Tag: : EU Law

Posted on

States as platforms under new EU (online platforms’) law

Published in European Law Blog, 4.07.2022

The recent political agreement on the Digital Services Act (the “DSA”) means that, once officially released, it will formally introduce into EU law the term “online platforms”: These (according to the Commission’s original proposal, at least) are meant to be “a provider of a hosting service which, at the request of a recipient of the service, stores and disseminates to the public information” (art. 2, point (h) of the DSA), whereby a hosting service, in turn, “consists of the storage of information provided by, and at the request of, a recipient of the service” (point f). Therefore, between the DSA and the Digital Markets Act (the “DMA”), that has also been recently finalised, a comprehensive framework for the regulation of online platforms is introduced in EU law, the first of its kind both in Europe and internationally.

What if, however, this framework was applied to states themselves? What if states fell within the definition of an online platform within this context?

1.   What is an online platform (in EU law)?

Platforms is a term that only recently entered forcefully not only the EU legislator’s but also general vocabulary: indicatively, Google trends indicate that interest was fairly low from 2004 until 2014 but has spiked since.

There is some repetition in the Commission’s definition: as seen, a “hosting service” is already defined in point (f), there is therefore no need to repeat “which, at the request of a recipient of the service, stores“. Accordingly, the “recipient of a service” is “any natural or legal person who uses the relevant intermediary service” (point (b). Consequently, leaving aside the part of the definition of an explanatory nature (“unless that activity is a minor and purely ancillary feature of another service and, for objective and technical reasons cannot be used without that other service, and the integration of the feature into the other service is not a means to circumvent the applicability of this Regulation”), the definition should read: online platforms store and disseminate to the public information at the request of their users. The digital, online environment is implied.

2.   The regulation of platforms

The EU’s first attempt to regulate online platforms came through the nick-named (since it does not constitute an “eponymous” EU act), “P2B Regulation” (Platforms-to-Business relations Regulation, Regulation (EU) 2019/1150 of the European Parliament and of the Council of 20 June 2019 on promoting fairness and transparency for business users of online intermediation services). In the Commission’s words, the P2B Regulation is the “first ever set of rules for creating a fair, transparent and predictable business environment for smaller businesses and traders on online platforms”.

The P2B Regulation does not refer expressly to online platforms in its text (preferring the, timid, step of only mentioning the “online platform economy”). Instead, it uses the term “online intermediation service” to denote information society services that allow business users to offer goods or services to consumers (art. 2, point (2)).

In spite of the potential mix-up between the P2B and the DSA approaches (the DSA using the term “intermediary service” in an entirely different manner, to refer to “mere conduit”, “caching” and “hosting” services, art. 2 point (f)), the fact remains that, to the Commission’s eyes at least, online platforms are indeed regulated by the P2B Regulation, at least as regards their relationships with their business users. In effect, the DSA is expected to govern the other side of the spectrum, namely the relationships between online platforms and their users or consumers (although, admittedly, it also covers business users too).

3.   Platform-specific rights

The EU legislator regulates online platforms in order to impose certain obligations upon them (much less to help them develop further or warrant their continued existence). From a B2B point of view, in the P2B Regulation the purposes of “fairness, transparency and predictability” are served through, indicatively, the requirement that online platforms communicate their rules and regulations in a plain and understandable manner, that if these rules change an appropriate notice period is observed, that any adverse actions taken by the platform (e.g. restriction, suspension or termination of an account) are justified and contestable, or that differentiated treatment is generally to be avoided.

A much more detailed and comprehensive protective framework is constructed by the combined forces of the DSA and the DMA, particularly for the largest online platforms, that are, aptly named, gatekeepers. This is to be expected, because in this case it is users (individuals) that are the beneficiaries of such beneficial treatment. In short, among others, the treatment of information found on online platforms needs to be carefully planned (in the sense that online platforms need to install takedown mechanisms, accompanied also by appropriate procedures so that affected parties can object to such takedown actions see Chapter III, Sections 1-3 of the DSA), and also risks need to be minimized (through risk assessments, or even crisis protocols, developed by online platforms or through the installation of compliance officers within their internal organisations, see Chapter III, Section 4 of the DSA).

