Too important to fail

Writing quite recently about demand in the US for ‘gigabit networks’, I reported one very credible view that the phenomenon reflects recognition of our growing reliance on the internet for all manner of infrastructure and services.  The proponent in question went on to explain: “that reliance is only going to increase and people will continue to want faster and faster broadband speeds for peace of mind.”  That begins to sound like an unhealthy spiral of addiction but is our growing dependence on communication networks (perversely) a ‘good thing’?  Well, it might be…

Earlier this month, a US court ruled that the Department of Homeland Security must make a plan to shut off the internet and mobile communications available to the American public. While President Obama quickly condemned former Egyptian President Hosni Mubarak for turning off the internet in his country to quell widespread civil disobedience in 2011, the US government apparently has the authority to do much the same thing - under a plan devised during the Bush administration. Details of the controversial "kill switch" authority have been classified but thanks to a Freedom of Information Act lawsuit filed by the Electronic Privacy Information Center (EPIC), DHS is obliged to reveal these within the next few months. 

Even assuming President Obama (or David Cameron) wanted to invoke such a measure, would it actually work?  Happily, the expert view seems to be that activating any kind of kill switch would do more harm than good.  According to Harold Feld, Vice President at Public Knowledge, a US lobby group focused on communications and technology policy, "I find it hard to imagine why an internet kill switch would ever be a good idea, short of some science fiction scenario wherein the network comes alive à la Terminator/Skynet.  At this point, so much of our critical infrastructure runs on the internet that a 'kill switch' would do more harm than anything short of a nuclear strike.  It would be like cutting off our own head to escape someone pulling our hair”.  A very similar argument is thought to apply to disabling mobile phones. The benefit of people being able to communicate on their cellphones in times of crisis is enormous, and cutting that off would potentially be very dangerous.  

At a time of growing concern about government security and surveillance issues, it’s heartening to find that the ubiquity of modern communications networks might actually be proof against ‘big brother’ measures.

Hail, the new economic wonder drug!

Well aware of the sundry benefits of faster broadband (not least the vastly improved performance of i-player!) I was fascinated to read recently that it’s also just the thing to supercharge economic growth.  According to analysis commissioned by DCMS from a consortium led by SQW (with Cambridge Econometrics and Dr Pantelis Koutroumpis), government interventions to upgrade broadband connectivity are projected to return approximately ‘£20 in net economic impact for every £1 of public investment’.  Wow!    

Unsurprisingly, I was not alone in feeling a tad sceptical about this conclusion: 

“Basically, BDUK has been getting a lot of flak and bad press…so the DCMS thought they best pay someone to write a nice report bigging up how much its broadband investment is going to bolster the economy”. (Computer Weekly) 

“The government, perhaps keen to cut through all the negative press, commissioned its UK Broadband Impact Study  in what appears to be a vain attempt to search for some good news about BDUK”. (The Register)

Within the Report itself, however, SQW appear to be refreshingly honest about its findings: 

“While recognising that there are still gaps in the empirical evidence base, and that the future is inherently uncertain, the study’s projections are the outputs from a rigorous and detailed analysis which draws on the best data currently available”. 

I was intrigued by that phrase, the ‘gaps in the empirical evidence base’, but it seems to stem largely from the Report’s later admission that: “The productivity impacts of increased speeds are, as yet, highly uncertain.” In fact, given the relatively recent introduction of high-speed broadband, SQW have had to invoke a labyrinth of tortuous logic and heroic assumptions to arrive at their central conclusion that: ‘an increase of 100% in the used speed in a year will lead to a 0.3% uplift in productivity  over the following three year period’. 

This sounds like a risky finding but, guess what…?

 “It also aligns…with research by Chalmers University of Technology, which found that a doubling of speed in OECD countries is associated with a 0.3 percentage point increase in GDP growth”.  What a coincidence!   

It turns out that the ‘research’ in question by Chalmers University (who??) consists of an econometric analysis of the past relationship between broadband speed and GDP growth in a sample of OECD countries.  A closer look at that analysis reveals this cautionary note from the authors: 

“This study concludes that the hypothetical impact of broadband speed on economic growth is statistically significant…. As the impact is modelled as linear, it needs to be judiciously applied when hypothetical country growth is far away from the sample means. The hypothetical impact is based on an elasticity measurement and any forward-looking simulation should be applied with care “. 

In other words, the past relationship established for this data set may not apply in the future for a different group…. As expected, therefore, take the ‘good news’ from DCMS with a pinch of salt.

Of broadband cats and pigeons

My goodness, what a flurry of feathers.… The Broadband Stakeholder Group (BSG) set out to conduct a perfectly sensible exercise, looking at the key statistical determinants of bandwidth requirement and how these determinants might change over the next ten years. Quite properly, they described this as “a model for forecasting bandwidth demand” but that choice of words may have been responsible for the subsequent furore and howls of protest when the ‘average demand’ turned out to be surprisingly low (19 Mbps).  For example, the FT reported that:

“A key government advisory group will raise questions over whether most homes in the UK are likely to need superfast broadband in 10 years’ time”.

Similarly, Sean Royce of Kingston Communications (KC) is reported as describing the BSG result as a 'red herring' that the Government might use as a ‘yard stick’ to help lower the bar for its own superfast broadband targets. 

The BSG has rightly defended its statistical results, and neither of the above policy claims would be justified, but I’m bound to agree with a further aspect of the criticism from Sean Royce, i.e.

“The second concern I have with the study is need versus desire. From our experience, there is a clear distinction between the broadband capacity that households need and the speed levels that consumers want. This isn’t simply about keeping up with the Jones’. It’s a recognition that…we [already] rely on the internet. That reliance is only going to increase and people will continue to want faster and faster broadband speeds for peace of mind.” 

Anyone in any doubt about the importance of broadband ‘aspiration’ need only consider the eruption in US demand for Gigabit capacity – probably sparked off by the pioneering deployment of Google fibre networks.  There is no shortage of editorial advice that ‘nobody needs gigabit capacity (yet!)’ but that hasn’t stopped the emergence of so-called gigabit envy.  The latest metropolitan examples are in Los Angeles and Las Vegas but there are now dozens of US cities planning Gigabit networks.  And the rhetoric isn’t confined to city mayors: earlier this year,

Julius Genachowski, then Chairman of the Federal Communications Commission, wrote an article entitled, ‘Why the U.S. Needs Gigabit Communities’.  It argued:

“We’re in a global bandwidth race, and we need to ensure the U.S. has a strategic bandwidth advantage. Without it, we risk losing our global lead on innovation, and we risk watching jobs and investment flow elsewhere….”

So much for the determinants of demand…