Why the ACCC should NOT authorise the NBN-Optus arrangement


The ACCC has released a draft determination to authorise NBN-Co acquiring Optus’s cable customers and the decommissioning the Optus HFC network. The document is an interesting read – but in my opinion, reaches the wrong conclusion.

As background, the Optus HFC cable was originally rolled out in the mid-1990s and currently passes about 2.4m homes in Brisbane, Melbourne and Sydney. It has about 0.5m subscribers. See the draft determination around paragraph 2.9 and following for details. The Optus cable has been upgraded so that its broadband services are about the same speed as the ‘entry level’ NBN services (see paragraph 2.22).

Given the Telstra-NBN deal, the Optus cable is the only fixed-line alternative to the NBN. The NBN wants to eliminate this potential source of competition and according to John Durie, Optus gets $800m from the deal. This will make Optus very happy. I have no idea what is the book value of the cable – but it would be relatively low. Optus lost the Telstra-Optus cable wars in the mid-1990s and its cable network has presumably never recovered its original cost. But that is history (or in economic terms, sunk costs). The ACCC evaluation must be on the future costs and benefits of the NBN-Co – Optus arrangement.

Overall, the ACCC gets the analysis spot on. It recognises that the NBN project will continue with or without the Optus arrangement. And in the absence of the arrangement, Optus will continue to be a limited competitor to the NBN but will probably, over time, have its customers move to the NBN anyway (see paragraph 3.60).

So why do I disagree with the ACCC’s conclusion? Remember that the test for authorisation is that the public benefits outweigh the public detriments. The ACCC argues that this test is meet. But in my opinion, the main public benefit that drives this conclusion is simply not a public benefit.

The ACCC’s lists three public benefits from the agreement. The latter two – some cost savings due to orderly migration and some environmental benefits – appear trivial. So the draft decision relies on the first public benefit summarised in paragraph 3.138.

The ACCC considers that the HFC Agreement is likely to result in benefits to the public from reduced duplication of infrastructure expenditure as a result of the cessation of operational and capital expenditure on the HFC network which is greater that the expenditure which NBN Co would incur in serving these same customers. The ACCC considers that these benefits are likely to be material.

Sounds good – so why is it wrong?

Remember that the Optus wires are already up and the NBN wires will be rolled out regardless of the ACCC’s decision. So the duplicated expenditure is NOT in the basic (natural monopoly) infrastructure but in operations and maintenance. Optus will keep on servicing its customers so long as the benefits to those customers outweigh the cost (i.e. they can charge them more than the operations and maintenance (O&M) costs). Those customers will have the option of the NBN but will choose not to take it. So those customers must value the Optus service at more than the O&M costs that could be saved by those customers moving to the NBN.

Why will they value the Optus service at more than the cost? No idea! Why do some people like red cars and some people like black cars? Why do some people like apples and others like oranges? Or, in the case of the NBN and Optus, why do some people like Royal Gala apples and some like Pink Lady apples? No idea! It is their preferences. But we know these values and preferences by observing the customers’ behaviour. And if the market shows us that those customers prefer to use Optus’s network – and to pay more than its on-going costs to keep using it – despite there being a perfectly good NBN alternative, then the economic conclusion is clear. The benefits to those customers outweigh the costs.

In brief, if the market keeps the Optus cable alive – even if only temporarily – then any ‘benefit’ from reduced expenditure of closing the Optus cable prematurely is necessarily outweighed by the loss to those customers who would have chosen to buy the services of that cable. So the main ‘public benefit’ is not a benefit at all.

The ACCC has fallen into the trap of ‘market design’. Suppose I can show you that it is cheaper to make just one type of car. After all, then we could have one big factory and save duplicated production costs. You might argue against this as we would have a monopoly and car prices would go up. So let’s get rid of any price effect and ask the following: is closing down all other car operators a benefit even if the car prices do not change?


Customers like choice and their actions in the market place show that they value that choice at more than any cost savings. As Henry Ford showed, if uniformity is cheap enough then many people will be happy with uniformity. So the cost savings of uniformity could outweigh the benefits to consumers of choice. But we need a market test to show this. And on the ACCC’s own analysis, Optus will continue in the absence of the agreement with the NBN-Co. So on that basis, there is no public benefit of avoiding on-going O&M costs. Rather there is a public loss because Optus’s customers, who for whatever reason prefer to stay with Optus rather than shift to the NBN, lose their benefits of that choice.

Should this be a lost ‘benefit’? Of course! It is the most basic economic benefit that there is. Consumers have preferences and buy things that they individually value at more than the market price. That is the source of all consumer surplus. So the NBN-Co – Optus arrangement will lower consumer surplus on the basis of the ACCC’s analysis.

The ACCC notes a number of public detriments from the NBN-Optus arrangement including a loss of competitive tension (see paragraphs 3.199 to 3.202). But my argument above is not about competitive tension. It is more basic. It is about consumer surplus and is simply based on revealed preference.

Should this reverse the ACCC decision? Yes! The Tribunal has noted that the relevant net benefit must be material and not trivial (se paragraph 3.207). But if the ‘benefit’ of avoided on-going O&M costs are recognised as a loss of consumer surplus rather than a ‘benefit’, then either the benefits of the arrangement are outweighed by the costs or the net benefit is trivial. So the ACCC should decline the authorisation.

In summary, the ACCC analysis misclassifies a benefit. By forgetting the most fundamental role of markets – to shift products to people who value those products at more than they cost – the ACCC has misclassified a loss of consumer surplus (a detriment) as a cost saving (a benefit). It is an easy mistake. But it needs to be rectified in the final decision. Which means that the arrangement should not be authorised and the fate of the Optus cable should be left to the market.


17 Responses to "Why the ACCC should NOT authorise the NBN-Optus arrangement"
  1. I don’t even know where to start, except that the idea that keeping Optus Cable installed is of no consequence is entirely wrong.

    I doubt Optus ever made money off HFC cable (you admit this), but that’s not a sunk cost. It costs them hundreds of thousands of dollars a year to keep active.

    The difference between revenue from users and HFC maintenance will only become worse with the NBN. What thinking person would choose Optus cable over the NBN?

    The NBN destroys any case for Optus cable to exist, so it’s only right that they compensate and effectively purchase the asset from Optus.

    If your concern is that the ACCC is forcing Optus to sell, this is only prudent to avoid court action in future.

  2. The car metaphor doesn’t really hold, another opposing but probably equally flawed metaphor could talk about power/water/gas and whether we believe there is value in having multiple providers of the base infrastructure instead of just retailers on top.

    I believe the real reason behind the deal is that NBN co wants pricing ubiquity but also has to roll out to areas where there is no HFC cable and no economic incentive to lay one.

    If optus was able to undercut the NBN by only offering services in these denser areas, then that leaves NBN two options:

    1) Drop their price in that area to compete and make other areas feel ‘ripped off’
    2) Don’t drop their prices and make people everywhere feel ripped off by NBN.

    What it really comes down to, is whether there is value in having ubiquity in access and price to NBN’s networks even though that means that dense metro areas essentially subsidise less dense areas.

    Personally I believe it makes sense to have that enforced monopoly provider who gives access to all, rather than leaving it to private competition and having black spots in the network where it’s not worth private providers treading.

    But this is the great debate that was missed in the early NBN days while the opposition and media where trying to claim wireless was the way of the future.

  3. Is the concern that the NBN will be pricing at average cost rather than its lower marginal cost, and so some customers will be driven to Optus (who will price closer to marginal cost) when they actually would have preferred the NBN if they had faced true marginal costs.