From its part, the DMA discourages “practices that limit contestability or are unfair” (in its Chapter III). The list of what exactly such practices comprise is exhaustive, however a few among them stand out: for example, exclusivity (in the sense of locking-in users to specific platforms) or preferential treatment (as in ranking or access to “query, click and view data”) are discouraged, while interoperability (allowing other providers to access and work with the gatekeepers’ proprietary systems) is encouraged; also, users/individuals need to at all times be able to port their data to the platform of a competitor.

4.    States as platforms

Could states be perceived as platforms? First of all, one could easily take away the digital part of online platforms: Although critical for the DSA/DMA purposes, the definition given in the DSA to online platforms may well apply in the real world too: platforms store and disseminate to the public information at the request of their users. The user could be a citizen (or an individual, depending) and the public could be the whole wide world, both digital and natural.

A state viewed as a platform would then be the intermediary in an information flow from its citizens (users, individuals) to everybody else. Its role would be twofold: First, it would store information and, second, it would make them available to anyone interested. As such, its intermediation is critical, if not necessary: individuals have no means to store their information in some repository, at least in a secure and possibly permanent manner. Similarly, dissemination cannot be done by themselves, at least not effectively.

Both roles are critical to our lives. Information on our births, various personal or family statuses or deaths need to be somewhere stored for societies to function in the first place. Once safely stored, this information, enhanced with information on our employment, residence, taxes etc. needs to be transmittable, again for societies to function in any meaningful manner.

Accordingly, both roles are nuanced. Depending on the type of societies each one of us is (lucky or unlucky) to be living in, control of storage and dissemination of this information lies to a larger or lesser extent with the state concerned. In other words, what information is stored on the platform and how and when it can be re-used depends on the (democratic or not) society one lives in. The state-run platform may apply larger or smaller restrictions.

While this blog post cannot constitute a political philosophy analysis on the various roles of the state over human history, it does invite a new perspective, enhanced by the digital. Online circumstances, and the primacy of the role of information (within the “data is the new oil” context) means that the role of the state perhaps needs to be reviewed (also) under this lens.

Is there any difference between (contemporary) online platforms, as perceived and regulated by the DSA and the DMA, and states? Yes, at least one that is of paramount importance, that of validation of the information concerned. States practically operate as platforms already: they already store, for example, our birth certificates and disseminate them to third parties at our own request. Essentially the only difference between that and what Facebook or Instagram do, that let us share, for example, our photos with our friends, is validation: States validate the information that they store prior to its disemination while online platforms today (mostly) passively store and diseminate it. For example, my birth certificate is validated by the Greek state to whomever I wish it transmitted, whereas I may tag my photo on Facebook but Facebook has no other option than to accept my word that it is actually me appearing on it. (It is a different story, and discussion, whether the states basically own this information, because it is within their existence that it was created in the first place).

5.   OK, so what if states are viewed as platforms?

What practical consequences can be drawn if states are viewed as platforms? While this is a preliminary analysis, a number of issues immediately come to mind. In essence, the whole discussion on platforms and platformisation needs to be transferred from the market to the polis (political philosophy).

Evidently, the Commission’s initiatives are predicated upon market locic. Speficically, the online platform economy and the dominance of online gatekeepers have imposed a protective regulatory approach whereby the market needs to remain contestable. Equating the state to a market would carry grave consequences to this understanding. The state does not need to be contestable, neither should its monopoly in relation to the provision of services to its residents be overthrown – at least, to this author’s preferences. This is why the aforementioned regulatory initiatives by the Commission may serve as hints or insights as to a possible new role of the state, but cannot possibly constitute the end result of such thinking themselves. They open up new perspectives but do not provide final solutions. In other words, automatic substitution of online platforms in the texts of the DMA and the DSA by the state would be unthinkable.

