This piece is authored by Richard Feasey, Senior Advisor. The views expressed in this blogpost are attributable only to the author in a personal capacity and not to any institution with which they are associated. In addition, they do not necessarily correspond to those of CERRE.
The long awaited and much debated Digital Networks Act proposals have finally been published by the Commission[1]. In this blogpost, I offer a short initial assessment of the contents, comparing them to the proposals in the CERRE report ‘Ideas for Future European Telecommunications Regulations’ which was published in September 2024[2]. I highlight the proposals (and reference the relevant Articles) that I think will be most significant for the future performance of the European telecommunications industry were they to be implemented as proposed. The views expressed are therefore mine alone and not to be attributed to CERRE or to any co-authors of the CERRE report[3].
Evolution not revolution
In the CERRE report we concluded: ‘we do not consider that the evidence of industry performance or market outcomes presented above suggests that a wholesale reversal of the approach to competition and regulation taken since 2002 is required or justified’[4]. We were more concerned about other forms of regulation, including spectrum auctions and measures to promote resilience, which introduced unbudgeted and unavoidable costs into the industry.[5] I think the latest proposals from the Commission are broadly consistent with our assessment and the most significant changes be proposed relate not to economic regulation but to spectrum policy.
The regulation of fixed telecommunications networks and services, for which the SMP regime is highly relevant, is to remain broadly unchanged. In 2024 we examined various proposals to replace or improve the existing SMP regime[6] but the Commission has, rightly in my view, decided against radical changes (Articles 72-6). The Commission clarifies that new fibre networks will be obliged to provide access to their in-home wiring even if they do not have SMP (Article 71), which I think was already implicit under the existing Code. The retention of the SMP and symmetric access regimes does not mean that the way in which NRAs apply those regimes will not need to change as the market itself evolves. For example, we anticipated in the CERRE report that a much more granular approach to defining markets would be required and that new entrant fibre networks are likely to find themselves regulated as a result. In line with this the Commission proposes that NRAs be required to collect more detailed market data to support their market analysis (Articles 183(2)& 183(12)).
On copper switch off we concluded ‘a common European target or deadline will not assist those operators able to retire networks well before that date and will not be met if no alternative fibre network is available at the location in question’[7]. The Commission proposes a deadline, but 2035 is unlikely to have much impact (Article 54). Although there is a threat that operators who fail to meet the deadline could be fined by NRAs I think that threat is unlikely to accelerate fibre deployment, induce operators to deploy fibre in areas where they judge it is not economic to do so or otherwise have any other impact (Article 61). Most operators will have decommissioned most or all of their copper networks well before the deadline in any event.
The proposals do not directly address a key issue we highlighted in the CERRE report, namely the wide differences in performance between different Member States within Europe[8]. It is not clear to me how or whether this can be addressed by changes to the law, as opposed to other actions. Greater harmonisation of regulation may be achieved by Commission proposals to extend its veto powers over NRA decision-making in certain areas (notably for symmetric regulation in fixed markets (Article 69(4)) and in spectrum licensing (Article 31&32)) and by the proposal to adopt a Regulation rather than a Directive. Assuming the aim and effect of adopting a Regulation is to force a greater harmonisation of regulation than is the case today, it remains to be seen how much this reduces the wide differences in market performance and outcomes that we highlighted. The aim must be to get the rest of Europe performing as Nordic telecommunications markets do but I am not sure whether the proposals yet offer the means of achieving this.
Positive changes in spectrum policy
In the CERRE report we were very concerned about policies that had led the industry to incur costs that were unexpected and unavoidable. We identified spectrum costs as the most significant sources and recommended that in future operators be allowed to retain spectrum under very long duration or perpetual licences, as the Commission proposes, (Articles 24 & 25) but on the understanding that they must ‘use it or lose it’[9]. We emphasised that this would require significant changes in the way NRAs apply spectrum policy, including a much greater capacity to monitor and assess how spectrum was being used and much less reliance on traditional market mechanisms such as auctions or trading. The proposal for a new database (Article 28) may help NRAs and other potential users how well existing holders of spectrum are utilising it, but I think something more will be required if the risk of revocation is really to encourage better utilisation. My concern here is that the Commission is relying on third party requests to share spectrum to trigger an investigation into whether it is being utilised (Article 27). In my view NRAs will need to be more pro-active in monitoring spectrum usage than that.
We also highlighted a tendency for the scope of regulation to expand in other directions and by other routes whilst the SMP regime was at the same time performing its intended function of reducing regulation as competition advanced[10]. An example of this has been some NRAs’ use of spectrum licensing as a way of imposing regulations on mobile operators which would otherwise be prevented by the SMP regime. The Commission now proposes to close this loophole as part of a set of measures (including veto powers) which are intended to harmonise and limit the contents of spectrum licences (Article 31and 32(4)). This is welcome but appears to apply only to new licences, so it will take time to have a significant impact. More generally, the Commission is proposing to acquire much greater influence over spectrum policy than it is has exercised in the past, including exclusive jurisdiction over the licensing of satellite spectrum (which I think has some justification) (Articles 17-19, 29-32, 38-43). I am supportive of these proposals but I expect Member States to be very wary.
