Thursday, July 02, 2009

Backhaul and all that

The Department of Broadband etc has issued its request for tender (DCON/09/47) for the regional backbone blackspots program. I was promising myself I wouldn't buy into this you all know I can't help myself. In no guise did I make a submission to the discussion paper and I haven't read those that were made (yet).

There are three parts of this program that worry me. The first is the voodoo economics. The second is about (the lack of) public information. The third is ultimate ownership.

Economics from another planet
The RFT says;

In regional areas where there is a lack of competitive backbone services there is little pressure on a wholesale supplier to offer low prices and higher quality services. This can mean that internet service providers and other service providers have limited scope to make new and/or higher quality services available to consumers in those areas at prices that are competitive, when compared to similar locations that have alternative backbone supply options.

This statement is potentially correct. However, the competitive state of backhaul provision is one of only three potential causes of higher backhaul prices.

The first cause is distance. Clearly the cost of building a fibre from point A to point B is a linear function of the distance from A to B. Most of the cost is in the installation of the fibre and bugger all is in the terminal gear. The further you start from where you need to get the traffic (in the first instance a mainland state capital) the higher the cost, and hence price.

The second is traffic density. A fibre can carry a whole lot of data. But the cost per megabit sent is the cost of the whole fibre divided by the total megabits sent. So if the backhaul is going to a small community the "price of backhaul" will be dearer than the going the same distance to a bigger community.

Only the third is the one mentioned in the RFT. That is the established theory that a monopolist will price above marginal cost, while a competitive market will price at marginal cost. There are a some important provisos though. The theory asumes a rising supply curve, but as we know our fibre will have economies of scale over all possible traffic volumes. The average cost curve slopes down and marginal cost is at all times below average cost. Even the competitive market won't price at marginal cost. The theory also assumes full competition and adding only a second provider means we get at best a duopoly. The theory of duopoly pricing is more complex. The Cournot outcome, which is based on the assumption that providers compete on quantity resolves to a point where prices are still marked up over marginal cost but at a rate equal to the sum of the squares of the market shares of the competing firms (the HHI). Bertrand version is competition on price and ultimately suggests that the provider who can achieve the lowest cost will eventually win the entire market. In other words pricing will reach the competitive model but once it does one firm ceases to exist and you get monopoly again.

Far more troubling are the conclusions of both an experimental (from slide 44) model or a more correct mathematical one, that competitive firms don't price at marginal cost. In one description of this tacit collusion is an emergent property of the dyanamic market. Through repeated iterations the two firms realise they can maximise profit by sharing the market at the monopoly level of output.

At this point empirical cases will be raised. That wherever new backhaul has been built the existing backhaul prices drops. Well that might be true - but unfortunately all too often the price drops to short run marginal cost as the incumbent is now prepared to treat fixed cost as sunk. It is a noted failure of competition policy - notably anti-competitive conduct laws - that this pricing behaviour by the incumbent is not drawing any consequence. But the second failure of competition law is that the transmission service is already a declared service. The ACCC has faffed around on pricing but ultimtely the Minister could simply use his determination powers under Part XIC to set a transmission (backhaul) price.

In fact, there is an even simpler version to incent the incumbent to price at long run marginal cost. That is to make a competition law that a monopoly provider cannot drop their price AT ALL once someone else enters the market. To deter efficient entry which would make the incumbent asset redundant the monopolist has to prce to the Chicago School assumption of long run marginal cost pricing for entry deterrence.

So building the additional infrastructure is not waaranted to reduce the monopoly pricing - there are better ways to do that. Building the infrastructure won't otherwise reduce average cost it can only increase it.

Lack of information
The only logical reason for a fibre backhaul program is therefore providing fibre where there is none now. That is the case on some of the proposals - e.g. fibre that goes through Barcaldine shire. However the citizenry whose money s being spent here doesn't know where there is fibre and where there isn't. This is despite the Howard Government's Broadband Blueprint proposal;

Access to efficiently-priced backhaul is important in a competitive market because it allows new carriers to provide a service in places where they do not own their own networks. The more competitive the backhaul networks the lower the prices
that will be paid by consumers. In the past, commercial sensitivities have prevented sharing of the extent of networks across Australia, particularly backhaul. For the benefit of all providers and government these issues need to be resolved as far as possible. For the benefit of all providers, the Australian Government will continue to work with industry to develop an interactive map of backhaul supply.

