Relaxing whilst doing Competition Law is not an Oxymoron

On the possible ex ante regulation of online platforms (I): lessons from the EU telecoms regime

with 7 comments


Before our world was turned upside down, a major theme in our community concerned the possibility of regulating ex ante online platforms. No matter how challenging the times, it is safe to predict that this question will not completely go away in the foreseeable future.

For the same reason, it makes sense to start a conversation about the possible ex ante regulation of online platforms, and more precisely about (i) why we would want it and (ii) what it would look like if it is ever going to see the light of day. Further posts will follow, as well as, hopefully, some sort of online discussion.

I would not break new ground by saying that it is an exercise that requires careful reflection. Any regulatory regime needs to be scrupulously crafted to ensure that the public interest is preserved. When there is so much at stake this is often a challenge. And claiming that something ‘needs to be done’ is not sufficiently compelling.

Fortunately, we have a template that gives us an idea of why and when to regulate online platforms. I see it as an insurance against sub-optimal outcomes. The EU telecoms regime (that is, the EU Regulatory Framework for electronic communications) asks the right questions and provides the right answers. Every time I look into it I realise that it is, in fact, an impressive achievement.

The EU telecoms regime has always been part of my research and teaching and thought it would be a good idea to start by discussing on the blog how a regulatory framework inspired in its approach would work, and whether there are good reasons to depart from the principles and approach enshrined in it.

Principles of regulation (and how they could apply to digital markets)

When the European Commission proposed the adoption of a new regulatory framework for electronic communications, it was aware of the challenge posed by innovation and technological evolution. Telecommunications activities were undergoing major changes; it was necessary to adapt regulation so that it would not become obsolete (that intervention would continue where it would no longer be necessary and/or it would be unable to address new issues).

The main principles of the framework can be summarised as follows:

  • Flexibility and adaptability: legislation sets the objectives and the procedure to be followed by regulatory authorities; it does not directly prescribe the remedies.
  • No regulation for the sake of regulation: the framework is premised on regulatory humility. Competition is seen as the best form of regulation, and as the default. Regulatory intervention only takes place where it can be shown that competition law would be unable to address the problem (mainly due to the problems that come with proactive remedies) and where there are major and lasting structural problems in a market.
  • Technology neutrality: the framework was crafted to ensure that it would not favour a particular technology over others, and that the market, not regulation, would pick winners and losers.
  • Bias against intervention in newly emerging markets: the framework is particularly careful not to intervene in nascent markets, where it is difficult, by definition to anticipate their evolution.

If ex ante regulation is ever adopted for online platforms, I hope the principles outlined above will also guide the regime. Two particular aspects deserve attention.

First, I cannot think of a good reason why an ex ante regime for platforms should not revolve around principles, as opposed to outcomes. In other words: it seems to me that ex ante regulation should provide for the objectives and the process through which the need for intervention is considered. By the same token, the details of intervention should be left to a regulatory authority. The temptation to enshrine particular remedies in legislation may be strong. Wisely, the EU legislature resisted it.

Second, it seems to me that neutrality should be a key guiding principle, in particular when (as acknowledged in the Special Advisers’ Report), there is much that we do not know about the gains resulting from practices in the digital sphere. There appears to be every reason to ensure that any ex ante regime does not have a built-in bias in favour of, or against, certain practices or business models.

The operation of the regime

As a principles-based regime, the EU telecoms regime leaves it to the regulatory authority to decide, on a regular basis, whether intervention is warranted and, if so, the remedies that may be needed to promote competition and innovation on a particular market segment.

How does the system work? If you are not familiar with the framework, you may not know that market definition (understood in the competition law sense) is central to the identification of regulatory concerns. The best explanation of when and why remedial action may be warranted is to be found, I think, in the Commission’s Explanatory Note on market definition.

The starting point of the analysis is to define, in accordance with competition law principles, a market on a segment that is potentially competitive and that is adjacent to a potential bottleneck. In telecoms, this adjacent segment could be, for instance, the one for retail broadband services (and the potential bottleneck the local loop).

In the digital sphere, this adjacent segment could be (I mention some examples that can be easily visualised) the online distribution of a particular good (say, books), a particular category of app (say, music), or a particular online service (say, travel). Each of these segments would relate to a particular bottleneck segment. Some good candidate markets that are less easy to visualise are extensively discussed in the CMA’s Interim Report.

The key question is whether these adjacent segments are effectively competitive. If they turn out to be effectively competitive – in the sense that no dominant position can be identified – then regulation would not be warranted. No matter how strong the position of a market player on the potential bottleneck segment, no remedial action will take place (again, no regulation for the sake of regulation).

If they are not effectively competitive, then the two main questions that follow are, first, whether the potential bottleneck segment to which these activities relate has ‘high and non-transitory barriers to entry’ and, second, whether a dominant position can be established on this market. As part of this assessment, it would be necessary to consider whether this position is likely to last in the foreseeable future or whether, instead, technology and innovation would undermine this position (for instance, by allowing the firms on the adjacent segment to circumvent the bottleneck).

Only where these two questions are answered in the affirmative will the regulatory authority be in a position to impose remedies, which have to be adapted to the specific demands of every case (they range from transparency and/or non-discrimination obligations to the functional separation of some activities).

The operation of the EU telecoms regime provides two valuable lessons. Again, I cannot think of a good reason to deviate from them.

First, regulatory action should be based on a market-by-market assessment of the conditions of competition, not on a broad-brush identification of companies as having a ‘special status’ on the basis of vaguer, more general criteria. Second, where a market is effectively competitive absent intervention, regulatory obligations would not be warranted. In such circumstances, they can well do more harm than good.

