Last time I came here, a month ago for the World Bank Infra4Dev Conference, it was clear the question was no longer whether rules are needed or not.
The question now is what rules should be set?
On my side, I had introduced my four golden rules regarding regulation and the digital ecosystem. Let me remind it to those already here a month ago or introduce it to the others:
- The first is to focus on structuring platforms. This can be done through “asymmetrical regulation”. Forget classical horizontal rules.
- The second is to give the power back to the people (not the State): consumers, entrepreneurs, NGOs, academics. This can be done notably by regulating with ex ante rules to give more choice to the users and by regulating “with data”.
- The third is to rely on a specialised authority, ensuring a permanent monitoring of the regulated parties. An authority with a “stick”.
- Ensure a good articulation between the national level and the European level. Both levels are relevant.
Today, I would like to get more into the technical aspects of the first of these rules: who should we regulate?
For a long time, we have adopted general regulations of the internet.
At the EU level, eCommerce and ePrivacy Directives and the GDPR give robust principles but they suffer from being horizontal and covering all types of players without distinction. These laws put a burden on all stakeholders and users and can heavily hinder innovation and freedom.
While the e-commerce directive was intended to create a safe harbour for digital players, Big Tech have recentralised the Internet.
This concentration is evil, not the internet in general.
Who are the Big Tech?
Many reports and authorities, the most famous ones have been co-authored by Diane Coyle and Jacques Crémer, point at firms with a strategic market status (Furman), “systemic players” and other forms of dominant platforms. The debate is flourishing and comparisons with other sectors, notably the financial sector, are of great use.
We at Arcep are currently working on a definition of “structuring platforms”, in the frame of a workflow launched by the Government “les Etats généraux du numérique”.
Our current thinking is that the firms to be targeted as structuring platforms should be the one that act as « points of control » in the internet. Not only at economical level but also technical. The legal definition could be the following (this is a draft to discuss):
“online platforms or operating system providers, which are able to limit significantly the ability of users to exercise an economic activity or to communicate online, notably to their importance as intermediaries.”
A bundle of indicia may be listed to meet this definition. We have prelisted a number of indicia to start the discussion.
A primary set of 3 indicia more of an economic type could be complemented by a secondary set of 4 criteria focusing more on the social impact of these platforms.
1st: The platform cannot be circumvented, either because it has a “bottleneck-power” (users become captive or dependent)
2nd: It has a certain size in terms of the number of unique users of the platform. In practice, given the transnational nature of the structuring platforms, the relevant field could for example be that of users within the European Union, unless they go so far as to only select those from a Member State. The consolidated turnover of the group to which the platform belongs could also be a proxy.
3rd: The platform is integrated into an ecosystem controlled by the group to which it belongs, and this ecosystem makes it possible to use leverage from one business sector to another.
4th: The platform is an essential gateway to a set of digital content (gatekeeper), via the implementation of algorithms classifying or referencing public information.
5 th: The platform has access to large quantities of high-quality data. These data are also difficult to access (or replicable) by potential competitors of the platform. These could relate to data of general interest.
6 th: In the case where advertising activities are somehow linked to the platform, the platform’s market shares on the advertising market are significant (used as a proxy for the attention of users that the platform is likely to generate).
7 th: The financial valuation of the company owning the platform.
The identification of the main players to target is the first step of what we call in the telecom sector, asymmetrical regulation.
For the last 20 years we have opened up and maintained opened monopolistic markets to competition by focusing on firms who had a significant market power. Once designated as such, you can impose some fine-tuned remedies to these players.
I am not advocating for a utility-based regulation here. Regarding the relevant remedies, we have to rely on duly processes where solutions are tailor made in regard of the problem to be solved. No one size fits all! Non-discrimination, data interoperability and data sharing should be part of the discussion, depending on the situation. But don’t forget the very basic task of the regulator to collect data and monitor the market.
Rules set at the European level by the European Commission could guide national authorities identifying the regulated players and relevant remedies.
We are quite used to such methodology in the telecom sector and we are quite confident we could identify quite precisely such firms either at national or European level.
Anyway, either at the national or European level you need a regulator to implement those remedies, to assess the evolution of the declining or increasing players, to collect, store, analyse, transmit and share collected data, to sanction, when necessary.
But this is not about replacing the power of Big Tech with the power of a regulator. Regulators must also be accountable and controlled. Finally, their mission should be to empower the many, not the State.
This proposal is to be considered as a draft to be discussed and challenged. All relevant input is very much welcome. We have plenty of great reports and we now have to decide.
• Elements of reflection relating to the characterization of structuring digital platforms (pdf - 498 Ko) (December 2019 - French version)