Bad Signals from DCMS


Yesterday was the final day of a rushed, three week long Government consultation into the elimination of the ‘partial mobile not-spots’ — areas where there’s 2G coverage from some, but not all, of the 4 mobile operators — which cover a fifth of the UK. The Government now considers such gaps unacceptable, and Sajid Javid has warned that he is prepared to legislate a solution should mobile network operators fail to come up with a satisfactory 'voluntary' response.

One of the options the consultation considers is the introduction of national roaming. Via government dictat, mobile operators would be required to enable customers to roam onto a competitor’s network if their home signal were not available.

As the ASI has warned in a submission to the consultation, national roaming would be a terrible idea.

Partial not-spots occur where mobile infrastructure is lacking. To address them we need things like more masts, more powerful equipment and more infrastructure sharing agreements. National roaming does nothing to achieve this, and on the contrary could harm investment and the quality of mobile networks across the board.

A system of national roaming rewards those who’ve invested least in their infrastructure at the expense of those who’ve invested the most. Were it to be introduced, networks could free-ride off the infrastructure of others where their own signal is weak or non-existent, and still 'provide' coverage for their customers. Roaming also creates a strong disincentive for any one operator to invest in infrastructure where there’s complete not spots or signal from all 4 operators is weak, as well as reducing the incentive to spend on general repair and upkeep.

Since mobile networks compete predominantly on coverage and the quality of their service, roaming reduces networks’ ability to differentiate themselves. With consumers less able (or less concerned) to judge the quality of an individual network, the return on investment further lessens.

Roaming could also have potentially disastrous consequences for network’s resilience. Were one network to experience an outage, customers would move en mass to alternate networks. This surge in traffic could overwhelm another operator’s infrastructure, leading to a domino effect of failures. This very real risk to critical infrastructure has long been acknowledged as a key argument against a permanent, ‘any to any’ system of national roaming.

For something that wouldn’t actually improve mobile infrastructure and could actually actively threaten it, national roaming wouldn’t come cheap, either. The government’s back-of-the-fag-packet figures put the cost of mandating roaming as between from £276-400m, compared with projected benefits of only £54-249m.

Creating a robust system of national roaming would be a lengthy, expensive, and complex procedure. There’s a very real risk that forcing mobile operators to divert resources towards roaming would result in the slowdown or scaling back of other projects, such as the rollout of 4G. To add insult to injury, consumers would also have to pay for the cost of establishing and operating roaming, even if it makes their service worse than it otherwise would have been.

For all of these problems, national roaming isn’t even an effective solution to partial not-spots. Roaming would be ‘non-seamless’, meaning that calls would be dropped when a phone switches from one network to another. This means that roaming would do very little to help those travelling by motorway or train and going through patchy areas at speed. Calls made where there’s weak signal also risk being dropped when they would have previously stayed connected, and in some areas connection could ‘bounce’ between operators as the phone tries to lock onto the strongest signal.

Roaming would also impact other, surprising elements of consumer's mobile experience. Roaming on another’s network means that you lose access not only to things like voicemail, but all data services. The practicalities of roaming mean that a phone will probably ‘lock on’ to a network for a few minutes before searching again for a home signal, which means that consumers could be left without internet and other services for a prolonged period of time, despite only experiencing a temporary loss in signal. In addition, a phone which constantly scans for signals and changes networks will deplete its battery far quicker than one locked onto the same operator.

To ask consumers to lose core mobile services and accept diminished handset performance in the name of tackling partial not-spots is frankly absurd. Whilst it may be possible to disable roaming on some devices until needed, the fact that it’s a good idea to do so simply highlights what an enormous waste of time and resources national roaming would be.

Everything so far suggests that introducing national roaming would be a mistake. But when you look at the scale of the problem of partial not-spots, you start to wonder why DCMS even launched this consultation at all.

DCMS point out that 21% of the UK’s land mass is covered by partial not-spots; but they also admit that mobile networks are already working to bring this down. Project Beacon, an infrastructure sharing project between Vodafone and O2 is expected, once completed, to bring this down to 13%, leaving just 2% of premises affected by partial not-spots.

It’s not even clear why the government is so concerned with land mass coverage statistics, anyway. When you look at the percentage of the population with 2G coverage, you see that every operator hits 99%. In addition, spectrum licence obligations mean that 99% of the population will have 4G coverage by 2017 (and developments like voice over WiFI may prove an effective way of  extending coverage and call quality). It’s somewhat misleading, then, to portray a lack of signal as a problem for a significant chunk of the population. Whilst losing signal in rural areas and when travelling can be annoying, it's millions of miles from clear that it justifies such extensive intervention from the government.

At best, national roaming would bring marginal benefits at great cost. At worst, it would be an expensive, time consuming and potentially destructive disaster. It runs the risk of reducing competition and investment, and sucks for both mobile operators and consumers. Hopefully the consultation will convince DCMS that national roaming is a terrible solution to a problem blown way out of proportion. Certainly, the department would be best to focus on projects that would actually improve mobile infrastructure, such as reform of the inaccessible and outdated Electronic Communications Code. National Roaming is one call that it would be good to drop.