As usual, your thoughts and comments would be most welcome.

Below you find, reduced to the essential, the logic of the framework.

Regulatory framework

Written by Pablo Ibanez Colomo

24 April 2020 at 11:42 am

Posted in Uncategorized

7 Responses

Subscribe to comments with RSS.

  1. Thanks very much for this interesting new saga on a basic topic.

    José Antonio Rodríguez Miguez

    24 April 2020 at 11:52 am

  2. I fully agree – and this appears in line with the Commission’s thinking. As Ms Vestager has just said at the ABA Spring Meeting: “ we would explore the possibility of ex ante regulation, listing clear-cut prohibitions and obligations for those who may be called digital gatekeepers”.


    24 April 2020 at 10:42 pm

  3. Very interesting. I agree with the idea that we should define a systematic and balanced approach to regulation before embarking on it. One such approach is to use the criteria for “proportionality”. The idea is that because regulation tends to stifle or slow down innovation and freedom of choice, we should impose regulatory measures only if they meet a proportionality test. Thus, regulatory measures should meet four conditions: (1) They should serve defined legitimate goals (excluding protectionism, because countries that favour regulation and protectionism over innovation and education will inevitably lag behind fast-growing economies and deprive their consumers of the fruits of this growth). (2) They should be effective and adequate to achieve those goals. (3) Regulation should be “no more than necessary,” in that there should be no less restrictive alternative. Top-down regulation by the state does not necessarily meet this “necessity” test in circumstances where there are reasonable technology-based or market-based alternatives. Of course, where there are serious market failures, regulation may indeed be necessary. And market power combined with high barriers to entry, which is the focus of your comment, is but one example of a market failure. Others include public goods, negative externalities, coordination problems, and imperfect information access. These play a major role, for instance, in our exploitation of the environment, leading to pollution, illness, and climate change. (We discussed that last year during your @chillin’competition conference, and, by the way, the article providing the more detailed thinking behind that “Sustainable Competition” talk has just this week finally appeared in CPLD). (4) A balance of interest test is required, to ensure that the disadvantages from regulation (including cost of enforcement and false positives) do not outweigh the advantages. This analysis should be based on evidence, not ideology or fear, or a desire to protect incumbents.

    I think the environment and climate should be our first priority, but continuing the focus on the digital, the UK House of Lords Select Committee on Communications published a report titled “Regulating in a Digital World” that may deserve more attention than it appears to have received. They recommend regulation of the digital world based on 10 underlying principles: Parity (the same level of protection must be provided online and offline); accountability; transparency (regulators and digital businesses alike must be open to scrutiny); openness to innovation and competition; ethical design in the interests of users and society; privacy; recognition of childhood; respect for human rights and equality; education and awareness; and democratic accountability, proportionality and evidence based approach. You may or may not agree with all of the conclusions of the report, but the I thought the principles they list and the analysis were interesting. See

    Maurits Dolmans

    25 April 2020 at 12:07 pm

  4. Thanks, Pablo.

    Effectively this regulatory template from the telecommunications is flexible (and robust) enough to take account of even those technical features that differentiate telecoms from data-driven markets. Network effects in data, that interoperability in telecoms largely mitigate, and economies of scope in data that facilitate ecosystems can be accounted for at the stage of assessing entry barriers and ease of access in the adjacent market(s).

    Vikas Kathuria

    25 April 2020 at 5:58 pm

  5. I agree and have drawn parallels myself (especially set top box regulation). It helps to have grown up as a lawyer during the roll out of the EU communications regulatory regime and its UK implementation, even though it makes me feel old… I assume the Furman Review panel had the telecoms SMP-based threshold in mind (as well as the Groceries Code) when suggesting the strategic market status trigger for regulation. While both SMP and SMS concepts are clearly linked to market power, they also appear to be designed to provide more adminstrable bright lines for intervention than dominance. Of course, what we don’t yet have is an EU-wide uniform framework, so there’s plenty scope for diverge and attendant complexity (before you even get started on Brexit).

    Becket McGrath

    29 April 2020 at 5:04 pm

  6. Many thanks Pablo. I also believe that the template of electronic communications (ECS) is helpful for digital platforms. However, I think that there is an additional criterion that we need to think about when assessing, in particular, whether we should impose mandatory access to data.

    Under the ECS framework it is assumed that we have the right tools to intervene in the market: economists/regulators have developed a broadly appropriate way to intervene and set the prices for accessing ECS networks at wholesale level. I am less sure that we can apply that template to calculating an appropriate price for access to data.

    The reason is that ECS or utilities operate in (i) mature, (ii) relatively low-risk markets, and (iii) derive their profits mainly from physical long term assets (think of telecoms cables, ducts, exchanges). Conversely, digital platforms are a very different business: (i) not a mature market (think of the growth of online advertising vs traditional advertising); (ii) high-risk markets (they derive earnings from advertising, which is highly pro-cyclical and therefore risky); and (iii) their profits come mostly from their investments in R&D (not physical assets).

    In my view, this means that on top of the 3-criteria test that we apply in ECS regulation to decide whether to intervene with regulation (which you have summarised in your post), we also need to ask ourselves whether we have the right toolkit to determine what is an appropriate price for data. If, contrary to the case in ECS, we are not well equipped to set a reasonable price for data, I think we should be cautious. The reason is that (i) there would be a higher probability of making Type I errors and (ii) the costs of those errors could be potentially higher than for utilities (because we could distort incentives to invest in competition for data – and Big Tech invest a lot in that).

    Jordi Casanova

    18 May 2020 at 7:04 pm

  7. […] On the possible ex ante regulation of online platforms (Pablo Ibáñez Colomo) […]

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.

%d bloggers like this: