The UK’s mobile sector is highly competitive. Ofcom, the telecoms regulator, has spent the last few decades carefully designing spectrum auctions and other measures to foster fierce competition among the four mobile network operators. I’m currently paying £10 ($12.70) per month for unlimited 5G data on the best network for coverage. When I was at university, I was paying over £60 in today’s prices. This allowed me to send 15 text messages per month, make a limited number of voice calls, and use no data whatsoever on a 2G network.
Now the CMA has issued its phase 1 decision in a merger case that would reduce the number of operators below four for the first time since Orange and One2One entered the market in the early 1990s. The current third and fourth operators in terms of subscribers, Vodafone (a UK-listed company) and Three (owned by Hong Kong-based Hutchison) argue that they are currently sub-scale and they need to merge. The deal would make the combined company the largest operator, with around half of the available spectrum and a market share approaching 40% of subscribers in a concentrated market.
The CMA has referred the case for an in-depth phase 2 investigation. Assuming that no phase 1 remedy will be offered or accepted, a final report will be due in Autumn. The CMA panel at phase 2 has a big decision on its plate. Last time a 4-to-3 telecoms merger was attempted, the CMA loudly opposed Hutchison’s proposed acquisition of O2. This was before the UK’s exit from the EU. It was the European Commission’s decision to take, but the CMA’s view was influential. The Commission blocked the deal in May 2016, was overturned on appeal at the General Court, and then won ultimately in the Court of Justice last year.
Although the Commission’s decision to block the deal was vindicated, it is fair to say that the market has changed since then. Even if we accept the Commission’s decision was correct, the likely outcome in the present case is not obvious. It seems commonly accepted that scale matters more than it used to, and four-operator markets are struggling to yield sustainable returns. The current situation looks sub-optimal and the UK is nowhere near the top of international league tables for 5G rollout.
Maybe my monthly subscription is actually too cheap? That is not an easy argument to make to a competition authority.
Mobile networks are important for the UK economy. They are essential for businesses to grow, and they are a vital market for consumers like groceries, energy and water. This case is arguably more important than the much-discussed Microsoft/Activision which only affected the tiny market for cloud gaming. This case has some arcane issues that are specific to telecoms markets (e.g. network sharing, spectrum holdings and mobile virtual network operators), but there are other issues that are of wider interest to merger control policy more generally. The CMA will need to decide where it stands on the trade-offs between competition and economies of scale. This case will test the CMA’s treatment of efficiencies arguments to an extent not seen since its increased interventionism in mergers began in earnest around five years ago. The CMA will also be careful to take account of Ofcom’s views.
The CMA would normally be expected to block 4-to-3 mergers nowadays, especially where there is little chance of new entry. But, is this case different? This blog post picks some of the interesting issues that seem likely to play a large role in this case.
Post-merger price rises (or quality degradation)
I assume the CMA will conduct its standard economic analysis of post-merger pricing incentives, which combine estimates of the merging parties’ profit margins with “diversion ratios” that estimate the proportion of customers who view the two parties as each other’s closest competitor. In a market with four close competitors, the CMA’s analysis will always produce a significantly positive number. In other words, the CMA will say that the merging parties will have a strong incentive to raise prices post-merger. The CMA’s phase 1 decision also says that Three is the lowest priced competitor, with an incentive to compete strongly on price, meaning that its removal could be especially damaging for competition.
The pricing analysis is only a static measure, but it puts the parties in a difficult position. They will need to prove to the CMA that their widely publicised merger efficiencies will in fact materialise and will be sufficient to outweigh the upward pricing pressure. Importantly, the parties will also have to prove that they have an incentive to pass on the benefit of these efficiencies to customers. It is rare that parties manage to persuade the CMA that their efficiencies are indeed “rivalry-enhancing” (in the language of the CMA’s merger assessment guidelines). In this case, the companies have been preparing for the CMA investigation with their lawyers and economists for several years, so we must assume they think they have a good argument. It is ominous for them that the CMA’s phase 1 decision already sounds sceptical of their arguments, but the phase 2 team will now look into this issue in more detail.
The merging parties may also argue that prices are currently extremely low, consumer expectations about network quality are rising quickly, and the UK is not currently high in the international league table for 5G coverage. Therefore, higher prices with higher quality would be a better equilibrium. That might be right, but the CMA would say that sufficient competition between operators is needed in order to force operators to achieve that high quality. And in the meantime, consumers would be paying more.
Network sharing
Building a mobile network is an expensive thing to do nowadays. Indeed, that cost is essentially the main reason for the proposed merger. The four networks currently pair up under two network sharing agreements, MBNL and Cornerstone, and therefore share some of the infrastructure costs. EE and Three are the partners in MBNL; Vodafone and O2 are the partners in Cornerstone. The proposed merger would therefore straddle the two network sharing agreements and significantly disrupt the delicate balance that currently exists.
There was a similar issue when the Commission investigated the previous 4-to-3 merger in the UK, Three/O2. One wonders what Vodafone said to the Commission about this aspect at the time, and whether it has now needed to change its story.
The Commission was concerned that Three and O2 would have had a full view of the network plans of the other remaining competitors. It felt that Three and O2’s presence in both networks would weaken EE and Vodafone as competitors and hamper the future development of mobile infrastructure in the UK, particularly with respect to the then-anticipated roll-out of 5G. By pursuing a strategy to degrade the MBNL network or delay its development, Three and O2 would have the ability to harm their largest competitor. This was seen as a plausible outcome in the eyes of the Commission.
