Bannerman Competition Lecture 2016

Summary: The 2016 Bannerman Competition Lecture was delivered by Dr Phillip Williams AM, Head of Legal and Competition Practice at Frontier Economics Australia.

Dr Williams' lecture, QCMA, forty years on reflected on the Queensland Co-Op Milling Association Limited and Defiance Holdings Limited (QMCA) matter, which the Trade Practices Tribunal rules on in 1976.

Published: 11 February 2016.

1 Introduction

It's a great honour for me to be giving this second annual Ron Bannerman memorial lecture. I first met Ron when I was a teaching fellow at Monash University immediately after finishing my honours year there. Although I was very junior I was invited to join Maureen Brunt, Bob Baxt and Jack Fajgenbaum who were teaching a seminar for law and economics students on trade practices and Ron was always invited to the Christmas party which was held every year at the end of the at the end of course.

On the last occasion I met Ron he reprimanded me; he had been retired for some years but I encountered him in the ground floor of the Federal Court building in Sydney. Even in his retirement he took a great interest in, the decisions of the Federal Court on Competition law and he came up to me expressing dismay at a recent decision by a Federal Court judge - I've forgotten the case and I've forgotten the identity of the judge. I agreed with Ron that the decision was a disaster. Ron urged me to write and publish a criticism of the decision. I replied perhaps too flippantly that if I wrote a criticism of every bad decision of the Federal Court I wouldn't have time for my academics. Ron thought that this flippancy was quite irresponsible and he told me so. As Maureen has mentioned Ron had many great qualities but one was this very strong moral code that he attended to apply to his own behaviour but he also attempted to apply to other people that needed a bit of encouragement such as myself.

I've agreed to talk about QCMA and Defiance Holdings. QCMA as almost everybody in the room will know arose because two companies, Queensland Cooperative Milling and Defiance Holdings, applied for authorisation for mutually exclusive merger proposals – each wanted to acquire Barnes Milling Limited. Each of them had been denied authorisation by the Commission and they were appealing to the Tribunal. The Tribunal rejected both appeals, however its reasoning differed markedly from that of the Commission and the reasoning of course that's important. What I want to do this evening is talk about some of the principles of competition and markets that were developed in that decision and discuss their relevance 40 years on.

Most of the key propositions that I'll talk about were drafted by Maureen Brunt, probably on her old typewriter with annotations in green biro. We have strong evidence that Maureen wrote these propositions. First is that those of us who have been privileged to have been taught by her will detect her drafting style and her thinking without much hesitation. Secondly, the president of the Tribunal at the time Sir Edward Woodward said as much in public, when he retired as Chancellor of the University of Melbourne. He confessed that his career as a judge was remembered chiefly for one decision (QCMA) and furthermore that the bits that were quoted from that decision were written by him but rather were written by the Maureen. So in giving the Ron Bannerman lecture this evening I'll be honouring not only Ron Bannerman but also Maureen Brunt.

2 Principles concerning competition

2.1 The analysis of effects on competition is central to the test for authorisation

The first proposition I want to talk about is this - that the analysis of the effects on competition is central to the test for authorisation. Now this might not seem a very surprising proposition to people in this room, but it was a new proposition at the time of the decision. The old section 90 (5) which was the test that the Tribunal was obliged to apply didn't mention the word competition and merely talked about public benefit, and there was a submission by Mr. Brennan – QC as he was then – to the effect that the Tribunal's analysis of competition should be secondary to its consideration of public benefit. Brennan appeared putting this submission for the commission. Mr. McComas who was representing QCMA put the opposite opinion – his submission was that the test in section 90(5) should be read in the context of the act as a whole and particularly with respect to section 50(1) which was of course the merger provision.

The Tribunal, happily for us, accepted the submission of Mr. McComas and rejected that of the Commission and said quite roundly that they were going to focus on competition as a way into this analysis of public benefit and they thought that Mr. McComas was quite right in saying that the provision that they had to apply should be read in the context of the act as a whole. So this set the ground for not only that decision but of course subsequent decisions and of course now the test is all the tests are slightly different and do mention the word competition.

2.2 Competition constrains market power

The second proposition which again won't seem so surprising - I get to more surprising propositions as we go on - is that competition constrains market power. That is, competition for market power is the opposite, if you like, of competition and market power is the freedom from market pressure.

I've always liked the decision of the Tribunal - consisting of different personnel of course - by that stage John Lockhart was the Tribunal chairman but Maureen was honoured in media council number two. There, the Tribunal had to consider a set of rules by which various players in the media, in particular advertising agencies and television networks, decided that they would engage in a system of what the Tribunal described as private regulation of advertisements that appeared on television. The Tribunal said that although these rules undoubtedly led to significant market power because they substituted a system of private regulation for market pressures, they agree that certain reservations basically authorised these rules because they thought that they were offsetting public benefits.