The theme was repeated in the Framework for the Future report from the Regional Telecommunications Review;

2.5.3: The Australian Government
a. regularly collect and prepare records of backhaul infrastructure for use by other Australian Government agencies for public policy purposes, and
b. assess the costs and benefits of making this information available to relevant market participants.

The Government Response was poor in this regard noting;

The outcome of the National Broadband Network process may have implications for the issues identified by this recommendation. Accordingly, the Government will consider this recommendation once the outcome of the National Broadband Network process is fully known. This will enable outcomes of the process to be factored into the Government’s consideration of this recommendation.

So the Government will decide AFTER the backhaul tendering process whether to obtain and release the data that would be useful BEFORE it. Sounds like sound public policy (not) to me.

Finally we come to the position of ownership. The RFT comments;

The Commonwealth's preferred model is that the Commonwealth will own the fibre and other passive/physical network elements, which may be transferred to the NBN operating company (NBNCo).

This I don't understand. We are investing in "competitive" backhaul but we may transfer that to the new monopoly access company (yes it will be a network level monopoly - I'm cool with that. I don't want two fibres.) The last place this backhaul should wind up is in the hands of NBNCo.

So there we have it, a tender designed to build new backhaul to reuce the cost (huh?) to places that we the public don't know have how much fibre (three fibres leave the NT already - one to Broome, one to Mt Isa and one dwn the guts through the Alice to Adelaide)and to be owned competitivly by the new monopoly. Oh and in the midst of all that the actual price per Mbps or MB on the new fibre is part of the tender process - but it won't be the same as, say, Gosford.

Hopefully more logic will inform the Implementation Study. But as I write that I note that there does not xist anywhere a statement of the scope of the "Implementation Study" - the most we have is the information in the REOI. First the timetable;

The indicative timetable for the NBN implementation study involves:
a) Engagement of Lead Adviser for the implementation study late June 2009;
b) Engagement of further specialist advisers by the Lead Adviser in consultation with the Department in late July 2009;
c) Commencement of the implementation study mid 2009;
d) Interim implementation study report during second half of 2009;
e) The completion of the final implementation study report in early 2010; and
f) Government consideration of the implementation study report following completion of the implementation study in early 2010.

The REOI was still quite vague about the scope of the implementation study;

The Department is seeking to appoint a high calibre recognised expert as Lead Adviser to provide high quality and timely advice throughout the National Broadband Network implementation study and to manage the production of an integrated implementation study report, that ensures that all issues are appropriately covered. It is expected that the implementation study will examine and provide advice on a wide range of commercial, financial, project management and governance and telecommunications issues relevant to the National Broadband Network.

Though the services required went into a bit more detail about the implementation study;

The implementation study will be multi-disciplinary and will need to include analysis and recommendations encompassing:
a) Advice as required in support of proposed legislation relating to the operation and governance of the network company, the regulatory regime, and ownership restrictions for retail telecommunications providers and other investors as required;
b) Advice on the overall funding requirements for the network rollout (quantum and profile) beyond the $4.7 billion initial funding injection;
c) Development of strategies to maximise the scope for private sector investment in the network company, subject to appropriate ownership restrictions and appropriate terms and conditions for participation;
d) Advice on the optimal capital structure for the network company over time;
e) Development of detailed commercial/financial and engineering analysis of the network roll-out and the implications for the network company;
f) Advice on how best to structure NBN Company arrangement from the outset so that the Government’s long term objective of privatisation can be accommodated;
g) Development of plans for the integration of the Tasmanian operation and backhaul network into the overall national broadband network;
h) Network design consistent with the Government’s objectives;
i) Development of strategies to provide procurement opportunities for local businesses;
j) Develop a detailed implementation plan for the roll-out of the National Broadband Network;
k) Development of recommendations as to the appropriateness of any foreign ownership restrictions for the network company;
l) Development of a risk management strategy for the national broadband roll-out; and
m) Stakeholder consultation.

So the Department in charge of the Implementation Study in issuing a request for expressions of interest for a lead adviser successfully communicated that it did not at that time have a fully detailed specification of what the study actually is (and to any extent prioritised network design as the EIGHTH item on the list), a read of the REOI left one unsure as to whether the Study was being conducted by the Department with the assistance of the lead adviser or by the lead adviser as a report to the Department. Maybe the Request for Tender issued to only a shortlist made that clear but I don't think that is public.

Hopefully all will be clear soon though as the REOI indicaed that the services of the Lead Adviser would commence in week commencing 6 July.

No comments: