Lessons from telecoms as regulators debate what to do with digital giants
Both the EU, in the form of the draft Digital Markets Act, and the UK, through the advice of the Digital Markets Taskforce, are considering applying ex-ante access regulation to digital platforms/gatekeepers. This could require the companies concerned to provide rivals with wholesale access to functionality of parts of their business on a regulated basis. For example, competitors might gain the ability to deliver applications to end users through access to app stores or real time access to customer data gathered by the platform.
Ex-ante economic regulation can be seen as having a dual role both:
- as a preventative measure, ensuring that companies with market power do not abuse their position, where ex-post competition law is not effective;
- and as an element of a proactive strategy to influence the future market structure by making parts of the value chain contestable.
While traditional ‘utility’ regulation has focused on the former objective, for example through retail price control, telecoms regulation has explicitly sought to introduce and facilitate competition where possible.
In many jurisdictions, access regulation has been a key tool in supporting the liberalisation of the fixed telecoms market from the starting position of a vertically integrated statutory monopoly to a market with a number of competing providers across the value chain.
Experience with access regulation in the telecoms sector has taught a number of lessons:
- To ensure access regulation is effective and proportionate, clarity is needed on the objectives of the regulation, the expected benefits and any costs and potential unintended consequences;
- There are significant practical difficulties in implementing access remedies even where the objectives are clear; and
- Economic regulation of the terms and conditions of access agreements is a complex and resource intensive task.
Keeping things in proportion
The first lesson from the telecoms experience is that access regulation may not be appropriate in all circumstances. For example, access regulation has not been introduced in mobile markets, with instead competition between vertically integrated operators developing in all jurisdictions. In a number of jurisdictions, such as the US, there have been policy decisions not to apply access regulation to fixed networks. Even in the EU, where access regulation is a key part of the common regulatory framework, some National Regulatory Authorities (NRAs) have chosen to withdraw existing access regulation in order to promote end-to-end (infrastructure-based) competition.
Under the EU telecoms regulatory framework, the market review process and the determination of Significant Market Power ensure that access regulation can be imposed only in a narrow set of circumstances. The framework explicitly requires NRAs to assess whether remedies, including access remedies, are proportionate and achieve the NRAs’ objectives. Decisions must be subject to consultation, and merits-based appeals of the decisions are available. These checks and balances ensure that access regulation is used sparingly.
The advice of the Competition and Markets Authority appears broadly in line with this experience, in that access regulation is proposed as one potential pro-competitive intervention following an assessment of the market. However, the European Commission’s draft DMA potentially requires all gatekeepers to offer some forms of access across the EU, as well as specific access to search data and application stores for relevant platforms, without any assessment of the impact of applying access regulation.
The second lesson is that establishing the technical parameters of access regulation is a long and complex process. First, a decision needs to be taken on the precise form of access. In the telecoms sector this can be informed by the market review process, which identifies the source of potential market failures and hence where in the network access needs to be mandated to address this failure.
Once the form of access has been defined, the technical implementation of access needs to be specified. Access agreements will, by their nature, be dense contracts between parties. For example, reference access offers in telecoms can run to hundreds of pages of technical and legal detail. These offers are generally developed under the auspices of industry working groups, including representatives of the regulators, the incumbents and access seekers.
In telecoms, regulators are also able to offer a quid pro quo to encourage incumbents to negotiate in good faith, since the market review procedure allows regulation to be removed from downstream markets as access regulation takes effect. So, for example, retail regulation was largely dispensed with across the EU following the implementation of effective access regulation. This provided a strong incentive for incumbents to implement effective access.
Access to internet platforms is likely to be more complicated because innovation in the underlying technology results in a moving target. Moreover, there is greater information asymmetry between the regulated company and the regulator or access seekers. The draft DMA also does not foresee a mechanism for deregulation when access regulation has resulted in effective competition, so there is little incentive for platforms to facilitate access.
Is it fair?
The final lesson is that even when the form of access is settled and technical issues have been resolved, determining and monitoring the terms and conditions of the access agreement to ensure access seekers can compete on an equal basis with the incumbent is a significant and ongoing task. Fair, reasonable and non-discriminatory (FRAND) conditions are well-meaning principles which are difficult to translate into clear-cut commercial terms based on economic principles.
Much of the initial focus on terms and conditions concerns pricing of access. The EU telecoms framework allows cost-based price regulation and permits NRAs to impose ancillary cost-accounting requirements in order to determine relevant costs. Even with this framework in place, establishing cost-oriented prices is a lengthy endeavour. NRAs have to exercise judgement when choosing between alternative methodologies and using complex models to determine and attribute costs. While costing methodologies were being developed, NRAs in many cases relied on benchmarking or ‘retail minus’ approaches in order to set access prices.
Setting access prices for internet platforms will be more challenging due to the difficulty of pinning down the relevant costs. Unlike national incumbent telecoms operators, the major internet players are global companies whose market position relies on a set of intangible assets developed through innovation in a competitive market. This raises theoretical issues in determining the appropriate compensation for investors, taking into account the need to incentivise innovation, and practical issues in attributing costs to access services. Even if the apparently weaker standard of ‘fair and reasonable’ is applied rather than ‘cost orientation’, these issues will need to be addressed. In addition, the short cut of using benchmarking or retail minus approaches to assess access prices is unlikely to be feasible given the lack of even approximate comparators for the platforms and retail equivalents to the regulated access services.
In telecoms, even where access prices have been determined, concerns have persisted about discrimination by incumbents between their own downstream divisions and competing access seekers. Cost-oriented prices offer some degree of protection, but non-price terms and conditions provide scope to discriminate, for example in quality of service or through cumbersome processes for access seekers. In order to address the potential for more subtle forms of discrimination, NRAs have in some cases ordered increased separation between the activities where the incumbent derives market power, such as access networks, and those where it competes with access seekers. The problem of policing non-discrimination is likely to be greater in dynamic and innovative internet markets, where the opportunity for differential treatment is greater and the level of transparency lower.
In telecoms markets, access-based regimes have been successful in allowing competition to develop in parts of the fixed value chain following liberalisation (arguably at the expense of deeper infrastructure-based competition).
However, ex-ante access regulation is not a quick fix. Effective implementation in telecoms has required significant resources for regulators and other stakeholders to specify the access offer, determine terms and conditions and monitor compliance with non-discrimination obligations. Implementation has been supported by a regulatory framework that has been devised to provide an incentive for the incumbent to engage in order to benefit from deregulation in downstream markets.
Designing and implementing effective access regulation would appear to be more challenging in digital markets. In part this reflects the complexity and dynamism of the sector but also the lack of experience globally in ex-ante economic regulation of these markets. In addition, regulating companies which have built strong market positions through a process of innovation in competitive markets raises knotty economic issues which did not arise when imposing access regulation on former statutory monopolies.
The framework, timescales and resources identified in the EC’s proposed DMA appear to significantly underestimate the challenges that the Commission would face in implementing and enforcing effective access remedies. The greater flexibility and discretion that the CMA suggests would be appropriate for a UK framework appear more realistic, but that does not diminish the challenges the CMA would face if it were to decide on regulated access as a remedy for a market failure.