What I've always liked about the decision is the way that they characterised market power as a system of private regulation. I will later in the talk refer to the McHugh case - I should say right from the outset that although McHugh had a very different approach from the Tribunal and Media Council, I should confess that I was called by the unsuccessful applicant in that case. I'll just say a few words about it later on, but in McHugh the Federal Court did come to the conclusion that, although the rules that they were looking at which were relating to the rules that the racing club's had not to allow as to be registered as thoroughbreds horses that were bred by artificial insemination, although this was a system of private regulation, the court found in that particular case that those rules didn't substantially lessen competition. I'll explain why a bit later.

2.3 The nature of competition is influenced by the structure of the market

A slightly more surprising proposition - at least surprising to people who have been trained more recently in economics - is that the nature of competition is influenced by the structure of the market. I'm sure almost everybody in this room will remember the five elements of market structure that the Tribunal pointed to in QCMA as needing to be analysed as bearing on the nature of competition in the market. The decision in QCMA is very much written in the light of this structure-conduct performance schema, which used to be found commonly in industrial organisation textbooks but is no longer there. That doesn't mean that this structure-conduct performance schema or a way of looking at things has been discarded. Indeed, any model of game theory which is used in current analysis of competition is based on certain building blocks of standard models, like that of Cournot Homogeneous products or Bertrand Differentiated products or joint profit maximisation. Each of these models makes certain assumptions about the structure of the market that they're analysing and then draws certain conclusions as to the conduct and performance of the market that will result.

So the Cournot model assumes a given market concentration, assumes a homogeneous product, assumes in effect blockaded entry because it doesn't allow for any entry, and then concludes as a result as that there will be a particular mark-up of price on marginal cost as a result. So these are models, which were standard of course at the time that QCMA was written and are standard today. They are the standard building blocks on which we construct all these game theoretic models - in effect infer conduct and performance from structure, so when the Tribunal in QCMA were instructing us that it's important to think of structure as determining key elements of market conduct. It seems to me that it is quite uncontroversial as a matter of economics.

2.4 Competition is a rich concept that cannot be reduced to a simple formula

Some criticism of the structure conduct-performance schema seemed to me to be based on a misunderstanding of what people like Professor Brunt and Professor Scherer were telling their students when they were talking about the links between structure, conduct and performance. Professor Brunt, in her published lecture notes for students who are undertaking the course that Bob and Maureen taught, pointed out (as of course did Scherer's textbook), that there are of course feedback loops from conduct to structure, so there was no question that they were supposing a very simple minded one-way pattern of causation from structure to conduct performance. They were quite alert to the fact that people/businesses by their conduct can influence the structure of the market. This is a proposition that I think is too often neglected by lawyers and economists who practice in this area. The Tribunal stated that competition is a rich concept that cannot be reduced to a simple formula, and I'll read you a quote.

He said this –

‘Since we give such importance to the relevance of competitive considerations in proceedings for authorisation, we had a few comments on how the Tribunal views competitions as the introduction to the famous being that it is known’.

But this is the important sentence that I want to draw your attention to:

‘However competition is such a very rich concept containing within it numbers of ideas that we should not wish to attempt any final definition which might in some markets settings prove misleading, or which might in respect of some future application be unduly restrictive. Instead we explore some of the connotations of the term’.

The Tribunal proceeded to talk about at least five different ways even in this passage as to how you might think about competition depending on the circumstances or the problem that you were analysing. So it seemed to me that people sometimes read QCMA as this list of structural factors and think if they just deal with those structural factors they've somehow analysed competition, whereas the Tribunal in QCMA was proposing a much more subtle approach, and said quite explicitly that the proper approach will of course be dependent on the problem that's to be analysed.

2.5 Competition should be considered over time

The next proposition is that competition could be considered over time. Many of you will remember that after the Tribunal lists these five structural factors they then say this - of the five factors that the Tribunal talks about undoubtedly the most important is number two - the condition of entry for it is the ease with which firms may enter which establishes the possibilities of market concentration over time, and it is the threat of the entry of a new firm into a market which operates as the ultimate regulator of competitive conduct. The Tribunal decision in the Cooper Basin case really hinged to a large extent on this very long time horizon. You might recall that the Commission urged, in the person I think of Stephen King if I remember correctly, urged the Tribunal to take a much shorter view of its consideration of competition than the Tribunal eventually took.

