At long last, the government has formalised its response to the Harper Review on the issue of misuse of market power, announcing that the legal test in section 46 will be changed to an effects test. Unsurprisingly, given the amount of debate over the last year, this announcement has led to cheers and boos worthy of a pantomime.

But for all the media coverage, I’ve not seen many attempts to outline, in practical terms, how the proposed change is likely to play out. So in Part I of this piece, I review some key cases to assess how the new test might work. Meanwhile subsequent parts will consider the application of an effects test in a variety of specific scenarios, such as predatory pricing, refusals to deal and supermarket issues. See Part II for refusals to deal and Part III for predatory pricing.

The proposed amendment and its legislative relatives

After more review and consultation processes than Cory Bernardi demands for the Safe Schools programme, the government announced on 16 March that it had accepted the recommendations of the Harper Committee in full on the question of section 46. Exposure Draft legislation is expected to be released shortly for comment (bearing in mind the Harper Report contains detailed drafting in its annexures – one might ask how much more comment is required!).

Whatever their precise form, the announced amendments would change section 46’s current tri-partite structure to a two-step process. Right now, to prove a contravention of section 46, you must show that a firm has substantial market power, that it took advantage of (“used”) that power, and that it did so for an anti-competitive purpose.  

Under the new regime, you’ll have to prove that a firm with substantial market power engaged in conduct which had the purpose, effect or likely effect of substantially lessening competition. It is this second limb that is generally dubbed “the effects test”. Just to add to the confusion, it can also be called the SLC test. The test is already used in various sections throughout the Competition and Consumer Act, specifically sections 45, 47 and 50 (it also assists in interpreting the authorisation and notification tests in sections 88-93).  

The SLC test and effects test
are one and the same

In each of sections 45, 47 and 50, some pre-condition must be met – always involving at least two unrelated parties – before the effects test is applied. For example, in section 50, there must be an acquisition of shares or assets before one applies the test. In the case of section 47, specific supply arrangements have to be in place. Under the proposed section 46, the pre-condition will be that a party has substantial market power – this means that the effects test will no longer apply only to bi- (or multi-) lateral conduct, but also to unilateral conduct.

How is the test likely to be interpreted by the courts?

One of the stated advantages of the effects test is that it is “known and understood”. That is true to an extent – the ACCC applies it ably several hundred times a year in assessing merger clearances, although it tends to struggle to articulate effects cases before the court.

Indeed, there’s reason to worry about the court’s application of the effects test, particularly given the current absence of specialist competition judges. As outlined above, the effects test already exists in several other sections of the Competition and Consumer Act, but these section have rarely been litigated over recent years. Consequently, there is little in the way of “leading” authority, and such authority as exists is very confusing. Indeed, if given just one word to describe the current judicial interpretation of the effects test, I’d choose “unpredictable”.

To give you a flavour, here are three recent leading decisions on the effects test (an absence of cases necessitates a broad interpretation of “recent”).

Rural Press: any lessening of competition is substantial if there’s market power

Rural Press v ACCC (2003) is the only case in which the effects test has been considered by the High Court; intriguingly, it is also the ACCC’s poster child for “proving” why we need an effects test in section 46. Rural Press was very naughty – the ACCC alleged that it misused its market power; engaged in conduct which substantially lessened competition under section 45; AND entered into a market sharing arrangement (illegal per se). The per se and effects claims were made out, but not the misuse of market power.

When considering the effects case, the High Court said that where there is limited prevailing competition (the case was dealing with an acknowledged monopolist), any lessening should be regarded as substantial. In other words, the High Court in Rural Press indicates that a firm with substantial market power could easily trigger an effects test.

While the logic of the effects analysis in Rural Press appears reasonable on its face, the decision has a funny place in history. Mostly considered as a misuse of market power case, few even remember that it includes an effects claim – let alone that it’s the only consideration of the test by the nation’s highest court. To my knowledge, it has not been relied upon by lower courts to determine subsequent effects cases (although the discussion below might give us a clue as to why).

Qantas: it’s hard to imagine how an effects case could ever be successful!

Less than a year after Rural Press, the Australian Competition Tribunal heard an authorisation application regarding the proposed merger of Qantas and Air New Zealand. The Tribunal is clearly not as authoritative as the High Court. But due to its make-up (comprising a Federal Court judge, an expert economist and a lay member), its decisions carry significant weight (indeed several are considered seminal). In Qantas, the Tribunal had to assess whether the benefits of the proposed merger would outweigh any likely “detriment” – detriment being assessed in terms of any lessening of competition.  

