This paper focuses on the relationship between EU Competition Law and ex ante regulation by specifically analyzing the case of regulatory sandboxes. It also provides some case studies of regulatory sandboxes and questions their impact of on the goals of promoting competition and innovation in the EU.
By Cristina Poncibò & Laura Zoboli[1]
I. INTRODUCTION
The European Commission has adopted a clear strategy to promote competition in the digital services markets through a form of asymmetric, sector-specific regulation.[2] In particular, it has recently adopted the Regulation of the European Parliament and of the Council for fair and contestable markets in the digital sector or Digital Market Act (“DMA”).[3] With this legislation, the EU has clearly expressed a certain diffidence towards the laissez-faire approach in general and, in particular, towards the emergence of the digital economy, as the absence of public regulatory action was too risky in the long run for the proper functioning of the single market.[4]
However, the choice to intervene through ex ante regulation also has some clear limitations with respect to the speed of technological innovation, as well as the risks of creating a burdensome and fragmented regulatory framework that ends up increasing in complexity and in implementation costs for companies and, ultimately, for citizens themselves.[5]
The decision to resort to such an articulated form of regulation, which is difficult to enforce in the EU member states, reflects the mistrust of the European institutions towards the effectiveness of competition law vis-à-vis digital platform operators and, in particular, large platforms (precisely: gatekeepers). Now, the aforementioned mistrust is rather worrying when one only considers the tumultuous development of new technologies ranging from blockchain to artificial intelligence; in other words, public authorities will soon also be called upon to grapple with the new digital service markets linked to the emergence of such innovations.
If competition law is actually ineffective, the fundamental error of the public actors with respect to the governance of the digital economy may be found in the rather long time lapse between the emergence of these new economic powers – the large platforms – and the ability of the authorities themselves to understand and appropriately regulate these realities, taking into account their particular nature and the associated risks (think, for example, of the protection of users’ personal data). In other words, it would have been preferable to intervene when this phenomenon came to the attention of markets and society. However, public authorities did not have the tools and methodologies at that time to ensure a dynamic approach and, thus, a prompt response also through the application of European competition policy. A reflection of this nature is particularly important not so much for the past, but for the near future where, as mentioned, public authorities will soon be confronted with new technology-related markets.
That said, this contribution considers a different model of regulation from the traditional regulatory one, which also includes the DMA. We are interested here in considering, in particular, those new pro-competitive regulatory strategies that are characterized by an experimental nature (regulatory experimentalism).
Precisely, the article dwells on regulatory sandboxes, i.e. an experimental approach aimed at public regulation of markets and, even more interesting for our purposes, at promoting competition in the new digital markets. It should also be pointed out that the term regulatory sandbox itself should be understood as a general term implying different mechanisms depending on the jurisdiction. These mechanisms are, however, united by their experimental nature and the mentoring function of public authorities with respect to the companies participating in the experiment.[6]
II. REGULATORY SANDBOXES AND COMPETITION POLICY
It should be noted how the growing interest in promoting competition policy through ex ante regulation also emerges in relation to the case of regulatory sandboxes.[7] A form of competition policy that, operating when a new market is born, would like to contribute to establishing the rules of the game (so to speak), or rather the regulatory framework that may well include economic as well as social objectives, such as sustainability. In this way, regulatory sandboxes could be used by public authorities to prevent the emergence of strong economic powers in digital markets and the consequent creation of barriers to entry.
Now, competition law scholars have not yet investigated the relationship between competition law and regulatory sandboxes, whereas public authorities seem to believe that sandboxes can generally foster both innovation and competition in fast-moving digital markets. For instance, the Financial Conduct Authority (“FCA”) established a regulatory sandbox in 2015 in order to promote effective competition in digital financial services markets.[8] Specifically, the regulatory sandbox should enable the FCA to collaborate with innovators, ensuring consumer welfare and promoting competition in financial services for small and medium-sized enterprises.
More generally, the CGAP (Consultative Group to Assist the Poor) and World Bank (2019) study on regulatory sandboxes[9] , identifies competition policy as one of the objectives of regulatory sandboxes.
Source: CGAP-World Bank study (2019), Motivations driving implementation of innovations facilitators.
In spite of the confidence expressed by public authorities in the capacity of these experimental tools to contribute to the regulation of digital services markets, it should be pointed out that empirical research is still at an early stage and that it will take time to obtain results and, therefore, to make an accurate assessment of the potential of these innovative regulatory tools.[10] In truth, regulatory sandboxes have been very positively received by national authorities, as they allow for a revisiting of the proportionality principle of European law, leaving more room for flexibility and activism with respect to innovation. In relation to our case, it is also possible to detect a sort of competition between systems, in the sense that the national authorities seem to be interested in competing in the search for the most up-to-date and promising methods of regulation.
As noted above, following the British example, several jurisdictions have chosen to create innovation-friendly sandboxes for companies and start-ups. It should come as no surprise, therefore, that innovation markets in digital services (especially in financial services) take center stage for the experimentation of these new tools. In this context, it is of paramount importance to distinguish what is truly innovative in the practice of digital market regulation from mere announcements that are often aimed at promoting a national regulation as the most favorable for companies wishing to establish their registered office and offer newly developed digital services.
III. REGULATORY SANDBOXES: CASE STUDIES
As mentioned, regulatory sandboxes have gained importance in the fintech sector, playing a crucial role in understanding how to regulate those technology applications in the financial sector with which the regulator was not yet fully familiar.
A. Financial Conduct Authority
The pioneer of regulatory sandboxes in Europe and what can be called the benchmark model in this field was the one launched by the FCA in the UK in 2016. The rapid rise of the financial technology sector and the resulting regulation (sandbox) in the UK led to a new methodology that, according to proponents, should ensure competition and consumer welfare and mitigate market risks, while encouraging the innovation needed by both market participants and consumers.
A key objective of financial market regulation should be to promote competition on the merits, ensuring that firms must comply with the same rules and bear the same costs. In this form of regulation, innovation can occur when firms seek to distinguish themselves from their competitors, rather than simply identifying a gap in existing regulation, which is often subsequently filled. [11]
Although the success of regulatory sandboxes is closely linked to the financial (in particular: fintech) and banking sectors, this model of experimental regulation is finding application in a very wide range of markets, including transport (e.g. drones, autonomous vehicles), energy, health, to name but a few.[12] In general, public authorities are trying to overcome the inherent limitations of traditional regulation. The EU institutions themselves recently launched a pan-European regulatory sandbox for innovative use cases involving Distributed Ledger Technologies, aimed at addressing sensitive issues such as data portability, inter-company data exchange, digital identity, and smart contracts.[13]
B. Artificial Intelligence
Equally interesting is the European Commission’s framework regulatory sandbox in the field of artificial intelligence. Precisely, this form of experimental regulation is governed by Title V of the Artificial Intelligence Act entitled “measures in support of innovation,” which encourages the competent national authorities to create spaces for regulatory experimentation and defines a basic framework in terms of governance, control and accountability.[14] These regulatory “testing spaces” for artificial intelligence would be aimed at creating a controlled environment to test such innovative technologies for a limited period of time on the basis of a program agreed with the competent authorities.[15] And interestingly, such a sandbox should in no way exempt participants from the obligation to comply with existing EU regulations, including the Data Protection Regulation.[16] This choice is indeed puzzling, as it risks depriving this instrument of one of its main features, as well as of a fundamental incentive for companies to participate.
As mentioned, each jurisdiction follows different paths in the creation of regulatory sandboxes to the extent that the very term regulatory sandbox should be correctly understood as a general definition that may imply very different experimental realities. More generally, it is evident that jurisdictions increasingly apply a trial-and-error process in order to ascertain what best suits the regulatory and business environment of each state. This naturally reinforces the differentiations between jurisdictions even in the EU. In this regard, it is possible to identify a number of recurring regulatory sandbox models in different legal systems – with respect to which observations can be made under the lens of competition protection.
First, we must mention the traditional (or standard) sandboxes that are accessible to companies that have certain requirements and may be interested in testing new digital services in a controlled environment. These are distinguished from sandboxes, which are applied on a case-by-case basis and in the presence of a few selected companies.[17]
Moreover, depending on the regulatory sandbox at issue, there are cases where the regulatory framework for sandbox participants remains completely unchanged, as compliance rules are not relaxed by public authorities during the trial period (as is the case under the Artificial Intelligence Act). This usually happens, for instance, when the sphere of personal data protection is concerned.[18] This helps to limit fears of favorable treatment for those who participate in the sandbox and consequent unfavorable treatment for those who are excluded. As will be seen below, this approach may reduce the risk of anti-competitive effects resulting from the creation of a regulatory sandbox.
On the other hand, the second recurring model sees regulatory sandboxes operating in a context where the authority can relax the application of certain rules. Once again, the example of this approach is the FCA’s regulatory sandbox,[19] which provides, at least theoretically, for the possibility to waive some of the existing rules in the course of experimentation. On the contrary, an actual relaxation of the rules can be seen in the sandbox used in the Netherlands in order to test innovative solutions in the energy field. In particular, the Dutch public authority authorized the energy cooperatives and associations admitted to the sandbox to deviate from the national energy regulations in at least some respects.[20]
IV. HELLENIC COMPETITION COMMISSION
The regulatory sandbox proposed by the Greek competition authority is a third and interesting example of the relationship between experimental regulation and competition law.
In particular, the Hellenic Competition Commission (“HCC”) has set up a sandbox in order to promote the goals of sustainability and competition in the Greek market.[21] In particular, the sustainability sandbox proposed by the HCC has applications in sectors of the economy, e.g. energy, recycling and waste management, industrial production of consumer products, food production and distribution, pharmaceuticals, to name but a few. Precisely, the sandbox under consideration would offer companies the opportunity to test new services that can promote sustainable development without significantly impeding competition in the relevant market.
In this way, the Greek authority would like to support small and medium-sized Greek companies that may be lagging in innovation and in the adoption of new sustainable technologies, by offering these private entities the opportunity to test innovative solutions in a controlled environment and with the collaboration of the authority itself. In such a model, companies may submit their proposals to the HCC and these proposals may also refer to the opportunity to allow agreements between competitors (horizontal agreements) or within supply chains (vertical agreements) in order to promote the public objectives mentioned above. Needless to say, such an approach could abstractly lead to a kind of derogation from the application of competition law in the concrete case in order to favor or balance other objectives. For the sake of completeness, it should be mentioned that such exemptions could also concern unilateral conduct on the part of undertakings that could constitute an abuse of a dominant position.
The latter case would indeed constitute an exceptional hypothesis according to the guidelines of the Greek authority itself. It should also be pointed out that the aforementioned sandbox of the HCC is specifically aimed at the most complex cases, in which it might be necessary to scrutinize the restriction of competition in order to achieve a certain objective linked, first and foremost, to the priority of the issue of sustainable development.
Once the companies have submitted their proposals, the HCC will assess their content and take into consideration existing European and national competition law, as well as the case law on the assessment and inclusion of public interests in the application of Art. 101(1) and (3) TFEU and the corresponding Greek framework. We feel it is important to emphasize that in this context the HCC should, for example, consider different criteria, indicators and performance keys concerning processes that are directed towards sustainable development goals.[22] After assessing the proposal, the authority could consider not applying the European or national competition law framework in the case at hand, or reject the request, considering the existing regulatory framework to be applicable.
In addition, the Greek authority would have to provide undertakings with guidelines to clarify the timing of the regulatory experimentation and the limits of such an exemption; and could also set certain conditions to be met by undertakings participating in the sandbox. The same authority proposes to constantly monitor the implementation of the experiment and to organize regular meetings on the progress of the initiative. Finally, participating companies should come out of the sandbox with a sort of “sustainability license” that would end up representing, in any case, a novelty of considerable relevance in the EU regulatory landscape.
Having framed the Greek initiative, is worth questioning whether the HCC is reverting to the previous practice of notifying competition authorities about agreements between companies, which has now been replaced by self-assessment undertaken by the firms themselves and, more concretely, whether the HCC has the financial and professional resources to take charge of this process, given that this area of practice is undoubtedly a novelty for national competition authorities.
V. PRELIMINARY CONCLUSIONS
The article analyzed the relationship between competition law and experimental regulation, taking the case of regulatory sandboxes.
What emerges, then, is a very fragmented picture of the sandbox phenomenon, in which each experiment differs in its objectives and methods. However, the above analysis allows for some considerations on the relationship between regulatory sandboxes and competition law.
It is, first of all, interesting to note how the aforementioned experimental approach ends up representing an area of convergence between competition policy and regulatory requirements. This also implies overcoming the traditional distinction mentioned above between mechanisms that operate ex ante and instruments that are instead only active a posteriori.[23] In this sense, it is true that regulation determines the field of action of competition law and, conversely, the latter influences the field of evolution of regulation. It is clear, therefore, that the aforementioned regulatory sandboxes can also be understood as a form of co-evolution of the two disciplines that is characterized by a kind of overcoming of the traditional dichotomy mentioned above. To clarify, they can overcome the traditional (di)vision of the work between regulation and competition, and, in this way, sandboxes can contribute to changing our perspective from a mere and unresolved issue of competence towards a new form of competition policy. The latter would apply from the moment public authorities define the rules of the game in the markets for new digital services for businesses and end consumers.
And again, while welcoming the attempt to develop new methodologies, even based on experiments, we must conclude that, at least in the EU, the shift towards an ex ante perspective will once again favor – as in the case of traditional regulatory regulation, such as the recent DMA – the role of the regulatory state. In other words, it seems interesting to underline how innovation, and specifically digitalization, is once again fueling the expansion of the fields in which public power is inevitably called upon to exercise itself. Yet, this model inevitably seems to be somewhat outdated and unconvincing with respect to the great changes taking place on the technological and social fronts, with the risk that the DMA will fail to achieve the results expected by the European institutions when it is applied in practice.
In other words, there is an expansion of the area of public regulation in nascent and innovative markets. At the same time, European and national competition authorities may be able to extend their scope of action through experimental mechanisms to include, as in the case of the Greek competition authority, far-reaching objectives, such as checking the sustainability of certain products and services.
The analysis also allows us to see how the objective of creating competitive digital markets through regulatory sandboxes is based on a kind of revisitation of the doctrine of experimentalist governance.[24] Well, experiments are not necessarily born equal: the structure of the sandbox itself may influence its proper functioning. Some actors may reap considerable benefits, while others may incur considerable costs. In short, it must be made clear that experimentation is not merely a technical activity, as it implies renewing our faith in the role of the regulatory state.
Our contribution confirms that a regulatory sandbox can produce both benefits and risks with respect to competition policy. On the one hand, it helps to improve collaboration between actors and the learning capacity of public authorities with respect to new emerging services and technologies. Thus, the fact that regulatory sandboxes represent regulatory regimes with reduced legal risks may also favor the evolution of such regimes in maintaining competitive market conditions. Moreover, regulatory sandboxes, if properly designed by supervisory authorities, could well create a level playing field for new entrants and mitigate barriers to entry. On the other hand, however, regulatory experimentalism may also exacerbate risks for both consumers and competition, as well as introduce some critical aspects related to the peculiarities of such mechanisms.
Last, it is possible to consider regulatory sandboxes as an alternative or, in any case, a complementary tool to traditional regulation, e.g. the DMA. The challenges of the present and near future call for a profound rethinking of the very nature of competition policy. In this context, the article highlights how the use of experimental regulation may contribute reshaping the application of competition law in the relevant markets, shifting the focus towards a kind of anticipatory competition policy, leading competition authorities to the forefront of market governance and assigning them greater responsibilities with respect to nascent markets.
[1] Associate professor at the University of Turin & Researcher at the University of Brescia, respectively.
[2] O. Borgogno & G. Colangelo, Data, Innovation and Competition in Finance: The Case of the Access to Account Rule in European Business Law Review, 2020, 31, pp. 573-610.
[3] Regulation (EU) 2022/1925 of the European Parliament and of the Council of 14 September 2022 on fair and contestable markets in the digital area and amending Directives (EU) 2019/1937 and (EU) 2020/1828 (Digital Markets Regulation or Digital Services Act), in OJ L 265, 12.10.2022, pp. 1-66.
[4] R.E. Matland, Synthesizing the Implementation Literature: The Ambiguity-Conflict Model of Policy Implementation, in Journal of Public Administration Research and Theory, 1995, 5(2), pp. 145-174.
[5] W. A. Kaal, Dynamic Regulation for Innovation, in M. Fenwick et al. eds., Perspectives in Law, Business & Innovation, Springer, New York, 2016.
[6] EMA, Regulatory Sandbox, 2020, https://www.ema.gov.sg/Sandbox.aspx.
[7] J. Kálmán, Ex Ante Regulation? The Legal Nature of the Regulatory Sandboxes or How to Regulate before Regulation Even Exists (November 21, 2022) http://dx.doi.org/10.2139/ssrn.3255850.
[8] FCA, “Regulatory sandbox lessons learned report” (2017) https://www.fca.org.uk/publications/research/regulatory-sandbox-lessons-learned-report.
[9] CGAP-World Bank: Regulatory Sandbox Global Survey (2019), https://www.findevgateway.org/sites/default/files/publication_files/surevy_results_ppt_cgap_wbg_final_20190722_final.pdf.
