Recently every pet shop parrot has been squawking “structural separation”, with scant reference to real facts and data.
Those who want to fulfil their own commercial aims by sacrificing Telstra follow the old theory that if you repeat misinformation often enough it becomes fact, especially in the complex area of regulatory policy.
We are repeatedly told that “separation” has worked in the UK, NZ and Ireland, playing on the old cultural cringe that anything done by one of these countries must be good for Australia.
The fact is that none of these countries has undertaken “structural separation”, that is, where the network is sold and operated by a company independent of the parent. Both the UK and NZ have opted for forms of functional or operational separation where new business units have been formed - OpenReach at BT and Chorus at Telecom NZ - to operate and manage the network mimicking a standalone business, but still reporting to their respective CEOs in BT and TNZ, and staying in company ownership.
The third leaf in the clover is Eircom, the incumbent operator in Ireland. Supporters of structural separation trumpet eircom as the model to follow because its owners, Australian private equity firm Babcock & Brown, requested the company be separated. In this case they wanted to sell the retail arm and retain the network operations.
But Babcock & Brown has withdrawn the structural separation idea. They have bigger problems to deal with. Eircom is encumbered by huge debts following two highly-leveraged debt-financed buy-outs, the second by Babcock & Brown. One analyst has estimated the debt to asset ratio was 117 per cent (lirne.net) at one point.
Eircom also desperately needs investment in its fixed networks. It has the second lowest maximum broadband speed in the OECD – just 3Mbps. BT is not doing much better, ranked 22nd out of 30 OECD (www.oecd.org) countries for maximum speeds offered by incumbents.
Babcock & Brown have gone to the Irish government seeking €150 million to assist with a network upgrade, but the government has knocked them back. BT has no plans to upgrade its access network with fibre, and the Government has launched an inquiry (www.theregister.co.uk) to find out why investment isn’t happening.
In the meantime, Telecom NZ has announced a very modest “cabinetisation” program in order to ward off structural separation and get a form of operational separation they can live with. This program would see the rollout of ADSL2+ services with roughly equivalent speed and population coverage to that already available in Australia – when it’s done.
Hardly models worth following, but it doesn’t stop SingTel Optus engaging a former regulatory staffer, now with a analyst group, to co-author a report which recommends – you guessed it – structural separation.
The report (www.cecg.com.au) didn’t get much media coverage, and anyone reading beyond the executive summary could see why. It is a curious piece of analysis which actually admits to the many efficiency benefits of having an integrated network provider, and that structural separation adds costs and disruption, but erroneously concludes there isn’t any other way to ensure competition on the network.
It patches over the fact that even the OECD is advising regulators to “actively consider all other policy options” because of the apparent effect of separation on investment. The paper does this by completely ignoring the investment question, wishing the network into existence and then talking about how to regulate a built asset.
The paper also gives a couple of misleading examples to support its arguments. It says that even within an operationally separated model, Telstra has exhibited “anti-competitive behaviour”. An example it uses to support its case is a Competition Notice issued by the ACCC against Telstra in April 2006 in relation to wholesale line rental. The paper fails to note this allegation was thrown out by the Federal Court and the judgment found that the ACCC had acted illegally in issuing the notice by denying Telstra procedural fairness and natural justice.
It also uses Singapore as an example of a structurally separated NBN tender in support of a similar approach in Australia. This is totally inappropriate and misleading. As CEG points out in the final lines of the paper, “there are a number of other operators with local fibre loops” in Singapore. Interpreted, this means that the Singapore tender is not the conversion of a national basic access network into an NBN, but a contract for a third high-speed network in addition to networks owned by Singapore Telecom and Star Hub. In this context Singapore Telecom has not been structurally separated, but has decided to take a strategic investment position in the new network. It is also important to point out that the Singapore government has committed AUD$800 million to the project, which will serve 4.5 million Singaporeans living in an area of 700 square kilometres, less than a half that of the Sydney urban area.
In this debate, rhetoric is too often structurally separated from the reality.