Where now for infrastructure access?


The Full Federal Court recently made its decision in an appeal about infrastructure access under Part IIIA of the Competition and Consumer Act. The specific issue relates to rail access in the Pilbara. But the key concern coming out of the decision is the Court’s interpretation of some of the declaration criteria.

Section 44H.4 of the Competition and Consumer Act sets out the criteria that must be satisfied before a facility can be ‘declared’ for third party access. Subsection (b) states that declaration can only occur if the Minister is satisfied:

that it would be uneconomical for anyone to develop another facility to provide the service.

This is meant to be a ‘natural monopoly’ test. And that is how it has been interpreted in recent years. However, the Court has now interpreted this subsection on its plain English meaning: Would it be privately profitable for anyone to develop another facility to provide the relevant service. Under this interpretation, the test is neither necessary nor sufficient to capture natural monopolies.

Part IIIA is the Australian ‘access regime’ for upstream facilities. It is part of the Competition and Consumer laws and developed out of the Hilmer report on National Competition Policy in the early 1990s.

For access, what the Hilmer report was interested in capturing was essential facilities. These have two characteristics. First they embody a natural monopoly technology. This means that it is more efficient (in the sense of minimising cost) for all relevant production to occur through one facility than multiple facilities.

The second characteristic is that the facility is necessary for competition. In other words, there are no alternative inputs to the service provided by the relevant facility that can economically be used to provide the relevant final goods or services.

Note that a facility that is a natural monopoly is not naturally a monopoly. In other words, other investors can and sometimes do build ‘duplicate’ facilities. This was common in the 1800s when investment in rail, canals and toll roads was relatively unregulated in some countries such as the UK and the Americas. Sometimes competitors built competing ‘natural monopoly’ facilities, often resulting in a price war that bankrupt one party and led to the duplicate facility being taken over by the other party. Indeed, this ‘destructive competition’ was the basis for government intervention, for example in the UK.

Arguably, the rolling out of duplicate and competing cable networks by Telstra and Optus in Australia in the 1990s is an example of this type of inefficient infrastructure duplication. Part IIIA aims to prevent this duplication by allowing competitors to get access to the ‘natural monopoly’ services. Rather than inefficiently building their own duplicate ‘essential facilities’, Part IIIA helps competitors access the existing facilities.

The relevant test in Part IIIA to capture services produced by ‘natural monopoly’ infrastructure is in s.44H.4.b of the Competition and Consumer Act:

that it would be uneconomical for anyone to develop another facility to provide the service.

This test has had a rather ambiguous history. It was initially interpreted on its plain English meaning. Would ‘anyone’ find it profitable to develop another facility? The ‘anyone’ might be the specific access seeker or anyone else. The test of profitability was often on a naive basis – at current prices would it be profitable? This has little to do with natural monopoly and often had more to do with the issue of how deep are the access seeker’s pockets. Deep pockets? Yes, you can develop.

The matter reached a head in an access dispute over railways lines in the Pilbara in the late 1990s. The access seeker made it quite clear that, if it didn’t get access, it would build its own railway line for its own use. But the existing railway line looked exactly like the sort of ‘natural monopoly’ facility that Part IIIA was meant to capture. So the National Competition Council made public its view that ‘criterion b’ was a natural monopoly test, and it has stayed so ever since.

The Australian Competition Tribunal kept to this (economically sensible) interpretation in the recent dispute between Pilbara miners FMG, BHPB and Rio.

However, the Full Federal Court has altered that interpretation.

The Parliament chose to frame criterion (b), so that it directed attention, not to whether the NCC or the Minister or the Tribunal judged that it would be “economically efficient” from the perspective of society as a whole for another facility to be developed to provide the service, but whether “it would be uneconomical for anyone” to do so. The perspective of this phrase is that of a participant in the market place who might be expected to choose to develop another facility in that person’s own economic interests. It is at this point that we are constrained by the statutory text to part company with the Tribunal (para 76).

To see why this interpretation makes little economic sense, consider the following:

  1. It may be profitable for a company to develop a duplicate ‘natural monopoly’ facility even though this is socially wasteful: The duplicate HFC cable networks in Australia provide an example of this. Whether or not duplication is profitable depends on the beliefs about post-entry competition. If the incumbent will ‘accomodate’ rather than ‘fight’ the entrant, then (inefficient) entry may be profitable.
  2. It may not be profitable for anyone to develop another facility to provide the service even though there is no natural monopoly issue: To see this consider standard free entry in a market with facilities that involve fixed costs. Entry will occur until it is unprofitable for anyone else to build a new facility and enter. There need not be any natural monopoly – free entry gets you to this point where excess profits are competed away. So even in highly competitive markets the test under criterion (b) may be satisfied.

In other words, the criterion (b) test, as interpreted by the Full Federal Court, may fail to capture natural monopolies and may capture facilities that are not natural monopolies.

Now, I am not criticizing the Court. The judgement is well written and makes the case clearly. The Court is just going on the plain reading of the Act. But this interpretation makes no economic sense. If the interpretation stands, it undermines the rationale of Australia’s essential facility access regime.

It is time for parliament to act. It needs to fix this part of our competition laws to make them clear. Otherwise Australia risks being left with a dysfunctional and potentially harmful access regime.

9 Responses to "Where now for infrastructure access?"
  1. It’s time statutes are written in mathematics and for mathematics to take over law. Being an ex-law student I’ve been unfortunate enough to endure what jurists call ‘legal reasoning’ ala rhetorical gymnastics. You might not eliminate ambiguity completely but at least arguments can be summed up in a couple of pages at most. It would be the greatest innovation in legal history, boosting productivity and eliminating underserved rents overnight.

  2. One interpretation I have heard of Part IIIA is that it decentralises the selection of when to regulate, confining to a selection of circumstances ‘that matter’. So under this view, access is only sought where it has some prospect of benefit, from the perspective of a market participant. But as you pointed out Stephen, its a different matter to then identify among these which are natural monopolies, and requires a different perspective. The Court’s interpretation seems to imply the first perspective be used for the second, in economic terms. Perhaps one way forward is to explicitly recognise a view like this about ‘decentralising’ the decision of when to regulate?

  3. DavidN
    “It’s time statutes are written in mathematics and for mathematics to take over law.”
    Whilst that idea appeals to me as a mathematician (I could have a new career as a lawyer) it doesn’t address the underlying problem, which is one of poor drafting rather than of ambiguity of plain English.  In fact, I think it would make poor (mathematical) drafting more likely.

  4. If only poor drafting was the most egregious issue with law …

    Alternatively, instead of judicial review you can have review by a ‘high panel of (independent) economists’ to decide on competition issues and the like. We have already have a panel of economists determining the fate of the Australian economy, so in a way deciding on IO issues would be a step down.

  5. woops, just realised the Australian Competition Tribunal is like the panel I was suggesting.

  6. DavidN,
    Poor drafting is clearly the problem in this case.  The words “economically inefficient” rather than “uneconomical” should have been used.

  7. I wish it were this simple. Pages of math, yes, but to adequately explain their meaning you need to use English anyway. Perhaps just a page could work on sub-additive costs, but then again, you need to explain in English was this means (as Stephen has). Replacing the words ‘economically inefficienct’ with ‘uneconomical’ does nothing for the apparent problem with the Court’s interpretation. Such a problem still arises with an interpretation of the relevant test being ‘economically inefficient from the perspective of a market participant’. And Stephen’s examples of errors would remain.

  8. I was thinking along the lines if you replaced legal arguments with mathematical arguments it would disqualify presumably most lawyers and judges, and they would have to be replaced with people who can ‘read’ mathematics and interpret it as abstracts of reality ala economists. Of course physicists and mathematicians could do it too but I think economists would have a comparative advantage with issues relating to social interaction. This way you can work around the issue of ‘interpreting’ what every single word means (part of what I consider rhetorical gymnastic). Instead of legal precedents you could use journal articles etc. etc. But I was only semi serious and it clearly isn’t a realist idea. I still stand by my view that poor drafting is the least of the problems with law, generally speaking and not limited to this specific case.

  9. Stephen, TPI/Fortescue and the NCC have applied for special leave to appeal the Full Court’s decision to the High Court. I would not however expect “mathematical argument” to be involved.

%d bloggers like this: