The IP Development Network Blog

Prologue
In my recent Broadband Dividend post, I looked at some of the reversals in rural decline directly attributable to the availability of broadband in those regions. There is a strong case for governments to ensure that there is ubiquity of access, particularly as it is an effective means to maximise the utilisation of scarce land resources and provide welfare and social services to citizens online.

Philosophy
Is there a right to broadband? We are not there yet, but there is a point, when broadband is a driver of economic and social prosperity well beyond the telecoms value chain, that this question will arise.

Past experience suggests that when there is a technology step, densely populated areas get the service first, but the rest catch up in time. It happened when broadband first arrived. There was a time when some had 512k, while others has just 56k - no-one died... Slowly but surely people caught up.

They did catch up, didn't they? Well yes and no... Yes they got "broadband", but because of line lengths "broadband" for them means 1Mbps. For people in LLU areas "broadband" can now mean 16Mbps or more. We are seeing two-tier services and two tier pricing where the haves spend less and get more than the have nots - twas ever thus. Is this wrong?

Is it wrong for petrol to cost more in Norfolk than in London? Perhaps not, but is it wrong to expect the residents of Norfolk to ride a horse to work because there are no public roads? Maybe yes... but if it is wrong, then who has the responsibility to right the wrong?

Warning: Capitalism at Work
Certainly, without legislation or regulation demanding it, the operators themselves cannot be accused of being responsible for the Digital Divide. They are playing strictly by the rules of capitalism which is the basis of their existence.

To be fair to BT, they built out the basic service almost everywhere, even where the revenue for doing so probably wasn't worth their while. They have blended their costs enabling them to build out to all but 0.4% of the population (subject to line conditions and length...) and have improved performance further with the ADSL Max upgrade. BT have invited competition to help them grow the market, but they have made Wholesale work for them.

But for a commercial entity, there is a point of diminishing returns, where taking a hit for the brand value of broader coverage no longer makes sense. Competition from LLU in the markets that were profitable brings this point clearer into focus for BT, especially knowing that the current installed hardware requires an expensive refresh via 21CN to go any faster.

Consider the competitive environment: at a recent industry conference, a senior BT exec revealed the following approximations of LLU competition. Around 1,500 exchanges have at least one LLU Operator, of which

450 have only have one LLUO
625 have between 2 and 5 LLUOs
400 have more than 6 LLUOs
1 exchange has 11 competing providers

So there is competition for BT meaning ADSL 2+ available already to 80% of the population. To keep up / leap ahead (depending on who you believe) BT is building out its 21st Century Network.

21CN Solution to a 21C Problem
21CN is starting with a controlled rollout in Cardiff and will then cover the rest of the country, starting with the bigger exchanges. It will take a while for it to filter down to the regions. During that time the haves might have 20 times faster connections again than the have nots, but eventually the have nots will catch up.

It might be a bit more than 0.4% of the population that miss out this time, but my exchange is in the smallest 3% and the switched-on date for 21CN is in the Spring of 2009. So some people will have to wait... It didn't matter last time, so why should we be concerned about it now? I'll get ADSL 2+, just not for a couple more years.

It seems clear to me that the 21CN commitment will deliver me and almost everyone else in the country who live within 3km of our exchanges, a substantial boost in bandwidth. We might have to pay more for our petrol, but at least we can put the horse out to pasture.

BT Wholesale fairly recently changed to a tiered exchange pricing model, meaning that the operator's access costs more in the most rural locations, a bit less where there are a few more people and a lot less where there is competition. Does this condemn the digitally poor to forever receive higher priced services? If it does, it is not however, BT's problem.

The Case for Subsidy
If the government's act of opening the markets risks causing social inequality by removing BTs ability to cross subsidise, it may have to be the government that take responsibility for rectifying the issue using its own means of cross subsidisation: taxation.

The Mechanics of the Digital Divide
For sure, there is enough bandwidth for most people now, certainly enough to meet their basic needs. But there is an underclass for whom service is not going to improve without new exchange builds. Those stuck more than 3km from their exchanges are effectively stuck below 8M, whatever technology is used and even getting to that rate may cost a bit more cash. QoS looks like being most attractive where speeds are lowest (sound familiar?).

To go faster than this, copper loops need to be shortened. About 2.1m households are connected on copper lines more than 3km from the exchange. This figure is as the crow flies, so bear in mind that copper wasn't laid by the Romans and so tends to be a fair bit longer than its straight line distance.

Shortening the loops is where Fibre to the Node (FTTN) comes in. In the UK, where there are just under 6,000 exchanges, these connect out via 90,000 or so Primary and Secondary Cross-connection Points (PCPs and SCPs) and 2.5 million distribution points. These PCP/SCPs would need a massive capacity upgrade (the F in FTTN), but there is also a big problem with physical space, power, cooling and security, all of which would need to be enhanced. But it doesn't sound like BT wants to do it, and they are returning cash to shareholders instead of saving up for the next big upgrade.

