The wonders of the Internet

From my hotel room in London, I read this SMH report, headlined “NBN benefits ‘grossly overstated'” which in turn refers to a report by “British telecommunications consultant Robert Kenny and Charles Kenny from the US Centre for Global Development” released (in London, as it happens) a couple of days ago.

Five minutes with Google is enough to determine that

* the Centre for Global Development is a genuine and reputable thinktank, with no particular axe to grind

* Charles Kenny is not what you might call an Internet enthusiast, having written, in 2002, a piece entitled Should we Try to Bridge the Global Digital Divide.

Kenny’s answer is “No”, which doesn’t seem to be the view of people in poor countries. According to this graph from Wikipedia’s article on the topic, the proportion of developing country people with Internet access in 2007 was the same as that of the developed world in 1998, which suggests that by now, developing countries must have already reached the 2002 developed-world access rates Kenny said they didn’t need.

That’s not to say that Kenny might not be right this time. But, so far, betting against the idea that people will want faster and more powerful communications and computers has not had a good track record.

30 thoughts on “The wonders of the Internet

  1. JQ, what do you make of this claim from the article: South Korea, cited as the world leader in providing fibre to homes, enjoyed annual productivity growth of 7.6 per cent per capita in the decade before it began the program and 3.8 per cent in the decade since.

    ”Many factors played into the growth slowdown,” the study says. ”But maybe the massive increase in online gaming, facilitated by the broadband revolution, played a role – the South Korean government estimates that as many as 2 million of its citizens are addicted to online gaming.”

    Seems a little overblown on the surface.

  2. The first thing Quiggan does is check the author’s credibility. The details of the report or argument are irrelevant. Unless of course, the conclusion matches ones personal view.

  3. The report is typical of what comes out of conservative US “think tanks” and appears to have more of a political aspect to it than actual hard data. There’s plenty of conjecture and suitable quotes, that’s for sure, but as a report on the future viability of fibre optics it lacks credibility.
    I suspect that the release of this report, at this particular time, has more to do with the House of Reps vote in the Australian Parliament today than it does with the long term viability of a fibre optic network.
    Scant and condescending reference is made to the fact that 10 countries around the world are rolling out fibre optics. As a friend of mine commented, “It looks like a pretty ordinary Uni assignment. I’d give it a Credit at best.”
    I wonder how the report would have read if Australias NBN was a privately owned, government subsidised entity.
    A word of caution. All such reports have an agenda. This one is no different.

  4. Julian :
    The first thing Quiggan does is check the author’s credibility. The details of the report or argument are irrelevant. Unless of course, the conclusion matches ones personal view.

    Well, there is concept that is becoming very popular with the plethora of PR releases that purport to be from “independent” think tanks. The concept is that of context. I find that if I read an article from a particular news site I generally have a feel for the context under which they write and report. If an external organisation I have not heard of is quoted, I’m often interested in what their context is – I want to know if they are pushing any barrows or on what they base their findings. Newspapers often just grab the conclusions and running with it; I want more context.
    Now I’d suggest anyone who doesn’t look for context in any reporting are perhaps themselves set in their views.

  5. @Julian
    but doesn’t every body check?

    i mean no one would be silly or negligent enough to broadcast something from (say)”The Onion” as verifiable believable news.

    would they?

    the medical proffession seem to be in (non-biblical) raptures and the on line further education industry seem to like the NBN idea rather a lot as well.

  6. The first thing Julian does is check the author’s credibility. The details of the report or argument are irrelevant. Unless of course, the conclusion matches one’s personal view.

  7. Julian, if you disregard the putative credibility of the report’s authors, then the content of the SMH report becomes “some people don’t like the NBN”, which is not exactly news. So, credibility is the first thing to check.

    An analysis done by an independent author or group with no particular axe to grind may be right or wrong, but it is worth checking. An analysis done by an author or group with an established line is worth reading, but you need to be aware that you are looking at a case in which favorable elements are likely to be emphasised, and others downplayed. A “report” from a PR front, however glossy is not even worth looking at. Fortunately, the Internet makes it much easier to work out which is which.

  8. Julian – with the huge rise in shadowy funding of PR companies whose remit is to spread uncertainty and doubt about various things such as global warming, DDT, cigarettes and Julian Assange, you do have to check.

    Similarly, you will most likely be distraught to discover, that the deposed Nigerian general offering to share his ill-gotten gains with you is not all he seems.

    Sadly, you can’t take the honesty of think-tank employees or random people on the internet for granted nowadays. For instance you can’t do a background check on me : /

  9. Julian – with the huge rise in shadowy funding of PR companies whose remit is to spread uncertainty and doubt about various things such as global warming, DDT, cigarettes and Julian Assange, you do have to check.

    Sadly, you can’t take the honesty of think-tank employees or random people on the internet for granted nowadays. For instance you can’t do a background check on me : /

  10. I haven’t seen the recent report, but the 2002 Kenny report seems fairly irrelevant.

    1) In developing countries, Internet access is most likely to exist via (ever-cheaper) smartphones, whose infrastructure is wireless, far less expensive than wired.

    2) In talking about this, it is fairly hard to make sense unless one carefully distinguishes:

    a) The bandwidth ranges available, and their prices.
    b) The fraction of people that could get access to each range.
    c) The fraction that choose to pay for that.
    d) And sometimes location.

    For example, I might be willing to pay more for better bandwidth to home, but not be able to get it any price. [We live in the hills above Stanford. Almost always, higher bandwidth was available down in Palo Alto where the population is much denser.

    Sometimes, one needs location availability more than high bandwidth. For example, if we’re on the East Side of Sierra Mountains, and I want to check a map on the IPhone, the difference between no map, and one that takes a few minutes to download is huge.

    3) Sometimes, more bandwidth matters a great deal, as if you want to to transfer Terabytes between supercomputers, or are download/uploading big multi-GBM files.
    But in many cases, there exist plateaus where:

    X is easily good enough for activity A, but unbearable for activity B

    Y = 4X is good enough for B

    In this case, being offered Z = 2X for $1.2X may not really be very interesting.

    In fact, getting Y = 4X may not even be interesting, if you don’t do B, or you would like to do B, but the backend infrastructure isn’t there.

    Current example: Y might be enough bandwidth to stream HD movies … but if there don’t get to be the services around to do that, many people wouldn’t care.

    So, I would suggest that more higher bandwidth is good, but at some point, reducing the cost and increasing the pervasiveness of each bandwidth level is way more important. At some point, for any given usage pattern, incremental bandwidth, CPU performance, graphics performance is *good enough.*

    What is the Oz government actually trying to do?

  11. @jquiggin
    These days it is imperative to check credentials – if only to check whether the report is objective and unbiased.. as clever Don noted there are so many Eggsperts and Pruffeshanall Schkepticks for sale to the highest bidder.

  12. Charles Kenny’s blog at contains enough material to mark him as a fairly interesting and committed development economist.

    And the content of the paper (co-authored with his brother, a telecoms consultant) suggests he knows a fair bit about telecoms issues.

    John, I’d really like your considered view of this paper.

  13. Again, can someone explain briefly what the NPN is actually supposed to do?
    In context of my earlier post?

    For example, one goal might be to radically increase bandwidth in big urban areas, with hope of enabling new services.

    A different goal might be to boost the bandwidth of people who currently have little.

    The geographic dispersion of USA or Australia make them different from, say Korea.
    It is expensive and takes long to get higher bandwidth when people are spread out, then when a large fraction of people live in dense apartment buildings.

  14. John your observations are right on the mark but your question as to what the government is trying to achieve? What is the sound of one of Tony Abbott’s (“ferocious”, one-Bible-believing opposition leader) one hand clapping the importance of science to civilisation? What was the look on Chris Monckton’s face before he was born? If another tree falls in the Amazon basin and turns out to be the final straw that tips the regional climate there over into a stable drought phase, does anyone hear it weeping? – these may be easier to answer.

    The current bandwidth gap between major Australian cities and even the largest minor cities and regional centres is huge. In Sydney compared to its surrounds I think it’s indicatively of the order of difference between average access to say 10Mbps ADSL2+ in the city compared to very variable wireless and mobile network speeds, satellite (slow uplink, long lags and latency), ADSL struggling to reach 256Kbps, or dial-up.

    In Australia’s huge landmass most (85%?) of the population lives in large cities on the seaboard. As you note, getting broadband to the 15% outside those centres is expensive.

  15. Those population estimates must be wrong, it’d be closer to half the populace living outside the areas of easy access to fast broadband I think. Our numbers are mostly distributed around the seaboard, true, but not concentrated to the extent of 85% in the capital cities where the broadband is available.

  16. This is a curious post. It sounds like a cautious defense of the NBN fiasco by raising doubts about a past Kenny forecast. The fiasco here is not necessarily the huge cost but the evasive, dishonest attempt to foist this scheme on Australia when it’s economic and technical rationale is unclear.

    The sorts of questions raised by John Mashey and Charles Kenny need to be answered.

  17. I saw the article but didn’t have time, energy or interest in checking the veracity of the author – although I did wonder. I was not sure about the Korean example of dropping productivity and increase in gambling. After all we have had a huge increase in gambling in Australia but it is not the sole province of the internet. The issue of on-line gambling however does show that there are almost always downsides to innovation which need to be controlled if possible.

    The Kenny article however does put me in mind of the quotation from the nineteenth century which goes along the lines of ” everything that can be invented has been invented”. We are never short of people who look to the past to determine the future because they have no imagination.

    There have also been many predicitions about computers which have been wrong because the person making the prediction has had no idea about the power of computing or its ability to create new usages. Who would have predicted five years ago the hold that social networking sites now have?

    That is one of the things about the NBN that the article didn’t really deal with – the unknown usages that will become available as a result of the increase in capacity.

    The Kenny article did create a question about how we should deal with the unintended consequences of the evils of society finding the increased bandwidth as tempting and useful as the social and economic spheres. Just the same he didn’t make the case that the bad outweighs the benefits even though this seemed to be the intent of the report.

  18. John Mashey, I like your examples. This is the kind of detail that the NBN debate rarely gets to, but should. There are a lot of variables. For example, what’s the real demand for streaming vs the demand for downloading onto an enormous hard drive? (On a 10Mb connection, a 4GB movie will take around an hour to download.)

    The Kenny brothers’ report goes into some of this. For more, Google the work of Andrew Odlyzko.

  19. I’ll do a proper analysis when I get a round tuit, but I’d make a couple of points quickly

    * This is one of those choices where there is no way of telling for sure, so calls for a benefit-cost analysis are really not that helpful. What might be more useful is to compare the NBN to the lower-cost alternative the Libs were talking about a while back

    * The quote mentioned by Tim @2 is not promising, both because of the use of “addicted” (I could estimate that 10 million Australians are “addicted” to TV) and because of the implication that gaming is not an economic activity

  20. John, Cost benefit analysis is never an exact science which ‘tells for sure’. Its always a pretty rough attempt to gather in all the info on benefits and particularly to be inclusive on the cost side. It is typically based on hypotheticals.

    I can’t see how you can compare the costs and benefits with those obtained with the Liberal proposal without doing a rough cost/benefit analysis on each.

    I’ve read assessments that suggest it will be a waste of money with the city massively cross subsidising the country and the subsidies being suppressed by a high tech chic image an d opposition to the scheme being treated as Luddism. As economists its important not to allow the pollies to be evasive on the use of huge amounts of our money & to reject their glib lies and evasiveness.

  21. For those who haven’t read the NBN implementation study, you should. It does address some of the questions raised here

    No, I haven’t read it all, so guilty as charged. But main points:

    1. Fibre to the premises for ~ 93% of households and businesses. Optionally up to 100 mbps.

    2. Most of the rest covered by wireless min speed 12 mbps

    3. The rest covered by satellite.

    4. The principle of universal access.

    The aim seems to be to run everything – voice, internet access, subscription (ie cable) TV, and whatever else through the fibre.

    Retail services to be provided by Retail Service Providers who pay for Layer 2 network access. NBN will not in general provide IP services – really important point. Subscription to multiple RSPs through single fibre link supported. Universal Layer 2 access is the basis of the governments view of facilitating competition for retail services.

    I was initially a bit dubious about NBN, but after reading the implementation study I’m thinking that the govt may have got it right.

  22. @hc
    Can anyone tell me if a cost benefit analysis of NBN would take into account the increase in employment and the multiplier effect on those employment incomes – or is the benefit just a shallow pricing X user – revenue exercise?

  23. What is the Oz government actually trying to do?

    Get re-elected.

    Seriously. We had a very strong balance sheet as a country (nil public debt) under our version of the Republicans (Americans would think of them as right wing Democrats or the liberal wing of the Republicans so they’re more of a centrist party) and they got turfed out for a big spending Labor Party (our version of the centre-left wing Democrats).

    As good earnest middle class types they decided super high speed bradband and climate change were key policies and wasted huge sums of money on both rather than either preserving the country’s balance sheet or spending it building genuinely productivity enhancing infrastructure.

    At the same time Hong Kong is rolling out a wireless broadband solution, seems to me that mobility and high speed will be far more useful than fixed point ultra high speed so it seems we’re building the equivalent of the Simpson’s Springfield Monorail.

  24. One of the major cost factors in the NBN is universal access. If the access was cut to (eg) 85% you’d expect much greater drop in cost. How could you quantify this benefit?

  25. Rob #27 – your bias is showing. You really should stop listening to Tony Abbott who has not yet discovered the on button on a computer.

    The NBN is building “genuinely productivity enhancing infrastructure”. Wireless is not a good option except for the most difficult to access areas for a number of well publicised reasons – including the massive numbers of towers required and the limitations of the technology. Hong Kong can manage it because of the density of population in a small area.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s