Chillin'Competition

Relaxing whilst doing Competition Law is not an Oxymoron

Archive for October 20th, 2021

Will the DMA deliver on its promises? Part I: substantive aspects (or markets do not become contestable by decree)

with 7 comments

How Tech Companies' Election Promises Have Held Up – The Markup

As the legislative train carries on its journey towards the adoption of the Digital Markets Act (DMA), academics, stakeholders and consultants have produced a great deal of work examining how the future regime can be improved and made more effective, whether from an institutional or a substantive standpoint.

One of the questions that is far less prominent in these debates is whether the DMA can actually deliver on its promises. Put differently: is it realistic to expect that the regime will achieve every one of its ambitious aims? It seems to me that this question is, at the end of the day, more relevant. Achieving the optimal design of the regime would be a futile effort if it appears that it is expected to achieve an impossible task.

A regime that promises what it will be exceedingly unlikely to deliver is bound to create frustration (and even the false impression that failure is not a function of the regime’s structural inability to achieve its aims, but rather the regulator’s lack of willingness to meaningfully enforce it).

Against this background, I thought it might be a good idea to start a debate along the abovementioned lines, if only because managing expectations about what regulation can and cannot achieve is always a good idea.

The DMA is not particularly explicit about how it intends to transform digital markets. Contestability and fairness are, in this sense, more high-level aspirations than operational principles. Beneath Articles 5 and 6, I identify the following three goals (which I discussed in this paper of mine too):

Address the risk of leveraging from bottleneck segments to neighbouring ones: this is a traditional goal in competition law and sector-specific regulation (think of telecoms regimes, for instance). The idea is simple: ex ante intervention seeks to avoid the possibility of the extension of dominant position from one market to adjacent ones.

Re-allocate rents (away from so-called gatekeepers to firms in neighbouring markets): there should be little dispute that the ambition (arguably the single most important ambition) of the DMA is to redistribute rents within and across digital ecosystems. This goal is achieved through several mechanisms. A straightforward technique is to ban or water down gatekeepers’ monetisation strategies (for instance by allowing the side-loading of applications). Another one is to limit gatekeepers’ ability to fight free-riding (for instance, by banning altogether MFN clauses).

Inject competition in bottleneck segments: the final, and most ambitious, goal of the DMA is to inject competition within bottlenecks (that is, markets where a gatekeeper enjoys a dominant, if not superdominant, position). In this regard, the regime seeks to change the structure of platform markets to introduce and preserve effective rivalry. For instance, the DMA does not hide its ambition to inject competition on the market for search engines.

So: can the DMA deliver on these promises? When it comes to the first two goals, I am ready to guess that the consensus is that they are, in theory and practice, at least achievable. Preventing leveraging and re-allocating rents is what utilities regulation has routinely done for decades.

The real question, in relation to these first two goals, is whether they are necessary and/or desirable. I am not uncovering any secret when I say that there is no consensus on this front. Some believe ex ante intervention in this sense is both necessary and desirable, some do not. My views are also well known, and I developed them in this paper.

The third goal is far more interesting. As far as I can see, the DMA has probably set itself an impossible, or virtually impossible, task. If a market is a natural monopoly (or displays extreme returns to scale, to use the expression found in the Special Advisers’ Report), regulation, alone, cannot change the tendency of a market towards concentration (or, more precisely, towards the emergence of dominant or superdominant positions).

The above is also true, by the way, of competition law. No matter how forcefully one may want to preserve rivalry on a market, if it does not structurally allow for it, it will not happen (which is why we have, for instance, the failing firm defence in the field of merger control, and the reason why dominance is not prohibited as such under Article 102 TFEU). It is important to bear this point in mind when debating whether competition law is working or is being applied effectively in digital markets: are some stakeholders frustrated with recent enforcement because competition law is now expected to achieve something it cannot and was never designed to do?

At most, competition law and regulation can intervene at the margin, and act against any practices that may accelerate the natural tendency of the market towards concentration (by acting, for instance, against exclusivity obligations). Ultimately, however, they will have to deal with the reality that the economic context allows: they cannot force, by regulation, different features.

Effective competition in a market that is structurally unsuited for it will come from technological and economic change, not from regulation. Regulation can, at most, accompany the process and perhaps reduce the scope of the bottleneck segment (not eliminate it). Telecommunications is a wonderful example in this regard (and the EU Regulatory Framework for electronic communications is the gold standard, in my view, of how regulation is to engage with changing realities).

The task is even more difficult as far as the DMA is concerned: dealing with natural monopolies is hard enough when the issues are well known and have been researched inside out (as in telecommunications or electricity). Just think of how much harder it is when the drivers of markets and technologies are not fully understood (and, as the Special Advisers point out in their Report, market dynamics in digital markets are not yet fully grasped).

I tell myself that a task for academics, moving forward, is to manage expectations: I just hope that the success or the failure of the DMA will not be measured against its ability to deliver deconcentration in bottleneck segments controlled by one or two gatekeepers. This point does not justify, alone, changing the regime, but it is a powerful reason to be realistic and to understand that the DMA will most probably underdeliver (and will do so by design).

(As you know, I have nothing to disclose in this or indeed any other matter).

Written by Pablo Ibanez Colomo

20 October 2021 at 12:27 pm

Posted in Uncategorized