  4. Thanks for your opinion, Stephen, however, I find the base concept in your article to be flawed.
    Optus customers, and all customers for that matter, will get plenty of choice because they choose the Retail Service Provider through which they will to provision their Internet service.
    When you send a parcel, you can choose whether you’d like to send it through any one of the numerous couriers that operate in Australia and each Service Provider will charge a different fee based on the level of service they provide to their customers, but you don’t have a choice as to the roads on which they drive the delivery vans. It would be absurd to suggest that each courier company (Retail Service Provider) should build, operate and maintain their own private road network to every premises.

  5. Steven, your article seems to hinge on the assumption that Optus HFC offers a consumer choice when compared against the NBN wholesale network.

    That’s not entirely straight-forward, seeing as Optus will be offering services over the NBN.

    There’s actually a very good model that can be used for comparison though: the existing ADSL/cable ISP market. Optus competes against itself in this market – providing retail ADSL over Telstra wholesale copper with its own DSLAMs to addresses right next door to the HFC cable. After an original price disparity, it no longer differentiates over the delivery method – broadband is broadband – and simply charges for additional speed where it’s available.

    It’s hard to conceive of market conditions where Optus would offer a different price between NBN fibre and its own cable, except if HFC was substantially cheaper to operate the service on. (Economics aside, that seems unlikely from a technical perspective.) NBN would also be capable of offering speed above that of HFC, and would have sufficient bandwidth to provide cable television. Optus would have a whole separate network without any remaining differentiation in service or in price.

    The main differentiation is that HFC is available now, and NBN would have to wait for fibre to be laid. Given the long time scale NBN has for deployment, and the vertical integration (and monopoly) that Optus currently has on HFC, there’d be a good commercial incentive for Optus to keep operating the network until the NBN is built. After that point, Optus had no legal obligation to keep HFC running if it didn’t turn a profit next to the NBN.

    So, yes, had the ACCC ruled differently you might get to run an interesting economics experiment in consumer choice against identical services. Strangely the ACCC was less fascinated with that, and ruled in favour of giving existing HFC consumers the choice between Optus and its competitors for retail services without waiting for the NBN fibre roll-out.

  6. Let me respond to a number of comments that appear to have fallen into the same trap as the ACCC. In economics we take consumer choice (or preferences) as our given. Consumers then make market decisions based on those choices and firms that meet those choices prosper. Consumer surplus (or the value consumers gain from transactions) is the net benefit to consumers from that choice (how much would they have been willing to pay versus how much they actually paid). So if:
    1. a consumer makes a choice in a a market; and
    2. the seller voluntarily offered the relevant product,
    then there is necessarily a gain to the consumer that outweighs the cost to the producer.

    The ACCC authorisation test is simple. Do the gains outweigh the losses? The ACCC has counted the cost saving by stopping Optus offering services as a gain. The ACCC has NOT counted the loss in consumer surplus. So they have counted as a net gain something that the market tells us must be a net loss.

    For those of you who wish to design the market (“consumers can have the choice the government lets them have” or “they will have enough choice, why let them have more?”), just think about the implications of what you are saying. Interventionists who hate markets make these claims all the time. Want to buy a foreign car? No – there is ‘enough choice’ from Australian made. Want to buy ‘brand-name’ beans? No – home brand gives you enough choice. Want to choose your own doctor? No – the state gives you enough choice, so take who we tell you to.

    That is why I call the error the trap of ‘market design’. It is far too common in public policy to have advocates of one side of a position argue that ‘they know best’ for the consumer. Limiting consumer choice may appear to ‘save costs’ as in the NBN-Optus arrangement. The cost savings are easy to measure. But if consumers are willing to pay those costs – as reflected by their actions in the market – then those ‘cost savings’ are outweighed by a loss of consumer benefits. How much? No idea. Consumer surplus is hard to measure. But there is an overall loss – that is what revealed preference in the market place tells us. To call that ‘loss’ a ‘benefit’ is wrong economics.

  7. Stephen,

    In this case the consumers are the ISP’s which have not being taking up anything much with Optus. Probably because Optus hasn’t been offering it!

