Long-term investment in high-speed broadband is crucial if the UK wants to secure a leading role in the global digital economy and keep its tech sector buoyant, writes Dan Lewis, the IoD’s senior adviser on infrastructure policy
• The UK suffers from some of the slowest speeds for internet and mobile of any developed major nation.
• We are now moving into a new era where speed, as well as reliability, will determine Britain’s place in the future digital economy.
• The UK runs the immense risk of falling further behind while its burgeoning tech sector decamps elsewhere.
• For mobile coverage, the UK still lags far behind in rural connectivity, with particularly poor reception indoors.
• Just 46 per cent of the countryside is able to access 4G.
• Data consumption is growing at around 50 per cent or more per year.
• While prices in the UK are relatively low by international standards, they do appear to have been rising – compared to other major European nations.
• There are two successful examples that the UK should pay close attention to when considering future broadband growth and investment: Lithuania and New Zealand.
• Britain must have ambition, vision and longer-term infrastructure planning.
• Business needs higher priority given to faster networks and competition for services to the workplace.
• Above all, we need new investors who seek a longer-term return beyond the typical three- to five-year telecom company investment cycles.
The UK may be the number one internet economy in the G20 – accounting for 12.4 per cent of GDP – yet it has among the slowest speeds for internet and mobile of any developed major nation, as well as some of the worst network coverage.
On downtime alone, the cost is substantial – estimated at £11bn per year. We are moving into a new era where speed – in upload, download and latency (the delay that happens in data communication) – as well as reliability will determine Britain’s place in the future digital economy.
But without a substantial shift – from short-term incremental improvement to long-term future-proof progress – the UK runs the risk of falling further behind while its burgeoning tech sector decamps elsewhere.
Britain’s telecommunications network and resulting broadband and connectivity woes did not appear overnight. They are the result of nearly two centuries of evolution and the compound distortion of decisions made long ago.
Starting in the 19th century under Disraeli, one by one, all the mediums of communication were nationalised: the telegraph companies (1869), the radio spectrum (1904) and, in 1912, the Telephone Company.
The long-term impact that reverberates today was the creation of a single dominant player, with no market competition to hold in check underlying costs or drive the requirement for efficiency and innovation.
The British Telecommunications Act (1981) was the first time in over 100 years that any government had seen fit to roll back the monopoly powers of the Post Office and set in motion a near-continuous period of demonopolisation that continues to this day.
Fast-forward to today and the greatest issue at stake for Britain is the 2015 Ofcom strategic review of digital communications, which will set the framework for the next 10 years and is very much dominated by the future of Openreach, the functionally separate infrastructure arm of BT.
So how does it all work, what are the connectivity chokepoints and how do we compare internationally? Turn the page to find out more…
Where are the broadband chokepoints?
Large rural areas – about 1.5 million premises, or 48 per cent of all rural premises, are not even able to access broadband speeds of 10 megabits per second (Mbits/s). For mobile coverage, the UK still lags behind in rural connectivity in 2, 3 and 4G with particularly poor reception indoors. Just 46 per cent of the countryside is able to access 4G.
Not-spots and city centres – a not-spot is an area which isn’t covered by a fixed or mobile network. In reality, most not-spots are partial not-spots as they tend to have one or the other. As well as remote corners of rural Britain having no mobile or broadband coverage, partial not-spots can also occur in city centres, such as the City of London and Westminster.
Suboptimal competition – the early years of Oftel were geared to creating competition where there was none. While Ofcom has done much in the intervening years to open up previously closed markets, BT’s market share appears to be increasing.
Poor knowledge about the copper network – BT does not have a clear record on the quality, positioning and connectivity of the copper network at the local level. The quality of the copper wire varies widely, depending on whether it is pure copper, a copper aluminium alloy, copper-coated aluminium or even just aluminium, all of variable thickness and of differing lengths to the premises, with a corresponding impact on the throughput of data.
This makes it hard for third parties to know what level of service they can offer customers. Similarly, BT offers a service of marking footways, to show where pipes and cables of other utilities are buried and at what depth to prevent other operators disrupting them.
However, a lack of proper record-keeping in the early days of the laying of cables and pipes means that there is even sometimes uncertainty about which side of a road they are located on, leading to unnecessary expense and time spent in digging up tarmac and pavements.
Limited knowledge about the whole network – unlike the power, gas and water grids that need to match capacity and demand in real time with prices, there are great unknowns within the telecoms network about quality, reliability and consistency of service. Actual Experience, co-founded by Dave Page and Professor Jonathan Pitts, has a cloud platform app that measures and captures the digital chain, identifying weak points in the internet network that result in sudden changes of speed or dropouts.
Data consumption is growing at around 50 per cent or more per year. These are the main drivers:
• Faster broadband speeds lead to more demand for internet use. Between 2014 and 2015, the average broadband speed increased from 23 to 28Mbits/s and average monthly data usage rose from 58 to 82GB. Equally for mobile, growth in data consumption has been just as pronounced, driven by rising penetration of 3G and 4G and an expansion of public WiFi hotspots.
• More consumption of video. Faster internet speeds have enabled the use of video on demand services, which are the highest consumers of data.
Today internet-connected TVs present a choice of standard definition or high definition. In the near future, there will be more widespread use of ultra-high definition. This matters because the higher the definition, the greater the bandwidth requirement and the slower the download time.
There are potentially several wildcards, however, which suggest that these forecasts may be a conservative underestimate. Immersive 360-degree virtual reality – headsets will be available for the first time this year – will require greater bandwidth than any flat screen.
At the other end of the scale, there is also massive potential for the many billions of new battery-powered wireless sensors making up the Internet of Things which will be connected to the web – sending out new packets and volumes of data that did not exist before, and to where the internet and mobile network
did not reach.
There are many competing and complementary broadband and connectivity technologies available or soon to be ready.
Fibre to the premises (FTTP). This is for now the highest-performing internet connection. By bringing a fibre cable into the home or office, symmetrical upload and download speeds become available at 1Gbit/s, with low latency rates. However, by comparison with other European countries, the penetration of FTTP in the UK is very low – 0.003 per cent1.
Satellite broadband. Many Britons have been receiving data from satellites since the late 1980s. Satellite broadband runs on the same principle but with higher bandwidth and can serve remote, poorly connected locations at much lower cost, more quickly. All of today’s broadband satellites operate in geostationary orbit at a height of 22,000 miles, which creates longer latency – 250 microseconds or more. However, download speeds are improving. ViaSat3 from 2020 will offer 1Gbit/s and largely uncapped data. It is well established in the US, with 700,000 subscribers but only a few thousand in the UK.
Fibre to the cabinet (FTTC). This brings fibre-optic cable between the cabinet (often a roadside box) and the exchange for a high-bit-rate digital subscriber line (VDSL). This is the superfast option being deployed across the UK by Openreach and
can deliver speeds at up to 76Mbits/s depending on the distance to
Fibre to the remote node (FTTrN). This essentially moves the fibre one step closer to the premises by connecting from the cabinet to the telegraph pole and so should be slightly faster than FTTC.
G.Fast or fibre to the distribution point. BT is trialling G.Fast, which promises download speeds of 300-500 Mbits/s and upload speeds of around a tenth of that with upgraded or new cabinets. XG.Fast pledges to deliver speeds of up to 5Gbits/s.
Hybrid fibre-coaxial cable using Docsis (a telecommunications standard that allows high-bandwidth data transfer to an existing cable TV). This is fibre-optic cable linked to a proprietary street cabinet, followed by a coaxial cabinet to the home. This is what Virgin Media has and it offers download speeds of up to 200Mbits/s, with the latest DOCSIS 3 upgrade.
Asymmetric digital subscriber lines and ADSL2+. In the UK, these are the standard broadband services and offer speeds from 0.5 to 24Mbits/s.
Aerial fibre. To avoid the costs of digging trenches in pavements and to the premises, in some cases it may be possible to deliver FTTP wound around existing telegraph poles
Line-of-sight or fixed-wireless access. Line-of-sight broadband connections essentially place a transmitting tower on top of a hill and relay either satellite or mobile wireless connections to anywhere that is within range and has line of sight of the connection.
Mobile wireless. In the UK, Relish has led the way with this offering, using 3G and 4G signals only for data to deliver 50-60 and up to 700Mbits/s to
a router in areas such as central London or rural not-spots that are poorly served by broadband.
Mobile broadband – 3G, 4G, 5G. The average download speed for mobile today is 6.1Mbits/s rising to 15.1 Mbits/s for 4G 2 ; 5G promises to have a speed of at least 1Gbit/s and may be available from 2020.
A superfast future?
In December 2010, the government published Britain’s Superfast Broadband Future, with the aim of the UK having the best superfast broadband network in Europe by 2015. So did the government achieve that?
Based on data compiled by Analysys Mason for BT3 , it appears that the UK achieved its aim of having the best network by 2015. But the wider picture is more complex and it does not mean that the UK could not have done better.
First of all, a national average of download speed would be a much more important measure, if the UK had an equal geographic distribution of population and economic activity.
However, it doesn’t. Very high bandwidth in remote corners of Britain will never have the same dynamic impact as in a concentrated bustling city centre. So it is concerning that London – Europe’s biggest city by economic output – has such a poor ranking compared to other European capitals, ranking 26th, just above Minsk 4 .
What can we learn from abroad?
All countries are different and work with varying legacy telecoms structures, regulatory environments, players and standards of living. However, there are two successful examples that the UK should pay close attention to when considering future broadband growth and investment.
Lithuania has the third-fastest connections in Europe. This is no small achievement considering it has a GDP per head around a third of the UK’s and a fifth of the population density. Nonetheless it is the leader in FTTH [fibre to the home] across Europe, with 35 per cent5 of households served and over 100 ISPs competing for business for 1.2 million households6.
How can it do this? In 2004, Lithuania’s equivalent of Ofcom, RRT, mandated the compulsory sharing of all passive infrastructure “…suitable for construction of electronic communications networks”. To further incentivise network investment, RTT went for low-cost access to ducts and poles.
Together, these measures led to a rapid build-out, predominantly (61 per cent) by the altnets (smaller network operators) in strong competition with the national incumbent, Teo, which had to make further investments to keep up7.
In Lithuania, the typical price per metre per month of duct access is €0.028 whereas in the UK it is €0.078. The bottom line is that competition and very low access costs gave Lithuania additional network investment and much faster broadband than would have happened if the incumbent’s privileges had been left in place.
Across the UK, many telecoms companies – apart from Virgin Media – have been calling for Ofcom to refer BT to the Competition and Markets Authority in order to split off BT Openreach as a separate company.
This has worked well in New Zealand. In 2011, overseen by the regulator and approved by shareholders, Telecom NZ was separated into two companies – Chorus and Telecom New Zealand, which became Spark New Zealand. Chorus, the Kiwi equivalent of Openreach, is responsible for the network infrastructure and Spark provides internet, mobile and fixed-line telephone services. Crucially, Chorus was spun-off, not sold off.
This meant that existing shareholders were able to see an increase in value and retain enough cashflow for future investments. New Zealand is racing ahead not just in delivering more fibre-network investment and subscriptions, but in a flourishing range of consumer choice in video on demand, ISPs and fixed-line services with a goal for FTTH to reach 75 per cent coverage by 2020.
The road ahead
The UK finds itself at a crossroads. The demand for data, largely driven by video and hand-held devices, is growing fast, but the fixed-access network investment has been woefully behind the curve, especially in comparison to many other nations.
The lack of network infrastructure competition has been a huge hindrance. BT has done well to extend the life of the copper network, but Britain cannot be ready for the future with a mere 0.003 per cent penetration level of fibre to the premises. Incremental improvements with pre-existing assets based on short-term targets are no longer appropriate in a non-linear digital world.
Britain needs to have ambition, vision and longer-term infrastructure planning. This requires looking beyond the next five years and estimating what demand and technologies will look like. With multi-gigabit speeds, many new capabilities become possible.
Virtual reality is coming to market this year, with headsets from Oculus Rift, Samsung, Google and HTC. A virtual world will be low cost, and more varied, spontaneous and flexible than the real world.
Above all, the marginal cost of adding new stock-keeping units – a measure of products in a given economy – will be close to zero and quickly exceed those in the real world. Virtual telepresence will lead to the death of distance.
Equally, the digitisation and automation of transport will be dramatic and requires low latency and frequent updates. Fully self-driving cars, expected on our roads from the middle of the next decade, will require high bandwidth in both directions.
Network Rail believes digitising the signalling system can increase capacity by 40 per cent. And Nats (the national air traffic system), plans to make huge improvements by reducing the gaps between planes in airspace from nine to 10 miles to just one mile. And flying drones – inevitably for deliveries, and maybe for people too – will rely on high-capacity, resilient two-way networks.
Britain already has many street cameras, but as the cost of data storage falls, much more will be recorded on video, which is demanding for networks.
Looking ahead, fibre to the premises has to be a big part of the solution for two reasons. First, in pure bandwidth limits, it is future-proof. At around 100,000 times that of copper8, once it is in place, maintenance costs are low, it lasts for decades and hardware upgrades to much faster speeds for optical splitters, optical network units, and terminals are simple and off the shelf.
A 1Gbit/s connection does not upgrade to 1,040Mbits/s but more likely to two, five or 10 Gbits/s. Second, it allows for high-speed modular innovation for connecting satellite, wireless, mobile and other technologies for those areas where laying cables is not financially practical.
Until 2030, policy must look at more immediate concerns. Priority has been given to connectivity to the home to serve the retail market. But business needs much higher importance given to faster networks and competition for services to the workplace.
Today’s cloud and video services depend on symmetrical upload and download speeds but they are held back by the current, mostly copper, fixed-access network.
Above all, we need new investors who seek a longer-term return beyond the typical three- to five-year telecoms company investment cycles.
That means setting a longer-term target for higher capacity and finding a way to bring in some of the $36trn (£25trn) pension funds that are comfortable with 15- to 20-year returns.
Upgrading UK broadband is cheaper, easier and much more likely to make a healthy return than many other unfunded and larger infrastructure ambitions.
The alternative of ‘business as usual’ runs the risk of spending twice or possibly three times on upgrading copper networks, while facing the threat of remonopolisation and increased regulatory oversight. The cost in lost opportunities before moving to fibre is not to be underestimated.
What do IoD members think?
Here are the findings of a Policy Voice survey on broadband in October 2015…
56% of members use broadband for video conferencing and 54% for downloading media files
57% store their data on owned or leased servers and 30% via the cloud
60% feel an increase in broadband speed would improve competitiveness and 78% their organisation’s productivity
51% think faster broadband encourages flexible working opportunities
49% are satisfied with the overall product or service provided by their broadband provider
IoD policy recommendations
• The Competition and Markets Authority should consider a market investigation into BT’s ownership of Openreach to clear the air.
• Substantially lower the physical infrastructure access (PIA) charges for ducts and poles and streamline and simplify PIA infrastructure regulations so they are not a barrier to entry.
• Set a USO (universal service obligation) target of 10Gbits/s across the UK by 2030.
• Design a new auction with the 10Gbits/s USO that aims to bring in new entrants and sets a ceiling on what share of the contracts any one supplier can win. The new rollout could start from 2017/18.
If you would like to take part in Policy Voice surveys, visit iod.com/policyvoice
For more information about Big Picture, visit iod.com/bigpicture
1 See Ofcom Connected Nations report – Section 1 Dashboard, p1
3 See Report for BT: International benchmarking report by Analysys Mason,
4 See Make or Break: The UK’s digital future – House of Lords Select Committee report on digital skills, February 2015
6 See case study on Lithuania by Fibre to the Home Council Europe (ftthcouncil.eu)
7 See Citi Research’s European Telco Regulation: Towards an open duct future,
8 See House of Lords Select Committee on Communications report Broadband for All:
An alternative vision, May 2012