[10] For example, as pointed out by D. Arner, J. Barberis & R. P. Buckley (“Fintech, Regtech and the Reconceptualisation of Financial Regulation” 37 Northwestern Journal of International Law and Business (2017), 373-385), the first sandbox experience in the UK covered only a very small fraction of the total number of financial services firms, a significant number of which are now in liquidation or insolvent.
[11] L. Bromberg, A. Godwin & I. Ramsay, “Fintech Sandboxes: Achieving a Balance between Regulation and Innovation,” 28 Journal of Banking and Finance Law and Practice (2017), 314-336.
[12] Council Conclusions on Regulatory sandboxes and experimentation clauses as tools for an innovation-friendly, future-proof and resilient regulatory framework that masters disruptive challenges in the digital age, November 16, 2020, https://data.consilium.europa.eu/doc/document/ST-13026-2020-INIT/en/pdf.
[13] European Commission, “Launch of the European Blockchain Regulatory Sandbox,” February 14, 2023, https://digital-strategy.ec.europa.eu/en/news/launch-european-blockchain-regulatory-sandbox.
[14] Proposal for a Regulation of the European Parliament and of the Council laying down harmonised rules on artificial intelligence (Artificial Intelligence Act) and amending certain Union legislative acts, {SEC (2021) 167 final}.
[15] Regulatory sandboxes in the field of artificial intelligence are also mentioned in: Council of EU, “Conclusions on Regulatory sandboxes and experimentation clauses as tools for an innovation-friendly, future-proof and resilient regulatory framework that masters disruptive challenges in the digital age” (2020); Commission, “EU Coordinated Plan on AI” (2018 and its 2021 review); EU Parliament “Resolution of 12.02.2019”; G20, “Ministerial Statement on Trade and Digital Economy” (2019).
[16] Regulation (EU) 2016/679 (General Data Protection Regulation), in OJ L 119, 04.05.2016.
[17] Jurisdictions where customised regulatory relaxation is permitted include the State of Arizona (U.S.), Brunei, Canada, Hong Kong, Indonesia, Malaysia, and Singapore, see D. M. Ahern, “Regulators Nurturing FinTech Innovation: Global Evolution of the Regulatory Sandbox as Opportunity Based Regulation” 60 European Banking Institute Working Paper Series (2020).
[18] See, for instance, Datatilsynet (Norwegian Data Protection Authority), “Sandbox for responsible artificial intelligence” (2021) ; CNIL (Commission nationale de l’informatique et des libertés), “Bac à sable données personnelles” (2021).
[19] FCA (Financial Conduct Authority), “Regulatory Sandbox” (2022) https://www.fca.org.uk/firms/innovation/regulatory-sandbox.
[20] E.C. van der Waal, A.M. Das, T. van der Schoor, Participatory Experimentation with Energy Law: Digging into a Regulatory Sandbox for Local Energy Initiatives in the Netherlands, in Energies, 2020, 13(2), 458, https://doi.org/10.3390/en13020458.
[21] Hellenic Competition Commission, “Competition Law and Sustainability” (2021). This mechanism was introduced following a public consultation by the authority. See Hellenic Competition Commission, “Public consultation: Proposal for the creation of a sandbox for sustainability and competition in the Greek market” (2021); see also the relevant press release of July 13, 2021.
[22] For example, the objectives of the Paris Agreement on Climate Change (C.P. Carlarne, “Balancing Equity and Effectiveness: The Paris Agreement & The Future of International Climate Change Law,” Ohio State Public Law Working Paper No. 477 (2019)) or the goals of the sustainable development agenda (see “Transforming our world: the 2030 Agenda for Sustainable Development,” UNGA Res 70/1 (September 25, 2015).
[23] D. A. Zetzsche et al, Regulating a Revolution: From Regulatory Sandboxes to Smart Regulation, in Fordham J. Corp. & Fin. L.,2017, 31, p. 23.
[24] C. F. Sabel, J. Zeitlin, Jonathan, Learning from Difference: The New Architecture of Experimentalist Governance in the European Union, in Sabel and Zeitlin (eds), Experimentalist Governance in the European Union: Towards a New Architecture, Oxford: Oxford University Press, 2010, pp. 1-28. On this point, see also Y. Svetiev, Experimentalist Competition Law and the Regulation of Markets, Hart, Oxford, 2020.