A Digitally Divided Britain
What happens when the bandwidth required by "basic needs" reaches the 8M ceiling imposed by the copper line length for the 9% or so who live more than 3kms away? The availability of 20+ megs in some locations will allow services to develop to fill the available capacity. If those services are life changing, in the real sense that they offer new economic opportunities or provide access to enhanced personal healthcare for example, you have a problem.

At that point you can probably say that there is a Digital Divide, because for some, use of life changing technology will be impossible because of network constraints. The Digital Divide is determined not by the exchange you are connected to, but by how far you are from that exchange.

Rectifying this will cost a lot of money and such investments are likely to have low yields as the worst affected exchanges are the usual suspects - those that are hard to reach anyway, with low densities even before you break the base down further by extending the core network to the cross-connection points.

Where are the Digitally Poor
There are no surprises when looking at which regions are worst effected. 24% of Lincolnshire's population live more than 3kms from their exchange. This is 25% in North Yorkshire, 28% in Cumbria, 29% in Dyfed and 30% in Co Down. For the Scottish Islands and Highlands and rural Northern Ireland this % is much higher, but over this entire area the problem base is only 93,000 households - just 10,000 more than are impacted in Lancashire alone.

So don't think of this as a problem that just hits the hermits among us who like to live in the middle of nowhere. There are 83,000 people in Lancashire (near Manchester), and 66,000 in each of Essex and Kent (near London) that also have lines longer than 3kms. Here the inequality problem may be more acute. If you move to the Isle of Tiree, perhaps you are expecting to get away from it all - so what if you can't get faster broadband? - but if you live in Kent some might expect to be connected better than you are.

It is in these areas where the Digital Divide might be quite stark. If a village is stuck on 8M and you can get 20M two miles down the road (in the same "good school" catchment area), you might see an impact on that most precious of post-Thatcherite Britain's benchmark - the value of our homes. Why? Because by that time 8M is not enough and puts you at an economic or welfare disadvantage.

Regional Answers from Regional Government
Regional governments have in the past attempted to get a local broadband provider off the ground in order to address the digitally poor, but levels of success have varied from minuscule to none. The problem in many cases was that the investment was a one off and when upgrades were required, the underlying cashflow was not enough to make it self sustaining. An ADSL exchange with 100 households is never going to be profitable so many of the companies that were created went bust.

Alternative Access to the Rescue?
So what of alternative access? WiMax for example could fill in the areas where scale and geography are a problem. But that's not where WiMax or any other new technology is being deployed because in spite of the competition from wired and other wireless networks, it is still more likely to be economically viable using WiMax over the top of existing networks than it is filling in the gaps between them. The Digitally Poor are digitally poor because service providers can make more money deploying their capital in more densely populated areas, regardless of whether the technology they use is WiMax, WiFi, 3G, LLU or FTTH.

Any spectrum auction where service providers need to pay just to have the ability to offer service simply accentuates the problem by making the the commercial viability of providing access to the digitally poor even more remote. It may be that instead of the operator paying the tax man for the spectrum, it may be necessary in some regions to work it the other way around.

For the rest of the market (the digitally rich in the towns and cities), the overlay of yet another competing technology has the opposite effect. It will prove baffling to the consumer with so much choice on offer, no doubt leading to further price wars and hara kiri amongst the operators. This is may be the unintended consequence again, as it has been so many times before when regulation has opened up new markets.

Certain use of the soon-to-be-available spectrum could offer coverage but might also significantly distort existing markets, where investments have been made on the assumption of stability. 3G of course, but also 21CN and LLU. Can that distortion be allowed to happen? But if it isn't, how can infill be achieved?

Summary of the Problem
So here we have a summary of the problem. Something that solves the problem for the digitally poor is likely to severely distort the market for the digitally rich which may cost more to the economy than leaving the bottom 10% with the scraps. But aside from the numbers, there is the looming question of whether broadband will be a pre-requisite for the economic progression and social equality of a country.

I do wonder what we would arrive at if we were to throw the whole picture away and design it from scratch, and it is tempting to try and right the wrongs when new opportunities arise like the Digital Dividend. I fear however that we were given a spade and some land about 25 years ago and we started digging. Now we have found ourselves in a hole and are tempted to use the spectrum auction to try and dig our way out.

Disclosure
Disclosure: I own 336 shares in BT worth just over £1,100 at the time of writing. I inherited these shares in two batches from my grandparents who bought them at privatisation. I have no intention of selling them, much in the way that I wouldn't sell the paintings and other memorabilia that similarly came to me from my deceased relatives.

Labels: BT, digital divide, FTTH, next generation networks, Wi-Fi, WiMax

# posted by Jeremy Penston @ 6/29/2007 05:44:00 PM

Friday, 29 June 2007

The Digital Divide