Navigation

FAQs:  Responses to common criticisms of the NBN Cost-Benefit Analysis

6th September 2014  |  Comments  |  Blog, Media Releases

The CBA was conducted by the Minister’s ‘mates’ and not an independent body such as Infrastructure Australia or the Productivity Commission.

The Cost-Benefit Analysis was too conservative when it estimated of future bandwidth requirements for Australian households.

The Cost-Benefit Analysis is wrong because it fails to recognise that bandwidth usage will  expand massively due to demand from applications not yet invented.

 

 

The CBA was conducted by the Minister’s ‘mates’ and not an independent body such as Infrastructure Australia or the Productivity Commission.

Claims

  • Jason Clare wrote: “Like many of his pre-election commitments on the NBN, Mr Turnbull has broken this promise. Instead of appointing Infrastructure Australia to do the cost benefit analysis, he has hand-picked former staff and some of the most vociferous critics of the NBN. This includes well known critics of the NBN Henry Ergas and Kevin Morgan, former adviser to Malcolm Turnbull Alex Robson and former Liberal Party staffer David Kennedy”.
  • Jim Chalmers wrote: “Communications Minister Malcolm Turnbull’s latest so-called 'cost-benefit analysis', performed by hand-picked former staffers and mates.”

Response

  • The assertion that the Government “promised” Infrastructure Australia would conduct a cost-benefit analysis of the NBN is untrue.  Regrettably, this falsehood has also been asserted by news outlets such as The Australian Financial Review without anyone ever bothering to actually fact-check its veracity.
  • Our broadband policy at the 2013 Election stated a Coalition government would conduct an “independent cost-benefit analysis and independent review of regulation”.  It did not specify Infrastructure Australia, the Productivity Commission or any other particular entity to carry this out.  While on the topic of what we actually promised versus what others claim we promised, I should also note we never stated changes to the NBN Co rollout would have to wait until; completion of the CBA – another commonly repeated myth.  What we promised prior to the election in the part of our broadband policy which described our three reviews of the NBN is exactly what we have done in office.
  • The claim I am ‘mates’ with some or all of the panellists and they were chosen because of this is farcical.  The four panellists’ biographies show they are distinguished and very experienced individuals with very strong backgrounds in directly relevant topics: telecommunications, regulation, competition, the digital economy, and public sector finances.
  • Dr Alex Robson is the only former staffer of mine who contributed to the CBA – he was an economics advisor in 2008-2009 (five years ago).  Dr Robson is a micro-economist at Griffith University with a strong background in relevant disciplines including telecommunications,  regulation and competition.  He contributed a methodology paper for the review (that is, he helped set out the analytical framework within which cost-benefit analysis was carried out – which is quite distinct from providing inputs and analysis that affect its conclusions).  Dr Robson’s work  was then peer reviewed internationally, by Professors Nicolas Curien (Conservatoire National des Arts et Metiers, Paris),  Kenneth Flamm (Lyndon B Johnson School of Public Affairs, University of Texas, Austin), Jonathan Pincus (University of Adelaide) and Cliff Winston (Brookings Institute, Washington).
  • To the extent that David Kennedy of Ovum, who helped support the review in it latter stages, was a former staffer of Senator Alston, that was many years ago.  Mr Kennedy and Ovum have a long and distinguished track record of providing objective advice to many industry players.
  • As to whether Infrastructure Australia should have conducted the study, it is important to remember that IA is a small organization with a very modest budget and few full-time staff.  It has limited telecommunications expertise and no history of reviewing telecommunications projects.  Indeed, even when considering other types of infrastructure, it does not generally undertake its own cost-benefit analyses from scratch – it instead typically reviews CBAs performed by others on projects seeking Commonwealth funding.  If we had asked IA to conduct the study, they would have had to outsource it to external experts.


The Cost-Benefit Analysis was too conservative when it estimated of future bandwidth requirements for Australian households.

Claims

  • Jason Clare: “When you do an analysis like this it depends on the assumptions that you put in, and if you get your assumptions wrong, then the result is wrong. The assumptions that they’ve put in on cost, on time, on revenue and on demand for the NBN, I think are all wrong. The report says that it expects that within 10 years’ time that only 5 per cent of Australians will want 43 megabits per second or more. Well, already you’ve got 28 per cent of Australians that are ordering 50 megabits per second or more.” (Source: Jason Clare, Doorstop Interview, August 27.)
  • Michelle Rowland wrote: “According to Turnbull’s cost-benefit analysis, in 2023 – that’s less than 10 years’ time - the bandwidth requirements of the median household will be 15 Mbps – which is less than half that of today’s reality.  The report’s discussion of demand refers to large software downloads, video uploads and busy hours within a household. There are frequent references to improvements in video compression, a matter the Minister was keen to pitch to the media when confronted with the absurdity of the premise in his cost-benefit analysis.  But as highlighted in much of the immediate reaction to the report, these are not the standard, static usage patterns one can anticipate in future. Nor do they reflect the utilisation of broadband beyond the internet, nor the pervasive nature of the internet as integral to everything rather than a mere application on a device.  The NBN is not about the download – it’s all about the upload.  Now that ‘the cloud’ is embedded in our vernacular it’s increasingly clear that its viability is intrinsically linked to the upload”.

Response

  • In any cost-benefit analysis, the assumptions that underpin its results are a crucial input – in fact an important requirement and benefit of any rigorous CBA is that underlying assumptions must be made explicit and the consequences of varying them tested.  The Vertigan panel’s work  therefore tested the impact on its results of a very broad range of assumptions about inputs – including the results of the CBA if the assumptions underpinning Labor’s original NBN strategy were retained.
  • The CBA found that under 98 per cent of the possible combinations of underlying assumptions tested, a multi-technology mix NBN delivered positive net benefits compared to an FTTP NBN.  Once MTM’s scope for future upgrade of technology was factored in, the multi-technology NBN is always more beneficial than FTTP.  The NBN currently being implemented has higher returns for the economy and society as a whole than the previous Government’s plan under any plausible assumptions.
  • The estimates of demand that Clare and Rowland refer to are important supporting information used in the cost benefit analysis.  But they are not the demand estimates used as the central case for the analysis. Rather, the CBA estimated the benefits of the different options using a study of consumer willingness to pay for broadband of different qualities.  This work used a large sample of consumers and was conducted by the Institute for Choice at the University of South Australia. The bottom-up study of bandwidth demand referred to by Clare and Rowland was actually used to conduct a sensitivity analysis of benefits. The cost benefit analysis also used actual uptake data for the NBN so far in conducting its sensitivity analysis.
  • Rowland’s claims about uploads are wrong.  The study explicitly took account of willingness to pay for uploads – and found that using any method, willingness to pay more for upload speeds was even lower than the willingness to pay for downloads.
  • All three methods of estimating current and future demand used in the CBA supported the same conclusion – that undertaking a multi-technology NBN had substantial net benefits compared to undertaking Labor’s FTTP NBN.
  • Clare is wrong in his interpretation of the demand study.  This work did not say that “within 10 years’ time only 5 per cent of Australians will want 43 megabits per second or more”.  The paper he refers to is a bottom-up study of technical demand for bandwidth – that is, actual usage by a combination of actual applications that people in households of a particular size will simultaneously use – a concept which is quite distinct from the bandwidth people are willing to pay for.  An explanation of the difference is provided in the broadband demand study on page 3. This differentiation is important because to assess the incremental economic value of superfast broadband, it is necessary to know how it will be utilised.  Bandwidth only has utility to the extent that it allows you to do things you otherwise couldn’t (and that you value). 
  • The study found the median user will probably use about 15 megabits per second of bandwidth by 2023 in the busiest periods  The caveats include that the median user is willing to cope with one minute per week of degraded service, and that there will be high end users who will use much more.
  • The paper’s author notes that dimensioning the network to suit the median household is a bad idea because half the users will get a service that doesn’t satisfy their usage (p.6): “Note that in offering figures for median demand, we are not suggesting that access network capacity should be based on this metric – by definition, such a network would (to some extent) disappoint 50% of households. Rather, access capacity should be driven by higher end users. Whether this means the top 1% or the top 5% (or some other figure) is a matter of judgement.”.
  • The more important study in terms of the NBN’s business case (and the one more pertinent to take-up by speed tiers) is the willingness to pay study (available in the main study, see pp.59-74).  This work took into account current take-up of the NBN and compared it to other countries – finding Australian consumers are subscribing to higher speed tiers (e.g. 100 megabits per second) despite having to pay higher premiums than abroad to move from lower speed tiers (see pp.70-71).  Note the percentage of people buying 100 megabits per second on the NBN has fallen from 23 per cent to 19 per cent.

 

The Cost-Benefit Analysis is wrong because it fails to recognise that bandwidth usage will  expand massively due to demand from applications not yet invented.