Despite Qantas and Air NZ having a combined market share of around 90%, control over terminal space and landing slots, and well-established networks, the Tribunal found that the emerging competition posed by Virgin and Emirates was sufficiently fearsome to constrain the combined entity’s conduct across various trans-Tasman markets. Subject to minor quibbles, it thus concluded that the merger did not substantially lessen competition; so, without really needing to consider the claimed benefits, the Tribunal said the merger should be permitted. (In case you’re wondering why it didn’t proceed, the NZ High Court vehemently disagreed. As the proposal needed approval in both countries, sign-off in Australia wasn’t enough.)

Baxter: the Full Court is more confused than punters trying to work out the allocation of Senate preferences

ACCC v Baxter (2008) is one of just two recent considerations of the effects test by the Full Court (the other being Metcash in 2011; for present purposes, Metcash – albeit a much clearer decision – doesn’t really aid our understanding of how an effects test would work in section 46).

Between first instance judge Allsop J (current Chief Justice of the Federal Court) and three appeal court judges, no clear approach to the effects test emerges. The various judgments descend into a confusing discussion of short-term and long-term assessments of effects with no discernible ratio. This is one decision I advise my students never to read, it’s such a schmozzle. If Rural Press is the soft end of the effects test, and Qantas the hard end, Baxter falls in no mans land. It is unintelligible in a way that sends shudders down the spine of any potential litigant.

So where does that leave us?

Bringing these cases together, what approach is most likely to be adopted under a revised section 46? I suspect that the small business lobby is hoping for a Rural Press approach – any lessening of competition, where there is little to none, should be considered substantial.  

In reality, however, that it is extremely unlikely. The removal of the “take advantage” element means that any causal link between substantial market power and an “anti-competitive” outcome has been removed (see Stephen King’s recent article in The Conversation). So, absent this causal connection, if “anti-competitive conduct” is too easily proven, then we come perilously close to convicting someone for having market power.  

The courts will not do this; nor should they. Our system of economic regulation is built upon the carrot of market power – we want people to strive to obtain it (we’d just prefer that they not succeed). If they do “win the race”, however, we can hardly turn around and punish them. It’s like asking a little kid to solve a quadratic equation and when, astonishingly, she does so, making her write out 100 times, “I must not be a smart-ar*e”.  

For this reason, we are likely to see a strict interpretation of the effects test in the context of section 46. In other words, decisions closer to Qantas. This is definitely not what the government has in mind but, honestly, is far preferable to decisions along the lines of Baxter. That said, we might have to suffer through some Baxter-like first instance decisions before higher courts settle upon their approach.

An added complication is the lack of local expert economists. How is this relevant, you ask? Well, the effects test as applied in Australia is different to the effects test as applied in, say, America (so much for it being settled and well-understood!). In Australian competition litigation, there is a critical lack of home grown experts. Increasingly, parties turn to overseas experts to plug the hole, particularly those from the United States. I’ve been involved in a case where the effects claim had to be dropped because an overseas economist would not (could not?) accept instructions about the Australian legal position as laid down by the High Court. Rejecting the authority of Rural Press, he instead applied the US position (focusing on the efficiency of the excluded firm) and said we had insufficient evidence to run the effects case. Sometimes that line between law and economics is awfully blurry...

The interpretation of a revised section 46 by the courts will of course have consequences for the effects test in other sections (including section 50, the mergers provision). After all, one of the stated benefits of adopting the effects test for misuse of market power cases is that it is known and we can draw upon existing jurisprudence. That cuts both ways – new section 46 jurisprudence will obviously affect the interpretation of the other sections. So if a particularly strict approach is adopted (to overcome the causation problem), then these sections are likely to become harder to prove as well.

The practical application of the test

See Part II for analysis of the potential application of the effects test in relation to refusals to deal, and Part III for predatory pricing.  I'll also address the supermarket space in a subsequent post. 

This is one of a series analysing the Government's response to Harper: links to my other pieces are listed at: https://www.dhirubhai.net/pulse/article/governments-response-harper-review-competition-policy-merrett/edit.  In the meantime, if you have any questions or comments, please email me at [email protected]