UK Broadband Policy: Why Both Tories and Labour are wrong

With Brexit, the UK might soon be the first country to ever impose economic sanctions on itself. In line with this current trend of economic masochism are both Johnson’s and Corbyn’s plans for full fibre broadband. Recently, both parties have been outbidding each other on this topic, promising from the Tories’ full fibre broadband by 2025, through the well-known benevolence of market forces and from Labour free full-fibre by 2030, funded by the taxpayer.

At the current rate a free subscription to Sky or Netflix might also be up for grabs. Both plans are unrealistic and will likely end up with large parts of the British population, especially in rural areas, left-out or a huge dent in the public purse.

The Tories and Ofcom’s undying belief in market efficiency:

The Tories, and to an extent Ofcom, believe that by unleashing market forces with light touch regulation and a £5bn subsidy package for 2m premises in the most rural (and most expensive areas to cover), the UK can be fibered up by 2025. Johnson and policymakers are right to believe that most of the UK can be covered through private investment and that other remote parts of the country, where building a fibre network is unprofitable, will require public subsidy. The problem with his plan is timing. The UK would have to ramp up its fibre construction rate from, to more than 5m premises a year. No other country in Europe has achieved this feat, apart from Spain, where at the time, unemployment was high, labour costs low and building regulation lax. Additionally, a hard Brexit with a salary cap, would affect the European migrant workforce needed to build these networks and likely create further delays.

The second problem with this plan is that market forces alone are unlikely to provide fibre coverage in areas where only one fibre network is likely to be profitable. By promoting infrastructure competition Ofcom and DCMS have created a vibrant fibre eco-system, but which has mainly focused on the low-hanging fruit. Although the UK has just passed the 10% fibre coverage landmark, ‘altnets’ and even the incumbent Openreach are still cherry-picking urban areas or medium density areas with low build costs and attractive demographics. Investors, with the current policy and regulatory frameworks, are weary of funding fibre projects in rural areas where they might be overbuilt.

Ofcom and DCMS have both recognized that in roughly 20% to 40% of the country’s premises only one network is likely to be viable. By offering subsidies to less than 5% it is recognizing the need for a de-facto monopoly in these areas but leaving the other 15% to 35% in a regulatory vacuum that discourages investment from new players. They are also proposing to favour Openreach in these areas through a regulated asset base model. Not only is this in contradiction with Ofcom’s own aim of promoting competition but in the current state of affairs, UK fibre deployments, may end-up looking much more like a postcode lottery than a comprehensive national infrastructure project.

Finally, Ofcom’s view on the benefits of having multiple competing fibre infrastructures in the same location is also questionable: Passive infrastructure fibre can easily be upgraded to support more bandwidth – today’s 1Gbps fibre networks can easily be upgraded to 10Gbps, without digging up any trenches. Innovation comes from the active equipment at either side of the network rather than the passive assets that are largely built to be future proof. Most people and policymakers would view creating parallel gas, electricity or water networks serving the same premises as madness. Why should broadband be any different?

Labour’s digital white elephant:
Although Corbyn’s British Broadband plan recognizes the fact that market forces alone might not be sufficient to provide fibre to all, it is likely to be disastrous for the public purse. Labour’s stated cost of £20Bn for full fibre coverage is way off current estimates, including BT’s. They are putting the cost at £30Bn to £40Bn plus the differed cost of losing out on Openreach’s current £5bn in annual revenues and the need to compensate private players like Cityfibre, FibreNation and the dozens of other altnets which have already began deploying their networks. Costs could also be driven up by the fact that a large national behemoth like British Broadband may not allocate capital as efficiently as private players and is also likely to slow down deployment rather than accelerate it. If in any doubt just look at what happened with Australia’s white elephant, the National Broadband Network, which is massively behind schedule, overbudget and which might require a huge write-off in the coming years - $13Bn is the figure currently estimated.

The second argument against Corbyn’s plan is more political. 95% of UK households already have access to speeds of 24Mbps or above which cover most basic internet browsing needs as well as streaming video, music, etc.… Using taxpayer money to provide free ultrafast fibre broadband is a bit like upgrading UK water infrastructure to provide free mountain sourced mineral water. Although fibre is a superior and cheaper infrastructure which will eventually replace copper, consumers and operators should make the choice of paying for it - not government.
The French and competition in the market:

France, curiously, provides a useful example where policy and regulation has led to an efficient use of public funds, wider broadband deployment, lower prices, innovative financing and continued investment.
By splitting the country into three distinct zones, very dense areas (VDA - 18% of premises), medium density areas (MDA - 35% of premises) and low density areas (LDA- 47% of premises), government and the telecoms regulator, ARCEP have created a tiered system with different rules that maximize private investment in each area.

In urban areas where costs per premise passed range between 200€ to 300€, operators are free to build as many competing networks as they’d like, with the only obligation of mutualizing the last connection point within a building. Although France has 4 major operators with FttH infrastructure and several smaller players, the reality is that only 2 networks were built in these areas as Free / Orange and SFR / Bouygues decided to team up to avoid excessive build costs.

In medium density areas, where costs range between 250€ to 500€, only two operators expressed interest in building non-overlapping fibre networks. The market was split 80:20 between Orange and SFR. These zones are de-facto monopolies – without subsidies - where anyone is free to overbuild. As each of the two networks is obliged to offer its services on the other and offer open access to other ISPs at a fair and non-discriminatory price the incentives to overbuild are highly limited.

Finally, in low-density areas, local authorities were given power to create their own networks or as is more often the case 20-year concessions. These were in partnership with private players, creating de-facto but not de-jure monopolies. The result is that €3.3bn in subsidies are being used to cover ALL 17m premises in rural areas. This in a country with a population density less than half of the UK’s. Private operators using subsidies are not allowed to cherry pick in the designated tender area. The project is currently both ahead of schedule and will be delivered below budget and the average subsidy request in the latest tenders has dropped from 50% of total project CAPEX to less than 10% - a rare feat for a semi-public infrastructure project.

So, without imitating the French, what lessons could the UK learn from this model?

• In countries where regulation offers a stable framework favoring an open-access wholesale model on a single network and thus high take-up in areas where only one network is likely to be viable, private investors are ready to accept lower returns and forego cherry-picking strategies.
• Without a well-organized competitive tendering process the cost for local authorities and government is likely to be much higher than needed.
• The stated initial roll-out costs of operators are often exaggerated (especially when subsidies are at stake) compared to those ultimately observed (strong learning effects, economies of scale, risk premiums, ...)
• Subsidies or favorable regulation like concessions should come at a trade-off for operators in the form of penalties for delays and coverage obligations.
• Interoperability between networks is key. With at least 26 alt-nets building fibre network in the UK, consolidation is inevitable. Setting standards ex ante will allow for an orderly process
• The profitability of these networks will eventually depend on the major ISPs' willingness to switch from copper to fiber. As such, an open and non-discriminatory wholesale model is the most suitable for fostering competition between different ISPs and lowering prices for consumers.
• Large incumbent operators have often deployed fibre in reaction to altnets. The later they do so, the more network and retail market share they will lose, and the greater the social and employment consequences.

We are convinced that the thriving alt-net sector along with Virgin and Openreach’s latest plans prove that there is no lack of private capital or ambition to fibre up the UK rapidly. However, the upcoming election’s victor, will hopefully choose a sensible path that recognizes that in many non-urban parts of the UK pushing for competing networks supplying the same service is folly. To have a hope of reaching the 2025 target and using public money efficiently the UK needs to create strong regulatory certainty in these areas to avoid overbuild. If in doubt, it can look at its own policy document from 2018, the Future Telecoms Infrastructure Review, which puts the price tag of a partial franchise model at £20.3Bn, 30% cheaper than the current enhanced competition model it is following.

Reef Read is a senior Manager at PMP's Telecom Infrastructure Practice, which provides commercial, strategy and regulatory advice on major telecom deals for investment funds and operators. PMP has offices in Paris, London, Montraal, Morroco and Brussels