Claims:

  • Jim Chalmers wrote: “The ‘future we can’t yet see’ is the real point of universal high speed broadband: to accommodate the yet-to-exist broadband applications that will inevitably come down the track, forever disrupting and improving the way we educate our children, deliver health services, do business and make life more comfortable for our seniors.” 
  • Rod Tucker wrote: “The Vertigan Report supports the image of Australia as a quiet little broadband backwater, comfortably conservative and not particularly interested in grabbing the opportunities provided by the ongoing digital revolution. It conveniently ignores or understates key international developments in broadband technology and deployment.”
  • Sydney University Associate Professor Kai Riemer told the ABC: "If you go back 25 years, which is roughly the timeframe of the report, and you try to imagine what the internet could possibly do for us in the future, there's no way we could ever imagine.  If we can't know the benefits and we extrapolate what we know today, then the conservative solution, the multi-technology mix, will always look better because it has lower costs."

Response:

  • The multi-technology mix is a preferred solution because it has lower costs and higher benefits. It has higher benefits because it delivers higher download and upload rates (and accompanying opportunities to participate in the digital economy) to households and businesses sooner.  In contrast the FTTP NBN extends the construction timetable during which many households and businesses are left waiting for high-speed broadband.
  • This suggests the critics quoted are completely off the mark: if marvelous new applications that cannot yet be imagined are in reality under development and imminent, the MTM NBN allows Australians to benefit from them far sooner than Labor’s approach.  It also brings forward the date when Australian IT firms can develop applications requiring high-speed broadband, and be confident consumers and small businesses will be able to access them – rather than waiting for fibre nirvana in the never-never.
  • Moreover, the criticism is illogical.  If we are certain that “build it and they will come”, why not build it to twice the capacity and do it twice as quickly?  Ultimately, all investment decisions must balance reasonable expectations of costs and demand – and the fact that this investment is being financed with taxpayers’ money, which could go to lots of other socially important uses, makes that all the more important.
  • Labor’s proposed approach is completely reckless. It assumes the worst that could happen is that demand increase more rapidly than expected. But the study was based on quite aggressive estimates of growth in demand for upload and download speeds – it is equally possible that demand growth will be slower than expected (say because technologies used to compress information develop even more rapidly than projected).  In such circumstances  rushing to invest in FTTP will squander billions in taxpayers’ dollars (bearing in mind the capital costs of FTTP are four times those of MTM).  It is reckless to spend billions of dollars of other people’s money on the assumption that risks only apply in one directiojn – no business would ever do that (or if it did, it would not survive for long!).
  • Nonetheless telecommunications networks are continually being augmented and upgraded.  Many commentators and analysts – such as the ACCC – previously took the view that this would not be possible to do when it came to FTTP.  However, movign incrementally from FTTN to FTTP is technically possible and many carriers around the world are taking a flexible approach to fibre – such as extending it to the node, cabinet or basement – allowing them to upgrade by pushign it further at a later date.  This approach protects investors from irreversible commitments exposed to downside risks while nonetheless ensuring we can secure the upside benefits should demand grow faster than expected.  As the Vertigan panel found, the MTM model is therefore ‘future proof’ – it manages the risks inherent in technological and demand uncertainty. 
  • Rod Tucker argues that it won’t be commercially viable to upgrade later: “With a sunk investment and strained finances, NBN Co will have no incentive or appetite to invest for many decades to come. Fibre will be stuck at the node.”  Again, this doesn’t make sense – if there is demand for higher bandwidth, it follows that future upgrades will make economic sense. Indeed, that is exactly what is happening internationally: telcos use FTTN to get to very high broadband speeds and then deploy further fibre upgrades when and where demand warrants it. Tucker nowhere explains why foreign telcos can do what he believes NBN Co cannot.
  • Additionally, it is wrong to assume the only upgrade path from FTTN and HFC will be through fibre. As the Vertigan Panel was told, wireless technologies are likely to develop that replace the final hundred metres of wireline links, offering very high speeds from the fibre in the street through to the home. While technology is always difficult to predict, wireless options could overcome the cripplingly high costs involved in upgrading lead-ins to fibre. On the other hand this means investments in fibre lead-ins today  could be wasted.
  • The NBN Co Strategic Review looked at the cost of future upgrades and found that deferring investment in FTTP by only five years would deliver benefits of around $2 billion – this comes from bringing forward revenues in deploying FTTN sooner and ensuring that when the investment is made, you are take advantages of improvements made in the intervening period.

  • Additionally, the cost-benefit analysis found that while at very, very high growth in consumers’ willingness to pay (greater than 250 per cent over 10 years) might tip the investment equation towards deploying FTTP initially, that advantage is wiped out once the ability to upgrade from FTTN is factored in. It is always better to deliver higher speeds to households and businesses early and then upgrade if and when additional bandwidth is required.

Write comment