Regardless of this new perspective, the role of a state as a, basically, service provider is not questioned. States, in one way or another (especially under the social contract theory), are basically a service provider to their citizens. A democratic state, for example, provides first and foremost safety and security, and then warrants all other basic individual rights. A developed state adds to the mix public infrastructures (roads, ports, airports), health and education. The list can get long (or short, depending on politics), however the fact remains that states basically provide services to their citizens.

Storage and dissemination of information is also a service. As said, it is a basic one performed inherently in any state (see, for example, the “informational state” of Bourdieu, p.213), therefore if there ever was a classification among services and types of states this would probably fall in the root category. However, what changes now, after some 3500 years of recorded human history other than, perhaps, the perspective, is the online environment.

Consequently, if states were viewed as platforms a basic proposition would be to apply whatever the EU legislator requires of online platforms to states as well. To a large extent this is already done: Accountability runs deep, at least in democratic states, through separation of powers and rule of law. There are, however, other rights that may be more disputed: The right to portability (meaning that anybody can change a platform/state at will) and thus abolition of exclusivity or the right to interoperability (meaning that states should enable other states to parallel administrative procedures) are perhaps harder for states to swallow. They therefore merit further analysis.

In addition, new developments may bring the role of states as platforms to the fore. Until today states merely digitise the off-line lives of their citizens, through e-government or similar applications. In essence, they strive to abolish the need for physical presence to perform administrative tasks, and thus increase productivity. At some point, however, this task will have been completed. If by then, assisted by technology and digital lifestyle, states start operating a digital space where their citizens could live in, their roles as platforms will forcefully come to the fore.

6.   Some concluding thoughts

States as platforms is a development determined by the internet. While their role as intermediaries in information sharing has accompanied them since the development of writing (which, after all, it was developed for similar purposes in the first place), it is the internet that may transform the states’ role in this regard. Particularly their relationship with their subjects/users/citizens may be affected. To begin with, one can start wondering which of the rights afforded to him/her by the EU legislator as imposed on Facebook or Instagram, he/she would also like to see imposed on their own state(s). The increased pace of digital transformation of our lives will unavoidably broaden this discussion even further.

Το Κράτος ως Πλατφόρμα;;

Εδώ και χρόνια κρατάω σημειώσεις γι αυτή την ιδέα, και να που τώρα τα σχέδια για την Digital Services Act (DSA) και την Digital Markets Act (DMA) της Κομισιόν, καθώς και το, φιλόξενο, European Law Blog που με εμπιστεύτηκε, μου έδωσαν την ευκαιρία να την λανσάρω. Προφανώς θα συνεχίσω πολύ πιο αναλυτικά, όμως ορίστε παρακάτω ορισμένα κομμάτια του κειμένου από τα αγγλικά (μεταφρασμένο αυτόματα – αλλά η μετάφραση δεν είναι καθόλου κακή!):

“Η πρόσφατη πολιτική συμφωνία σχετικά με την Digital Services Act σημαίνει ότι, μόλις εκδοθεί επίσημα, θα εισαγάγει επίσημα στο δίκαιο της ΕΕ τον όρο “διαδικτυακές πλατφόρμες”: […] Τι θα γινόταν, ωστόσο, αν αυτό το πλαίσιο εφαρμοζόταν στα ίδια τα Kράτη; Τι θα γινόταν αν τα Kράτη ενέπιπταν στον ορισμό της online πλατφόρμας στο πλαίσιο αυτό;

[…]

Θα μπορούσαν τα κράτη να γίνουν αντιληπτά ως πλατφόρμες; Πρώτα απ’ όλα, θα μπορούσε κανείς εύκολα να αφαιρέσει το ψηφιακό μέρος των online πλατφορμών: Αν και κρίσιμος για τους σκοπούς της DSA/DMA, ο ορισμός που δίνεται στην DSA για τις διαδικτυακές πλατφόρμες μπορεί κάλλιστα να ισχύει και στον πραγματικό κόσμο: οι πλατφόρμες αποθηκεύουν και διαδίδουν στο κοινό πληροφορίες κατόπιν αιτήματος των χρηστών τους. Ο χρήστης μπορεί να είναι πολίτης (ή άτομο, ανάλογα) και το κοινό μπορεί να είναι όλος ο κόσμος, τόσο ο ψηφιακός όσο και ο φυσικός.

Ένα κράτος που θεωρείται ως πλατφόρμα θα ήταν τότε ο μεσάζων σε μια ροή πληροφοριών από τους πολίτες του (χρήστες, άτομα) προς όλους τους άλλους. Ο ρόλος του θα ήταν διττός: Πρώτον, θα αποθήκευε πληροφορίες και, δεύτερον, θα τις καθιστούσε διαθέσιμες σε κάθε ενδιαφερόμενο. Ως εκ τούτου, η διαμεσολάβησή της είναι κρίσιμη, αν όχι αναγκαία: τα άτομα δεν έχουν τα μέσα να αποθηκεύσουν τις πληροφορίες τους σε κάποιο αποθετήριο, τουλάχιστον με ασφαλή και ενδεχομένως μόνιμο τρόπο. Ομοίως, η διάδοση δεν μπορεί να γίνει από μόνη της, τουλάχιστον όχι αποτελεσματικά.

[…]

Ποιες πρακτικές συνέπειες μπορούν να εξαχθούν αν τα κράτη θεωρηθούν ως πλατφόρμες; Αν και πρόκειται για μια προκαταρκτική ανάλυση, ορισμένα ζητήματα έρχονται αμέσως στο μυαλό μας. Στην ουσία, η όλη συζήτηση για τις πλατφόρμες και την πλατφορμοποίηση (platformisation) πρέπει να μεταφερθεί από την αγορά στην πόλιν (from the market to the polis).”

Posted on

The “act-ification” of EU law: The (long-overdue) move towards “eponymous” EU legislation

Published in European Law Blog, 26.01.2021

The recent release by the Commission of the draft Digital Services Act and the Digital Markets Act may have attracted significant attention from stakeholders for their substance. However, this contribution is not about the content but the titles of these acts. In the author’s opinion, these two instruments are the latest addition to an emerging trend among the EU law-makers to release, “acts” or at least “eponymous” pieces of legislation. This trend shall be referred to here as “act-ification” of EU law. This trend is to be welcomed in that it signifies a new confidence and self-assuredness of EU law. After more than half a century since it came into life, EU law now seems to feel confident enough to release “acts” or, at any event, eponymous pieces of legislation, immediately recognizable by Europeans.

  1. Identifying a trend

The Commission’s draft Digital Services Act and Digital Markets Act (each one also fitted with acronyms to even further ease reference – DSA and DMA) are the latest additions to a Data Governance Act and a European Climate Law. These followed the Cybersecurity Act of 2019, which in turn followed the, perhaps EU law’s Big Bang moment in terms of public awareness and direct effect on the (EU) ground, General Data Protection Regulation of 2016. If we were to indeed identify a trend and then try to make a distinction, it could be claimed that, while eponymous EU pieces of legislation were not unheard of in the past (see the case of the ePrivacy Directive below), the significant turning point came during the Commission’s Priorities Programme 2019-2024. Almost all of the above instruments fall under it, signifying an underlying consistent and persistent (but not uniform, e.g. Regulation on serious cross-border threats to health or the Directive on the resilience of critical entities that follow the old pattern), cross-sectoral approach.

For the purposes of this analysis, “acts” or “eponymous” EU law are considered only those who formally and officially carry a name in their title. For the moment, this is accomplished by means of parentheses that follow the “normal” name of the statute in question. In fact, in these parentheses, the statute’s name is included and the text preceding them describes their subject-matter (see below, 2).

Under the above clarification it should be noted that EU legislative acts naming has undoubtedly not been unheard of in the past. For example, the ePrivacy Directive (or, more accurately, the “Directive on privacy and electronic communications“) has carried such a title since 2002. However, this by no means constituted either the rule or even a frequent occurrence. In their vast majority, even famous EU pieces of legislation only carried an informal title, given to them even by their law-makers themselves (e.g. the NIS Directive, the Water Framework Directive or the Industrial Emissions Directive).

  1. Titles of EU legislative acts

Titles are given to EU legislative acts according to the EU Interinstitutional Style Guide. According to its Section 2.1, “the complete title of an act comprises:

  • the type of act (regulation, directive, etc.),
  • the number (that is, the abbreviation(s) that apply (‘EU’, ‘Euratom’, ‘EU, Euratom’, ‘CFSP’), the year and the sequential number of the act),
  • the name of the author of the act,
  • the date of adoption (the date of signature for acts adopted jointly by the European Parliament and the Council),
  • the subject matter,
  • in acts with double numbering, the number assigned by the author (see Section 1.2.2 ‘Double numbering’)“.

Furthermore, “where the title of an act is amended by another act or is the subject of a corrigendum, the amended or corrected title should always be cited thereafter“.

Additional guidance on the titles of EU legislative acts is provided in the EU Joint Practical Guide (“JPG”). After clarifying that “the ‘title’ comprises all the information in the heading of the act which serves to identify it” (JPG, 7.1), the guide continues to suggest that “the title of an act shall give as succinct and full an indication as possible of the subject matter which does not mislead the reader as to the content of the enacting terms. Where appropriate, the full title of the act may be followed by a short title” (JPG, 8).

Consequently, within EU nomenclature it is the “short title” in the JPG that is of concern in this analysis. Nevertheless, the JPG is quite discouraging in this regard. According to section 8.4 JPG “a short title for an act is less useful in Union law — where acts are identified by a combination of letters and numbers (for example ‘(EU) 2015/35’) — than in systems which do not have such a system of numbering. In certain cases, however, a short title has come to be used in practice (for example, Regulation (EC) No 1234/2007 = ‘Single CMO Regulation’). Despite the fact that it may seem a simple solution, referring to acts by a short title creates risks for the accuracy and coherence of legal acts of the Union. This method should therefore only be used in specific cases where it significantly aids the reader’s understanding” (JPG, 8.4).

The JPG is even more condemning as regards short titles appearing on the title of an act: “The creation of a short title when an act is adopted by adding it after the title of the act should be avoided, since it only renders the title more cumbersome, without actually resolving the question of whether or not the short title should be used, either in the act which created it or in subsequent acts. While the risks outlined in point 8.4 must always be borne in mind, it is possible to refer to an act by using a short title in order to make it easier to understand the act in which the reference is made. In this case, the short title chosen will have to appear in brackets in the body of the text of the act in which the reference is made, like any other abbreviation” (JPG, 8.5). Notwithstanding the formalistic approach, that puts the cart in front of the horse, the fact is that the JPG’s request to include the short title in the body of the act and not on its title, has not been followed by any one of the abovementioned “famous” EU pieces of law (see 1), including(!) its own example (the “Single CMO Regulation”).

“Eponymous statutes”: A Common, rather than Roman, law practice?

If the JPG got one thing right as regards short titles on acts it is the fact that “they are less useful in Union law – where acts are identified by a combination of letters and numbers (for example ‘(EU) 2015/35’) — than in systems which do not have such a system of numbering”. Indeed, naming statutes does not seem to be a European legal trait at all. In many European states legislative acts are invariably numbered, usually in a sequential manner according to their publication in the government gazette, followed by their date of publication. While they also carry a title, this tends to be long and descriptive and is never followed by any short title. Consequently, reference to them within the same jurisdiction is usually made by using their number only. Exceptions are, perhaps, the German and Austrian legal systems. Here short titles in the type of an abbreviation are included in the title of acts (see e.g. the German Federal Data Protection Act, the BDSG), and therefore replace numbers when referring to them. Another exception, of course, is codified legislation. In Roman law systems, codes carry names, as most prominently demonstrable in the civil codes of EU Member States.

However, where short titles, or eponymous acts proliferate, are in Common law systems, most notably in the US. Although US law is indeed numbered and codified law, law-makers use short names for legislative acts extensively. It has been pointed out that“[the Congress] has named its handiwork either in the statute itself or in a later measure. These names may be descriptive […] or they may memorialize some individual. That individual may be a sponsor of the legislation, a beloved or respected congressional leader (living or dead), or a private citizen to whom the legislation in some way relates” (Strause R E B et all, p.11). This enthusiasm has ultimately led to the introduction of so-called Popular Name Tools, to translate the short titles of laws to actual codified legislation. How are US law-makers implementing these short titles? – Interestingly, by actually following the (EU’s) JPG recommendation (in 8.5). For example, Section 1(a) of the Sarbanes-Oxley Act specifically sets that “[t]his Act may be cited as the “Sarbanes-Oxley Act of 2002”.

Why name a statute at all?

The most obvious reply to this question would be to better remember it. Alphanumeric reference (e.g. “Regulation 2016/679” instead of the “GDPR”) is basically addressed to experts in the field. However, the law should not solely address a handful of experts. It should instead aim to address the public at large. Everyone needs to be aware of the law in force and have access to it. In other words, every piece of legislation, especially if aimed at directly affecting individuals’ lives, needs to be easily remembered and referred to in everyday life routine.

The EU’s JPG has expressly identified this need in that it finds that “drafters must […] consider what information should appear in the title in order to prompt a reader who is directly concerned […]” (JPG, 8.1). It is against this background that the EU’s inclination against short titles has to be questioned. Even more so where none of the “famous” EU legal acts mentioned above have followed the JPG’s advice. Although discouraged by the JPG, short titles have been frequently used and indeed achieved their purposes – as most demonstrably shown by the GDPR.

The use of short titles in EU law may also serve a different purpose – EU law penetration and the creation of a European public space. In the last few years we witnessed an increasingly direct impact of EU law on the everyday lives of Europeans, most famously through the GDPR. While EU law always had an impact on its citizens, the latter was, arguably, far less obvious to the individual. The GDPR however touches a central subject matter in the lives of citizens – data protection. It is therefore no surprise that citizens are well aware of this piece of legislation and refer to it by name. EU law might hence do well to continue the use of simplified titles, in an attempt to illustrate the importance of EU law for the individual, and to enable the citizen to refer to these EU laws that directly concern them in a simple and uniform manner. These citizen-friendly titles are likely to raise awareness and ultimately contribute to the creation of a European public space. An alphanumeric system could not possibly fulfil this mission.

Conclusion: A trend that needs to become the new rule?

Short titles for EU legislative acts are to be welcomed, even if the EU may ultimately need a US-like “Popular Name Tool”. Short titles are indispensable means to raise awareness among intended recipients of EU legislation. They could ultimately create a European public space, within which commonly and readily identifiable EU statutes are employed in their original names by all Europeans. Europeans should not be expected to use an alphanumeric system to refer to legal acts that directly affect their everyday lives. In this context, the EU JPG’s Section 8 should be amended in order to adopt – if not a positive then at least – a more neutral approach towards short titles in EU legislative acts. Even more so, where it has been shown in this contribution, that short titles for EU legislation already seem to be a reality anyways. Ultimately, the trend of “act-ification” of EU law serves a dissemination and awareness purpose, critical in assisting EU law to prove its ability to deliver immediate effect and protection of EU citizens.