Tuesday, September 06, 2011

Telstra ongoing embarrassment from its CEDA foray - Updated

One of the observations Dr Phil Burgess made to me about Australia was the lack of engagement by Australian business in political debate. He set about changing that.

Most particularly he made it clear to think tanks and research institutes that he was only interested in funding organisations that supported the views that Telstra espoused.

The Committee for the Economic Development of Australia (CEDA) was one body that heard the message. CEDA commissioned a report that was ultimately released as Growth 60: Australia's Broadband Future - Four doors to greater competition. It is not totally available on line but many libraries have copies.

I am not suggesting that CEDA was funded by Telstra to undertake the research, however the bulk of the contributors were at some stage paid by Telstra for writing similar pieces elsewhere.

Confusingly the CEDA link to the report dates it as "Wednesday, November 18, 2009", however, the one chapter available online dates the report as 2008. The introduction also refers to "current policy" as FTTN to 98% of the population.

Having been in the audience for the launch (and at one stage considered as one of the authors) I know the report was released at the time of the conclusion of tenders for NBN 1.0 in November 2008.

This date is supported by the excellent review by Stuart Corner. He noted;

Leading economic think tank CEDA (Committee for Economic Development of Australia) has criticised the Federal Government's obsession with building the National Broadband Network, saying a subsidised $4.7 billion FTTN rollout is unnecessary. CEDA's report, Australia’s Broadband Future: Four doors to greater competition, maintains that focussing on FTTN to the exclusion of competing access methods will decrease competition.

The CEDA report contains six chapters written by eminent economic researchers, all of whom argue among other things excessive regulation and structural separation of Telstra is unwise, while the promotion of competing access networks including cable, copper and wireless would provide the necessary competitive marketplace.

Yet CEDA as recently as April 2011 has been peddling this report as if it is informative about the current policy position.

This is compounded by the fact that the AFR continues to use the report's commissioner, Michael Porter, to write opinion editorial pieces on the NBN; the latest offering is Fibre rollout wastes billions. These pieces ignore the reality of what happened in the NBN 1.0 tender and make fanciful and inaccurate technical claims.

I will move to a systematic response to the column in a moment. My first point is though that CEDA and Porter's interest in this matter was spurred when Telstra wanted NBN 1.0 stopped. In particular Telstra did not want any solution that involved their structural separation - the key reason they declined to bid.

Now Telstra's position is totally different. They would like to get on with their business and deploying NBN 2.0 is what they want to achieve. David Thodey has made it clear that shareholders should not vote on the NBN and separation deals on the assumption that there is an alternative policy. It might be worthwhile for Telstra to have another quiet chat with CEDA.

Moving to Porter, he asserts that "by avoiding fibre overbuild in metropolitan and regional areas, HFC can indirectly finance up to $20 to $30 billion of NBN savings."

This assertion simply does not stand up to scrutiny. HFC exists mostly only in metropolitan areas, there is very little in regional areas. HFC passes somewhere around 70% of Australian households, but it is a well known fact that it has extremely low penetration into multi-dwelling units. That is a technological limitation due to the difficulty of feeding a whole coax cable into existing spaces.

The total NBN construction cost of $43B includes the cost of the satellite and wireless solutions. The implementation study provided detail (below) of the cost per premise connected by percentile.

Even if we assumed that HFC did cover 70% of households, these are the households it costs $2 to $3K to connect. It is noteworthy that it was this analysis that saw the FTTP footprint extended to 93% from 90%. That means that Porter's upper bound of saving $30B, being three-quarters of the capital cost, is totally beyond reach. $20B may be reasonable to estimate the amount of capital not spent.

But here's the other problem - the outcome is not comparable. The 100Mbps speeds being quoted for HFC are shared speeds - not what is available to each premise. To provide a comparable service over the life of the asset requires additional investment. Further, as mentioned above, HFC simply is not available at all the premises it passes.

Porter asserts

Most independent experts would prefer NBN Co to request tenders by location and speed, not technology. Regional markets, with support from the private and public sectors, could sort and blend bids and technologies, and deliver content competition.

and concludes

To be commercially viable, NBN Co should invite all parties to bid to build upgraded and expanded cable, wireless and fibre in metropolitan areas. It will then start to have a better balance sheet owing to savings on fibre rollout - estimated at well over $20B.

The difficulty is that neither Telstra nor Optus - the owners of the two big HFC networks - have any interest at all in this strategy. It doesn't deliver comparable service and does not defer forever the need for a fibre to the home investment.

It is interesting to note that Porter's latest contribution only credits his university post. It is unclear from the CEDA site whether he is still also CEDA National Director Research and Policy. (Update - I have been informed that Porter is no longer with CEDA. His replacement is Nathan Taylor, Chief Economist). CEDA could profitably leave its 2008 report - properly dated - on its website and delete the more recent blathering.

Novae Meridianae Demetae Dexter delenda est


ian said...

There is little point in logically commenting on most of the NBN critics. They are like the climate change deniers.

The difference is that NBN is well defined problem, and there is no reason to question the technology recommendations of the eminent experts group. (The implementation may leave some room for criticism.)

Apart from the penetration, HFC is not a viable long term solution. Its depenance on exposed mechanical connectors is a problem waiting to happen - in some of the older US networks cable struggles to deliver TV. Optus had to take special measures to ensure the SNR was adequate for telephony, and this was on a brand new rollout.

One major factor none of the critics mention is the cost of OPERATING a network. OPEX swamps CAPEX in the log term. Mixed technology networks are just a recipe for escalating OPEX.

Stuart Corner said...

The "excellent review" of the CEDA paper attribute to me was in fact by my colleague, Stan Beer.