    If Optus couldn’t compete with Foxtel on a reasonably even basis how would it compete with the NBN for ISP’s with inferior speeds.

  8. The draft determination suggests that the consumers who might stay on the HFC would cost the economy more (para. 3.82). But as Stephen points out, a reason we know for sure that such consumers would choose to stay is that they value it more. Stephen actually went to great lengths to describe why any other reasons for such a choice are irrelevant for the purposes here: if they choose to stay its because they prefer it over the NBN. Simple logic, not abstract economics. Therefore, if such consumers would have made the decision to stay, and for whom some cost to the economy would be saved by approving the arrangement, they are also those who must lose something if their HFC choice is taken away. An economic analysis cannot measure one and not the other.

  9. On your Marx – Nooooooooooooooooo.
    ISPs are intermediaries – not ‘consumers’. ISPs have no ‘preferences’ in the sense the word is used in economics. Consumer surplus is derived from final consumers only. ISPs will maximize profits by doing what the final consumers want them to do.
    And on the basis of competition, Optus is still operating. So its on-going profit (i.e. ignoring the massive sunk cost of the HFC cable which is irrelevant for any future decisions) must be non-negative. That means that some (about half a million) consumers prefer it to the alternatives (Foxtel, ADSL, wireless, reading newspapers, whatever!). So there MUST be positive consumer surplus that outweighs the avoidable costs. If not, Optus would have shut down the cable services years ago.

  10. Stephen,

    The ONLY consumers for the NBN are ISP’s.

    They are to be the only wholesaler.

    We then get our stuff from the ISPs otherwise known as the retail market.

    You and I wil lnot be buying from the NBN but from optus, TPG etc

    The speeds on the optus cable are slower than the NBN and have little potential to go faster unlike the NBN.

  11. It does seem to be a very strange decision for the ACCC to make and contrary to basic economic reasoning.

    The mere fact that NBN is willing to pay $800 million simply to destroy a bunch of cables is evidence in itself that this is a bad decision.

    The combination of what appears to be universal prices to the consumer due to political imperatives and the servicing of very high cost areas means that unit costs under the NBN will be high. Optus would presumably be able to offer a substantially cheaper service for basic NBN services, which is all most people will want for the first decade.

    There is some value to the NBN in customers it would not have had if Optus were still operating. But i doubt it’s $800 million (which is $1600 per current Optus subscriber). Instead, it will make its money back by forcing the consumer to pay more than it would have in the absence of this decision.

    The ACCC seems to completely ignore lower prices for these services as a public benefit.

  12. On your Marx – No. As a matter of economics a consumer refers to an economic agent who gains ‘utility’ from a good or service that they consume. A consumer has preferences and may gain consumer surplus from his or her purchases.
    An intermediary is an economic agent who purchases inputs, applies technology and sells outputs to either other intermediaries or consumers. An intermediary may make profit (which becomes income for consumers who own the company’s shares) but an intermediary CANNOT gain consumer surplus. The only consumers in the economic sense of the word of NBN services are the final consumers of internet services.

    So the only BUYERS of NBN services will be ISPs. The only CONSUMERS of NBN services will be those people who consume NBN services as part of a product that they purchase (probably from an ISP).

    This is not just a semantic difference. it is at the heart of all economic analysis of the costs and benefits of market interactions. The ACCC decision is incorrect because it neglects the consumer surplus created by the Optus network.

  13. I am sorry Stephen I was a little late in understnding how theoretical you were being.

    I can tell you as an Optus customer that they do not want to use their cable when the NBN comes around.

    As I said the speed in limiting. Fine for now as it is competitive in speed in terms of ADSL2 etc but not later.

    This is why as I stated earlier Optus is not really in the wholesale market.

  14. I’m not sure I understand. Stephen, you argue that the ACCC decision is wrong, because you believe that consumers who choose to stay on Optus’ own cable network are deriving some benefit, thus we should not discount the lost utility should they be forced on to the NBN.

    It seems that ACCC thinks this utility is zero, and I would concur.

    Having a quick looked at Optus cable internet plans, they appear comparable, indeed almost identical, to current NBN plans? From the Optus website they have the 120GB plans at $59.99 for NBN, and $59.99 for cable. If I recall, there is some dodgy deal about having a phone line with the NBN, but Optus customers will get discounts for bundling etc.

    So let’s continue with the assumption that consumer surpluses from the ‘choice’ (or really, the lack of motivation to realise you can change your internet plan) of Optus cable are zero once the NBN is available, then we simple have a cost on Optus during a transition phase to NBN.

    Under such a reality your claim seem strange.

    “… if the market keeps the Optus cable alive – even if only temporarily – then any ‘benefit’ from reduced expenditure of closing the Optus cable prematurely is necessarily outweighed by the loss to those customers who would have chosen to buy the services of that cable.”

    But I doubt that Optus would keep servicing customers on its own cable once the NBN was rolled out in that street. Why would they incur the ‘duplication expenditure’ highlighted under para 3.138 given they are also an NBN retailer – it makes no sense. Just force your customers to switch.

    It seems that this is more of a political sweetener for Optus, given the government take-over of the fixed-line internet market. And on what economic grounds can we argue with political strategy?

  15. You entire article is ill researched and full of dogmatic rather than realistic economic theory.

    You claim optus passes 2.4m homes, it passes 1.4m premises as shown in the ACCC release.

    The government allowed Telstra to compete with Optus causing huge overbuild and asset write offs for both companies. What it got us were to very sore companies that have no intention of repeating the debacle as a competitive environment with entire competing networks duplicates costs while not duplicating the consumer base to justify it. This is why a NBN is a natural monopoly in Australia, especially when we consider it on a national scale.

    Now NBN Co. has a responsibility to cross subsidize both the less profitable FTTH areas and the Wireless/Satellite areas. If Optus were able to eat away at the most profitable areas we are basically saying we want NBN Co. to compensate by charging higher prices to other areas. You cannot have competitive network, especially when one network has social and nation building obligations while another can cherry pick.

    We should all know that in real life a market is not perfectly mobile due to many reasons such as contracts and laziness to find a better deal. I mean how many people by brand name product just because they have not tried generic? If the market was able to recognize the benefit of the NBN without switching to it first this would be much easier, however, there is no underestimating ignorance.

    Your car analogy is completely incorrect, as analogies usually are. With data communications the differentiating factor is speed, price and reliability. The consumer should not care which technology delivers or whether the cable is coated green or black. The fact is the NBN pricing is similar or better and it provide a better service, along with this these consumers will support cross subsidization of less profitable areas in order to bring ubiquity of service to fruition.

    I personally wish this contract was not necessary, ideally it is $800 million that would not be spent, however, in order to support long term viability and to ward off an imperfect and uninformed market it is necessary. Who knows, maybe if Tony Abbott and Malcolm Turnbull actually told the truth about this network it would not have been necessary, the sheer amount of disinformation out there is staggering.

  16. Many of the comments above seem to have missed the central point Prof King makes. Absent authorisation, Optus is likely to continue to operate its HFC (if not, then no need for the agreement anyway), in which case there must be a net surplus to the society (ie consumer benefit is greater than producer cost). This should be counted as a loss given the ACCC’s decision. Instead the ACCC considers it a gain arising from avoided opex of HFC.

    In my view, the ACCC does not count the surplus as a loss probably because they assume those consumers will end up getting services on the NBN in the absence of the HFC (ie same consumer benefit). Therefore the main difference between factual and counterfactual is the cost differential between the HFC and NBN of servicing those customers who would have otherwise stayed with Optus. Given the NBN has a lower cost, the ACCC concludes there is a saving (or benefit) to the society.

    Would this be a fair characterisation?

%d bloggers like this: