is a step in the right direction—especially if the concepts
of contestability and fairness are interpreted appropriately.
Second, all authors also believe that substantial dialogue between the
Commission and the gatekeepers on how exactly to interpret the obligations
that the DMA imposes on individual gatekeepers will be important. But the
DMA gives the Commission substantial discretion in deciding how many and
what issues to take up in the further specification process. We fear that limited
resources the Commission currently intends to devote to enforcement will lead
to a less than optimal amount of dialogue not just in connection with the
specification process but also in connection dialogue that should occur around
the gatekeepers’ annual compliance reports and other matters.
Below, we repeatedly caution the reader that our conclusions depend on the
quality of the implementation of the DMA, and it should be remembered at all
times that whether the goals of the regulation can be achieved depends critically on its implementation. Our analysis is predicated on the assumption
that the DMA will be applied to the handful of largest digital platforms.10 Some
regulations would be unwise if applied generally, but have, we believe,
positive consequences when applied to this select group. In particular, the
extent of “unintended consequences” is likely to be lessened for this group.