How the Carter Tax could work

1 Matching funds to needs

As far as broadband is concerned, the key proposal in Lord Carter’s Digital Britain Final Report is to charge what he delicately calls a “supplement” of 50p a month on the rental of every telephone line in the country (barring those provided under social telephony schemes). The money raised will go to a “Next Generation Fund” to deliver “at least 90% coverage of next-generation broadband by 2017.”

Many commentators have said such funds would be quite inadequate and the objective is far short of what is required. But Point Topic’s research shows that this modest tax should be more than enough to support the rollout of superfast broadband to serve over 90% of the UK population. With a flexible approach it should be able to take coverage up into the high nineties.

On the other hand, our current forecasts suggest that 2017 looks a rather slack target for achieving 90% complete coverage. We think it should be brought forward by two years. We also believe that we should see a much more ambitious definition of what taxpayers are going to get for their money than Carter has given us so far.

Based on some simple assumptions, our research shows that market demand alone should be strong enough to bring next-generation broadband to around 73% of the UK population. We estimate that, provided it was efficiently used, a cross-subsidy of £70m a year could extend coverage to another 19%. Reaching the final 8% will be more difficult, but with the Carter Tax able to raise at least £170m a year there should be resources to spare to do it.

The key revenue assumption behind this is that the equivalent of 60% of present-day broadband users will be prepared to pay an extra £1.50 per month (including VAT) to get the benefits of next-generation broadband, rather than current products.

The matching cost assumption is that the average investment required to provide fibre-to-the-cabinet (FTTC) broadband to a single street cabinet, able to deliver over 40Mbps to each user in its service area of about 1.6 sq km, will be £50,000. We assume this will be amortised over 10 years with a regulated rate of return of 8%.

On this basis it turns out that next generation access needs about 300 customers per square kilometre to be viable on a purely commercial basis – corresponding to a total broadband density of about 500 lines per sq km after allowing for the 60% take-up assumption.

More generally, Point Topic believes that broadband density (the estimated number of broadband lines per sq km) is the best single indicator of the “next-generation attractiveness” of an area. For example, this is the first measure which communications providers should consider to decide where they should deploy their next-generation investments. On the other side of the market it is what development agencies should look at first to see which parts of their region are most likely to lack NGA-appeal and need special support.

2 Broadband density zones

Point Topic’s model for estimating the relative attractiveness of different areas for rolling out next-generation access (NGA) networks is explained in more detail in a separate UK Plus Short Report, 'Assessing 'NGA Attractiveness'', together with more detailed results. Here we provide just a high-level overview.

Broadband density varies hugely, from over 10,000 lines per sq km in business districts and apartment blocks to under 1 in the most thinly populated areas. Point Topic has devised a range of nine density zones to be able to view this range in a manageable way.

In turn we have merged the detailed nine zones into four broad areas which focus on the main differences as far as distributing the Carter money is concerned. The pattern of these areas for the UK as a whole is shown in Map 1 while their basic statistics are provided in Table 1.

Figure 1: UK NGA Appeal (click for larger version)

The four areas range from the highest-density Area A, dark blue on the map, where next-generation broadband can be financed by market demand alone, to the lowest density Area D, brick red, which will need case-by-case solutions to bring broadband to the most remote places in Britain. Area D looms large on the map because it covers over 75% of the land area of the UK but only 4.4% of the population while Area A looks much less significant than it should, with 73% of the population but only 3.7% of the land area.

Table 1 Figures for the four Density Areas

Map 2 gives a close-up view, focusing just on London and the South-East, to show how the different areas are distributed between the towns and the countryside.

Fig 2: NGA in London and the South East (click for larger version)

Area A includes the three most high-density zones (9, 8 and 7) and the great bulk of Britain’s homes and businesses, comprising anything from city centres and apartment blocks to leafy suburbs. It includes almost all the areas covered by Virgin Media’s cable network, which is already offering a type of next-generation service with download speeds up to 50mbps.

Area B comprises the next three zones (6, 5 and 4) with broadband density ranging from 500 to as low as 30 lines per sq km. These areas include suburban fringes, villages and the more populated countryside, with 19.3% of the UK population. Point Topic estimates that the extra revenue generated from next-generation services in Area B areas would be about £25.3m a year, £69.2m short of the revenue required to finance it. This gives an indication of the subsidy needed.

Area C has been identified with just one density zone, Zone 3, because this is the transitional area where straightforward subsidy should give way to more complex solutions. We estimate that total extra next-generation revenue in Area C would be only £4.1m a year while the amount required to finance the rollout on the basic assumptions would be £126m a year. But Zone 3 is essentially small villages and hamlets with no continuous built up areas. Here next-generation broadband will be a matter of delivery to isolated points rather than across broad areas, and the technology and economics is likely to be considerably different from what prevails across Area B.

Finally, Area D comprises the most remote areas of all, density zones 1 and 2, with nowhere supporting more than 10 broadband lines per sq km, and often much less. The same points apply as with Area C only more so. Satellite broadband could be one solution but these empty moorland areas will often also have very low costs per kilometre for laying fibre-optic cable. Subsidies will still be needed but with local knowledge and initiative they could be much lower than appears necessary at first sight.

While the four Areas show the NGA appeal of different parts of the country at a broad strategic level it is necessary to look at the nine Zones to see the full detail. Map 3, focusing on the seaside city of Brighton and its hinterland of Sussex countryside, shows how sharply broadband density varies between tightly-packed town centres and open spaces such as the South Downs. Working at this level planners, marketers and campaigners can see immediately where the market is likely to support investment or where special intervention is likely to be required – and even what form it might take.

Figure 3: NGA Density Zones (click for larger version)

3 What kind of service?

Lord Carter’s proposals will not succeed unless they deliver a broadband service for the 2010’s. This means broadband with the performance which enables ordinary households to access the high-performance internet applications they will want to use. The public is gradually realising that having a high download speed is not the biggest issue for their broadband; increasingly they want more modest speeds which are continuously sustained and, in particular are not choked off at peak hours. “We just want to be able to watch iPlayer,” as one newspaper interviewee put it when asked to comment on the Digital Britain report.

Peak download speeds will not be a problem for next-generation customers. BT’s FTTC rollout programme should be able to provide up to 60mbps. Virgin Media is already offering 50mbps and working on 200 – although the cost of that is likely to be high. But high peak speeds alone will not meet the needs of broadband users in the 2010s.

To have a quality experience with iPlayer, especially in high-definition versions, users will want to receive a sustained, continuous data stream of at least 2mbps, maybe 4. To join in video exchanges with their friends they will want at least 1mbps upstream as well. They will want low latency and jitter for good quality voice and successful online gaming. All this will define a service which represents a step change beyond the broadband we know today, so much so that it really needs a different name – “superband” is one suggestion.

Simply building out next-generation access services which offer scores or even hundreds of megabits downstream will not be enough. The delivery of superband features and quality of service must be a requirement of next generation infrastructure delivered with the help of Carter money – and hopefully that which is financed by the market as well. Such Quality of Service will also demand substantial investment in the “middle mile” backhaul and core networks, which we believe will be covered by the costs we have assumed.

The Digital Britain timescales also look unambitious. Although BT has committed to achieving next-generation coverage of 40% of the UK by the end of 2012 our current forecast is that it will slip by a year to end-2013. But by then BT and its competitors should be rolling out at the rate of 4m more homes and businesses passed every year. Even just maintaining this rate would take them to the 25m level needed for 90% coverage by mid-2017. A target of 2015 for 90% coverage would set a more appropriate challenge.