I also note the strong presumption in the proposals that spectrum ought to be shared rather than assigned on an exclusive basis, as it has been to mobile operators (Articles 15 & 27). I am not sure how much this, together with a stronger hand by the Commission, will impact the telecommunications industry. I suspect each assignment will continue to be decided on the merits. Similarly, how prices for spectrum will be set is not really addressed in the proposals (Articles 25(4) and 29) but will be the subject of a separate recommendation within 12 months of adoption. If Member States fail to follow that recommendation then the Commission proposes powers to allow it to set prices itself by means of an implementing act (Article 29(4)).
More symbolism than substance on pan-European markets
The Commission and others have long advocated measures that are intended to support cross-border consolidation of networks and the provision of telecommunications services on a pan-European basis. The latest proposals contain a number of measures which might be said to contribute to this aim. These include a simplified authorisation regime with passporting to allow an operator in one Member State to provide services throughout the Union without further administrative burdens (Article 10). It also includes pan-European numbering ranges (Article 47) and a standardised wholesale broadband access product that would be available from any SMP operator in any Member State (Article 81). Greater harmonisation in the implementation of regulation generally (see above) may also assist. However, I do not expect the structure of European telecommunications markets, or the services within them, to be much affected by any of these.
Simplification and streamlining of the general authorisation regime is a welcome measure and will lower entry barriers a little but have minimal impact for firms already in the market. Pan-European numbering has been tried before without success. The standardised wholesale broadband access product is attractive in theory but it is not clear whether or how far it can be achieved given the differences in network technology and topology around Europe. Importantly, I would expect implementation of these measures to involve significant costs both for the industry and NRAs. I would therefore want to first be confident that there is meaningful real world market demand for the things they are intended to enable. Otherwise there is a risk that these measures, like the other ‘single market’ measures we have seen in the past (caps on the prices of international calls and the Roaming Regulation) impose an additional regulatory burden of European operators which their counterparts elsewhere in the world do not have to bear.
One aspect of regulation where I might have expected the Commission to go further but it has instead gone in the opposite direction is universal service. The specification of the minimum internet service to which all citizens in a Member State are entitled i left to NRAs to define, (Article 88), whilst the Commission has been quite willing in recent years to set coverage and availability targets for the entire Union. Perhaps the difference is explained by the fact that one is an aspirational target and the other is an enforceable legal obligation. I note that the Commission proposes to give itself powers to address these issues using implementing acts (Article 89) and appears to anticipate undertaking a more fundamental review of universal service (as I think it should) by 2034 (Article 92(3)). Other aspects of the proposals also appear to give NRAs greater leeway in deciding how any universal service obligations should be funded than in the past (Article 90) but the costs for the industry of funding universal service provision in Europe have never been all that significant in practice.
Additional costs for the industry and work for NRAs
I am concerned that implementing the changes properly will involve a significant cost and effort for both the industry and NRAs. An example of this, anticipated in the CERRE report, are actions on network resilience and the Union Preparedness Plan which BEREC is expected to develop and, whilst recommendations, the industry is nonetheless expected to pay utmost regard to (Articles 5 and 6).[11] The key issue we identified in the CERRE report remains how substantial the additional costs to improve resilience will be and how they are to be funded.[12] Since the proposals do not reference public funding I assume that these costs will be borne by the industry.
Similarly, changes to the authorisations regime (Articles 9&10), a fundamentally different approach to spectrum licensing and monitoring (Articles 15-32), the introduction of European numbering (Article 47) and standardised wholesale products (Article 81), copper decommissioning and fibre transition plans (Article 56&60), extending contractual protections to microbusinesses and non-profits (Article 99), anti-fraud measures (Article 103) and requirements on NRAs to collect more granular network, spectrum and market data (Articles 28&183) will all mean significant workloads and costs for both operators and NRAs. I support of some of these proposals and others may be justified on their individual merits, but it is very important that their proponents recognise the additional costs – including opportunity costs – which they will impose on the industry (and NRAs) if they are to be implemented properly.
At the same time, I have not identified any significant reduction in consumer protection or other similar regulations, as we proposed in the CERRE Report[13]. I recognise that some articles that were introduced into the Code but rarely applied, such as those relating to co-investment and separation measures, have been removed and the Commission may reasonably argue that greater harmonisation should reduce the overall regulatory burden implied by those obligations that remain (and that for many operators most of the costs of complying with consumer protection regulation have already been sunk). Nonetheless, for all the talk of simplification, the Digital Networks Act will remain a vast and complex piece of legislation by which the Commission still seeks to micromanage the industry.
Kicking ‘fair share’ down the road?
Much of the debate in recent years has focussed on claims by European operators that Amercian content providers such as Netflix or YouTube were not contributing a ‘fair share’ to the costs that telecommunications operators incurred in deploying and operating their broadband networks and should be obliged, through regulation, to do so.. There are no such proposals in the Act[14].
Instead the Commission proposes measures to contribute to what it calls ‘eco-system co-ordination’(Article 191). This echoes, but does not go as far as adopting, an earlier proposal from Vodafone for a Framework for Responsible Use to be agreed between content providers and operators[15]. I am not clear why BEREC is well placed to produce guidelines on these issues or how it is expected to ‘promote co-operation’. Constructive dialogue between content providers and network operators is obviously desirable but my understanding is that it already occurs. NRAs will also be required to conciliate in individual disputes if they arise between firms (Article 192) although this appears to me to amount to littlemore than organising meetings.
It will apparent that I am deeply sceptical about this aspect of the Commission’s proposals. It is difficult to predict what impact these changes will have but I think they will encourage some operators to manufacture disputes to test the regime or demonstrate that it has ‘failed’ in some way. In the past NRAs and BEREC have not shown much appetite to intervene in these matters and I think the proposals put them in an impossible position, including exposing them and not the Commission to any geo-political risk. The proposals include a threat that the Commission may revisit these arrangements after 3 years of operation, presumably to consider something else if they are judged to be ineffective (Article 193). What impact this will have on the conduct of the firms in the meantime is also unclear. Whether the Commission thinks its proposals will actually help anybody or has decided to defer any conflict with American content providers until the geo-political environment changes is unclear to me.
On the other hand, the Commission has not taken the opportunity to revisit the Open Internet Regulation, other than allowing it to specify how ‘specialised services’ are to be defined (Articles 93&94). We had suggested in 2024 that changes be made to ‘include the possibility of providers of internet access services being able to charge on a commercial basis upstream content or internet service providers for differentiated connectivity services that improve the experience for the user’.[16] The Commission may argue there is sufficient room for interpretation in the existing text to ensure that innovation is not impeded in any way and that its proposals would give it powers to ensure this. I can only hope they are right.
Changes in appeal standard?
One change that might pass unnoticed is the removal of an explicit right to an appeal of any NRA decision expressly ‘on the merits’ (Article 197) and a requirement only that it be ‘effective’. Whether (and in which Member States) this amounts to a raising of the bar for challenging decisions taken by NRAs I leave to lawyers to advise upon, but I think it is a potentially important aspect of the proposals which should be of concern to operators.
Conclusion
Overall, I think the proposals represent a significant improvement on the rather muddled thinking of the Commission’s earlier White Paper[17]. I remain concerned about the costs of implementing some of the proposals, particularly the suite of pan-European or single market measures which the Commission attaches importance to but which I think will have little practical benefit for either industry or users. The changes to spectrum policy are radical – as they need to be, whilst the maintenance of SMP regulation and other aspects of fixed market regulation seems to reflect a proper recognition by the Commission that fixed telecommunications markets are already performing reasonably well in Europe. Operators should be concerned about any weakening of their rights to challenge decisions by NRAs
I think the Commission is on much weaker ground when it comes to ‘eco system’ co-ordination and Open Internet regulation. I worry that the former will encourage unnecessary and time consuming disputes and put NRAs in an invidious position when they could be devoting their attention to things that will have positive impacts.
[1] https://digital-strategy.ec.europa.eu/en/library/proposal-regulation-digital-networks-act-dna
[2] https://cerre.eu/wp-content/uploads/2024/09/CERRE_Ideas-for-the-Future-of-European-Telecommunications-Regulations.pdf. A longer report containing work by various other co-authors was also published but references here are to the summary paper which I largely authored.
[3] I was Vodafone’s Director of Policy between 2001 and 2013 and so have many similar assessments in the past. I left the industry in 2013 and my only interest in the Digital Networks Act is as Senior Adviser and co-author at CERRE.
[4] Ibid p.25
[5] Ibid p.25
[6] Ibid p.31-35
[7] Op cit p.38
[8] Ibid p.25
[9] Ibid p.37
[10] Ibid p.13
[11] Ibid p.47-8
[12] On 20 January 2026 the Commission also announced the Cybersecurity Act 2, which is intended, amongst other things, to allow the Commission to prohibit and/or require the removal of equipment from high risk vendors and for competent authorities to impose financial and other penalties in the event of non-compliance.
[13] Op cit p.41-2
[14] There were some differences in view amongst the authors of the CERRE report so we did not comment at length on this issue, but were critical of the Open Internet Regulation, see p.42/3
[15] https://www.vodafone.com/news/newsroom/public-policy/responsible-use-of-networks
[16] Op cit p.42
[17] https://digital-strategy.ec.europa.eu/en/library/white-paper-how-master-europes-digital-infrastructure-needs
