Written evidence submitted by Ofcom (RUT0311)

The state of fixed broadband in the UK

Key findings on fixed broadband

• 89% of UK premises (25.5m) can receive superfast broadband speeds (defined as having a download speed of at least 30 Mbit/s)
• But less than 2% of UK premises get ultrafast services based on full fibre networks.
• Overall 1.4 million UK premises (5% of the total) cannot get a download speed of 10 Mbit/s. Ofcom has previously published research that demonstrates that a download speed of 10Mbit/s is currently sufficient to allow multiple users to simultaneously use the internet, including web browsing, video streaming, video calling and gaming - but we acknowledge that this minimum will need to increase over time.
• Availability varies substantially between urban and rural areas, and between UK nations. For example, 59% of rural UK premises (2.3m) can receive superfast broadband speeds. 25% of the UK’s rural premises (approximately 960,00) cannot get a download speed of 10 Mbit/s and the average download speed in the UK’s rural areas is 21Mbit/s.

Source: Ofcom Connected Nations 2016

Please note that the UK Government and Broadband Delivery UK define superfast broadband as having download speeds of 24Mbit/s or higher. Connected Nations 2016 reported coverage of services at these speeds in June last year was 90%.

Availability of superfast broadband (30Mbit/s) in rural areas, by premises (%)

                                                                                                                                                            Source: Ofcom Connected Nations 2016

Premises unable to receive a download speed of at least 10Mbit/s (%)

                                                                            Source: Ofcom Connected Nations 2016

Why are broadband speeds lower in rural areas?

• Lower population density and more challenging terrain make the rollout of superfast broadband a significant challenge across many rural areas. These conditions make it difficult for companies to provide a service at a reasonable cost per connection. To address this, the UK and devolved governments are providing public investment to further the rollout of superfast broadband.  Such schemes include Broadband Delivery UK and the Digital Scotland programme.
• One of the main reasons rural properties are less likely to receive decent broadband speeds is the distance between premises and the nearest telephone cabinet or exchange. The resistance of copper wire increases with the length of the wire, so speeds drop as the distance between the premises and the exchange or cabinet increases.
• Many properties in rural areas are on exchange-only lines, where premises are connected by copper wire directly to the nearest exchange, sometimes a great distance away.
• For fibre-to-the-cabinet (FTTC) installations, where fibre optic cables run to the nearest cabinet and a copper line runs from the cabinet to the premises, the distance covered by copper can still have a big impact on speeds. However, even those who live too far from the cabinet to receive superfast broadband will still benefit from the upgrade at the cabinet, as the reduction in the length of the copper wire will still improve their broadband speeds.
• Ofcom wants everyone to share in the benefits of a modern digital society. Our starting principle is that no one should be left behind. A broadband universal service obligation is one potential key tool in ensuring people can receive decent broadband services.

Ofcom technical advice to Government on a broadband universal service obligation

• The UK Government set out its intention to introduce a broadband universal service obligation (USO) in November 2015.  The Government said that its ambition was for a download speed of 10Mbit/s to be available to all on reasonable request.
• The current legislative framework for a broadband USO derives from the Universal Service Directive. The provisions of the Directive have been implemented in the UK through the Communications Act 2003 which sets out the process for the imposition of universal service obligations on providers. The Digital Economy Bill proposes to amend the Communications Act 2003 to make express provision for a broadband USO.
• In March 2016 the Government commissioned Ofcom to provide technical analysis and recommendations to inform its decisions on factors such as speed, eligibility, affordability and funding.
• Ofcom published its advice to Government in December 2016. Ofcom’s report examines how the provision of a USO could work in practice. It also considers how the specification of the USO could affect both the number of premises that are eligible and the costs that could result from meeting a variety of technical specifications.
• Ofcom set out three technical specifications for the Government’s consideration:
• (1) a standard broadband service (10 Mbit/s download)
• (2) a more highly specified standard broadband service (10 Mbit/s download + 1 Mbit/s upload)
• (3) a superfast broadband service (30 Mbit/s + 6 Mbit/s upload)

    Number of premises and costs for each scenario

         Please note that the 2017 forecast of the number of premises and the costs is based on expected future  

              roll-out of publicly-funded programmes and commercial roll-out.

• In designing any intervention, we highlighted that Government might want to consider the extent to which it should be designed to take into account further future growth in broadband usage. For example, in 2004 Ofcom defined broadband as 128kbit/s, sufficient for real time delivery of e-mail. By 2010 our focus had moved to video, so the Universal Service Commitment was set at 2Mbit/s. In 2013 we noted that a typical household, with multiple simultaneous users, required 10Mbit/s and last year evidence started to emerge that households with connections above 40Mbit/s consume significantly more data.
• In addition to the technical specification the Government asked us to consider affordability and pricing.  In our technical advice, we found that current broadband pricing does not prevent most people from taking it up. However, there are issues of affordability for a small proportion of consumers and so, given the policy aim of social, economic and digital inclusion, a social tariff for those consumers is likely to be appropriate.
• The Government also asked Ofcom to consider reasonable requests and cost thresholds. The cost-per-premises of delivering decent broadband to the very hardest-to-reach premises could be very high. Premises in the final 1% have an average cost that ranges from £2,780 per connection for standard broadband to £3,350 for superfast broadband. Those in the final 0.5% can cost between £4,460 and £5,100. The cost of serving the most expensive premises is estimated to be around £45,000 in all three of our scenarios. This could support the need to introduce a reasonable cost threshold to limit the upper bound of the costs. Policies on universal availability often include such limitations. Our technical advice considers the effect of different reasonable cost thresholds. Where the actual cost exceeds the threshold, we recommended that some form of backstop solution (e.g. satellite broadband) should be made available.
• Few providers would be capable of effectively providing a USO, and none of the operators have so far expressed interest in providing a broadband USO. The most efficient outcome for any broadband universal service may be to make use of, and build upon, existing scale communications networks. This would imply BT may be best placed to be a designated provider for the UK, and KCOM for Hull. This is because existing UK network structures mean that the majority of eligible premises will be connected to BT’s copper network or will live in areas where BT’s network is present, except in Hull, where KCOM has an extensive presence. We may also not consider it proportionate to impose the obligation on smaller providers unless they have volunteered.
• We recommended the use of eligibility criteria to ensure that a USO provider does not overbuild existing commercial / community broadband networks.  This could be done by limiting eligibility to those premises where there is no network capable of offering the USO technical specification (or more). 
• The Government indicated that its preference was for a USO to be funded by an industry levy. In Ofcom’s advice we took a cautious approach to estimating the potential effects on consumer bills, assuming an industry fund is implemented and that the costs of industry’s fund contributions are fully passed through to customers. Given competition in retail broadband services and the need for any industry fund to be competitively neutral in its design, it is reasonable to expect that at least a proportion of an increase in providers’ underlying costs would be reflected in consumer prices.
• The actual amount paid by consumers will depend on a number of factors, including the period over which costs are recovered and which companies contribute. For example, assuming a seven-year cost recovery period and costs recovered from fixed broadband providers, on average consumers might see increases in household bills ranging from just under £11 per annum to deliver a standard broadband universal service to just under £20 per annum to deliver a superfast broadband universal service. Including mobile operators in any industry fund might reduce the average effect on household bills by just under 15%. Whether and how to actually pass on any costs to customers would be a commercial decision for contributors. This analysis is therefore illustrative only.

• The Digital Economy Bill, currently before Parliament, contains provisions enabling the Government to create a USO for broadband. Once Government has decided the specification of any USO, Ofcom’s main role would be one of implementation, for example designating the universal service provider(s).

Further Information

March 2017

3