The Commission also felt that the behavioural remedies offered by the parties would have been difficult to implement and monitor effectively. The CMA is famously wary of behavioural remedies, so it is difficult to imagine it being more open to this kind of remedy than the Commission was in 2016.
It is difficult to see how the analysis of network sharing would be significantly different eight years on, despite the nature of the networking sharing agreements having changed. The joint venture partners’ assets are less intertwined in the Vodafone/Three case than was the case in the Three/O2 example because they share less of the active equipment, but there is still a large amount of shared work. The merger would still produce incentives to frustrate the MBNL joint venture.
It is not therefore surprising that the CMA’s phase 1 decision highlights the ability of the merged company to disrupt the effective functioning of the network sharing agreements. It notes the awkwardness of the merged company gaining access to both of its competitors’ commercially sensitive information. It also raises the concern that the merged company will be able to predict its competitors’ strategic moves, and being able to compete less aggressively as a result. These issues would be difficult to solve through contractual arrangements as the partners would still need to co-operate with each other in agreeing future investments. It is difficult to see how the network sharing regime works with only three operators.
Spectrum
It seems strange that the CMA’s phase 1 decision summary does not mention spectrum, but it still seems likely to play a large part in the final outcome at phase 2. At the moment, the four mobile networks have fairly similar levels of the available capacity. Of course, this is not an accident. Ofcom carefully designs spectrum auctions in order to foster competition between them. Ofcom has produced a situation where the operators are not completely symmetrical such that coordinated effects become more likely, but they are also not so asymmetrical that competing with the leader is hopeless.
If the merger goes ahead, the combined Vodafone-Three would hold around 60% of the C-band spectrum that is considered valuable for 5G networks. It would have over half (54%) of all useable low band spectrum which is, for example, over four times the amount of spectrum held by EE (13%) that is considered valuable for 5G networks including indoor coverage and innovative 5G solutions such as the Internet of Things and connected cars.
It would be a similar story for infrastructure sites. The four operators currently have around 18,000 to 20,000 sites each. Post-merger, Vodafone-Three would have 36,000 sites, although it seems to intend to reduce that to around 26,000 sites.
The current balance would therefore be disrupted. The parties argue that this will give the other players a stronger incentive to invest to catch up, and the presence of only three operators means that they will each have the scale to do so. The parties will also say they have a strong incentive to act as a wholesale player for mobile virtual network operators (MVNOs) who piggyback on the four networks and often serve specific customer niches. They will argue that Vodafone and Three customers will receive a merger benefit from Day 1 because the two networks can use each other’s capacity.
The other two networks, EE and O2, would eventually be able to build more infrastructure if necessary (subject to the UK’s restrictive planning laws), but they cannot do anything about an in-built spectrum disadvantage. They would face a significant disadvantage in both spectrum and the number of sites, so there is a risk that competing with the merged company is hopeless.
The CMA will want to carefully consider whether the merged Vodafone-Three can disincentivise their competitors’ investment with the threat of huge spectrum deployments in response to their investments. Perhaps the large spectrum holding makes competing with the merged company almost hopeless in the medium term. The spectrum issue seems more of a problem than traditional horizontal theories of harm relating to the accumulation of a large market share for customers.
National security
The UK has a regime for screening mergers that may raise national security issues. It seems that Vodafone has some sensitive contracts with the British intelligence agencies, although the exact details are difficult to discern. It also has some less sensitive contracts with the Government. The merger therefore potentially raises issues because Hutchison is Hong Kong based and therefore arguably subject to the Chinese Government’s control.
The UK has mandated that Huawei’s equipment be stripped out of the mobile networks by the end of 2027 at huge expense. Much as I would prefer the Chinese Government not to read my conversations with my wife about what time the kids need picking up, it would be strange to treat these as more sensitive than Vodafone’s top-secret contracts.
Whether this issue will be important in the eyes of the Government, and whether it can be solved through commitments, is another key issue for this case.
Conclusion
It may be true that scale matters in modern telecoms markets, and it may be true that the weaker network operators are not making sustainable returns. However, that does not necessarily mean this merger will be cleared.
When the CMA hears arguments from companies and investors that they are not making enough money, that may sound to them like vigorous competition is working its magic. The CMA’s remit does not include worrying about investor returns, unless it reaches a point where companies are failing to invest and innovate. The key question is how consumers will be best served in the medium and long term. Will the proposed merger help or hinder the development of the connectivity that consumers want and the economy needs?
The merging parties will need some bullet-proof data on their merger efficiencies. They will need some very good answers on issues such as network sharing and spectrum hoarding. Arguments about what happened after telecoms mergers in other countries such as Italy, Ireland, Germany and the US have been traded back and forth for a long time, but the present case will ultimately be decided on its own facts.
It is also notable that the merging parties seem to be treading a careful line on their future intentions. The CMA case would be easier to win if Three could credibly claim to be exiting the market. Some commentators have argued they will inevitably exit at some point, but the company itself seems unwilling or unable to make that argument. For its part, the CMA has said that both parties are “currently viable and competitive businesses and that they would continue to invest in their networks absent the Merger”.
The merging parties should prepare for a rough ride in phase 2.
The post Vodafone/Three: UK telecoms consolidation will be given a rough ride through the merger control process appeared first on The Platform Law Blog.