The Tribunal took the view in the AGL Cooper Basin decision that it really should take into account the long-term effects of a decision that it might make as to whether contractual arrangements which were undertaken some decades earlier (two decades earlier) should be allowed to continue or should not. Of course the most remarkable decision that we have of jurisprudence on time horizon is the decision by Justice French in the first AGL case which I think everybody would regard as a tour de force. The decision by Justice French was to a large extent contingent upon taking a long view of the time over which competition should be considered in a similar way to that proposed by the Tribunal in QCMA.

Justice French in his decision had this to say:

‘The ACCC has made subsequent submissions about price spikes said to derive from economic withholding by LYP. I am prepared to accept that there are periods of high demand where a generator may opportunistically bid to increase the spot price. I do not accept that such intertemporal market power reflects more than an intermittent phenomenon, nor does it reflect the long-run phenomenon having regard to the possibility of new entry through additional generation capacity and the upgrade of interconnections between regions, does not amount to an ongoing ability to price without constraint from competition’.

So we have a lot of authority both in the form of the decisions by the Tribunal and of this famous, I would say, decision by Justice French in the AGL case that a long time horizon should be considered when analysing the effects on competition of particular conduct. I think however we should accept that proposition at the same time balancing by the mourning of the Tribunal that we don't want to reduce competition to a simple formula.

In particular I don't think we should ever say that easy entry solves all competition problems. My favourite example for students was the Net Book Agreement which was the subject to an inquiry in the United Kingdom prior to the books case that Maureen referred to in her remarks. The Net Book Agreement was an agreement that involved both horizontal agreements among all the publishers and also horizontal agreements among all the retailers that all the retailers in the United Kingdom would charge the prices recommended by the publishers. Do some of you might be old enough to remember they used to be on the backs of British books that used to say Net Book. They used to have a price and then say net. The net was the reference to the Net Book Agreement. The first book to be netted under the Net Book Agreement was Alfred Marshalls 'Principles of Economics', published in 1890 and then the Net Book Agreement continued. It was enforced quite effectively for about a hundred years even though entry both into retailing and into publishing of books was relatively free.

Indeed the freedom of entry was probably bad because the excess profits in the retailing and the publishing were eliminated not by pushing prices down, but rather by so fragmenting the market among such an inefficiently large number of publishers and retailers that their unit costs rose and the excess profits were eliminated that way. So it was doubly bad - it was bad because the prices were very high and that restricted the number of books that were published and sold and furthermore, it was bad because the market was fragmented and so there was productive inefficiency as well. So I think the Tribunal was right to say yes; consider particularly perhaps merger cases over a long run time horizon but that doesn't mean that they threw overboard their proposition that you shouldn't reduce competition to a simple formula.

2.6 The identification of competition is assisted by considering the performance of a market.

I said perhaps the most controversial thing I'm going to say but I think, well it's something that I think has changed over time since the decision of QCMA. The proposition is put rather tentatively in QCMA that the effect of conduct on competition is assisted by considering market performance. I think this is standard economics. I have a very favourite quote in the written version of my paper, a quote by a Frank Fischer to the effect that economists, when looking at effects of competition on markets, will always examine the effects of conduct on both structure and performance.

Indeed the textbook on Massimo Motta which was perhaps the best textbook at the moment on the economics of Competition law, defines market power according to the learner definition which is basically the mark-up of price on marginal cost. Indeed in Professor Brunt's lectures that I've referred to earlier, that she is to distribute to the Monash trade practices class, she said exactly the same thing; that when looking at the effects of particular conduct on competition I will want to examine the effects on structure, conduct and performance. However when it came to QCMA, the Tribunal was a bit more cautious. When it came to QCMA, the Tribunal had this to say:

‘Competition may be valued for many reasons as serving economic, social and political goals but in identifying the existence of competition in particular industries or markets we must focus upon its economic role as a device for controlling the disposition of society's resources’.

That's the closest that went to saying that performance was relevant. Now the reason for this rather tentative statement was undoubtedly that the Tribunal was aware that, certainly at the time, there was a debate about the extent to which they were able to consider performance when they were considering effects on competition, because there was dual structure in the act that seemed to imply that performance was something that could be considered under authorisation proceedings under the heading of public benefit, whereas competition being different could be considered by the court when considering the question of liability. Professor Brunt was quite aware that this duel enforcement structure might narrow the test for anti-competitive conduct so as to exclude considerations of market performance.

She stated in her lectures (I kept a copy), she said this:

‘It may be, [so she's not nailing herself to this proposition], that the existence of the dual enforcement system that is judicial enforcement coupled, with provision for case-by-case administrative exemption on grounds of public benefit, will serve to narrow the tests of anti-competitive effect before the court. It may be that where authorisation is available the court will be drawn to a standard of pure market power in association with evidence bearing on market rivalry. But monopolisation and price discrimination are not subject to authorisation and hence the court might find it difficult to strike down practices which make some obvious contribution to good economic performance’.

So she was aware that there was this very tricky line that the Tribunal had to trade when dealing with competition; on the one hand this duel enforcement structure seemed to indicate that certain matters were available to the courts, but other matters were reserved for authorisation matters. Now the law may have moved on a bit and I know many of you will know that I sometimes act as a bit of a lawyer as well as an economist but you'll have to forgive me. In particular recent statements by the Tribunal starting with Chime Communications (No 2) are quite definite that when considering competition, at least within the context of a decision by the Tribunal, they will consider the effects of conduct on structure, conduct and performance. Indeed tribunal in the person of Justice Finkelstein as it was, made exactly the same statement in the Pilbara case. We have one statement by the court, not such a clear statement, but the process of reasoning it seems to me by the court in McHugh is very, very similar. In McHugh, as I said before, the matter before the court was: were these rules of racing that all the racing club's in Australia subscribe to that prevented horses bred from artificial insemination as being registered as thoroughbreds for the purpose of racing and thoroughbred races.

A key passage in the courts reasoning was this. I'll read it out and what I wish to say is this, that I think what the court was saying there was this: although this was a system of private regulation as distinct from market-based forces, the rules were justified, sorry I shouldn't say that, the plaintiff had failed to establish that the will should be struck down, because the plaintiff had failed to establish that they significantly damaged the performance of the market.

This is what the court had to say in McHugh: It follows from my findings that the applicant has failed to establish the claimed increasing competition in the breeding market (there were different markets that pleaded) from the removal of the restrictions in that market, that the act that sorry, start again.

‘It follows from my findings that the applicant has failed to establish the claimed increasing competition in the breeding market from the removal of the restrictions in that market, that the applicant has not established that there would be the flow-on effects in acquisition market leading to an increase in competition in that market for which he contended’.

It's a complicated sentence but the next sentence is the important one.

‘Further, I am not satisfied that if artificial insemination were permitted in Australia, there is likely to be an increase in the number of high-quality yearlings product for sale, or that prices were likely to be lower for high-quality yearlings. I am not satisfied that there would be a significant demand for artificially insemination bread thoroughbreds in the thoroughbred acquisition market, or that international or domestic purchase is seeking to breed or race in Australia only, would be likely to purchase artificially insemination bred thoroughbreds. I am therefore not satisfied that giving effect to the immune provisions has the effect or is likely to have the effect of substantially lessening competition in that market’.

So it appears that both the Tribunal and possibly, if McHugh is followed, even the court might be prepared to accept evidence about the effect of conduct on performance of a market as indicative of the effect on competition in the market. I think so far as the Tribunal goes, perhaps it's quite clear as so far as the court will go, we have to watch this space a bit.

3 Markets

3.1 Markets are analytical devices

Only a couple of remarks about markets - of course the decision in QCMA is often quoted as authority for what is meant by markets and how one goes about approving them, but the Tribunal in QCMA is quite cautious when it comes to markets. In particular it accepts the proposition that Mr. Brennan put to the counsel for the Commission, when Brennan said that mere specification of markets cannot be determinative by itself of some ultimate issue. The Tribunal quoted that with approval. Of course it went on to define market as a field of actual and potential transactions, which led Justice French in the Taprobane case to say make a statement which you said any time since that a lot of these words that we have in economics, like market or whatever are really metaphors which I think is certainly true.

3.2 Markets and sub-markets take into account substitution in both demand and supply

The Tribunal introduced in QCMA also the notion of submarkets, because markets according to the Tribunal in QCMA are merely analytical devices, not determinative of any ultimate issue. Some markets are perfectly acceptable. The objection that some lawyers have to the use of submarket language arises because those people believe that market definition can determine ultimate issues in litigation. But that wasn't the view of the Tribunal in QCMA and it certainly, I would say, is not the view of a decision of the Full Federal Court in Singapore Airlines and Taprobane. In both QCMA and in Singapore Airlines, which is the decision I like to get my students to read when they're talking about markets. Markets are analytical devices used as part of the reasoning process by which one explains the likely effects of particular conduct on competition. So to speak of some markets is neither here nor there it's just something that's bit less than the market.

4 The future of QCMA

Finally, just some remarks about QCMA. The rule of law of course means that, or it implies that old cases live on. In the United States jurisprudence and competition law cases as early as Addyston Pipe and Trenton Potteries are still in the technical - students are still required to read them, and for very good reasons. Of course fashions come and go in economics probably rather more rapidly than they do in the law, as we've seen the language of structure, conduct and performance has now been got rid of from the economics textbooks even though we still reason from structure to conduct to performance.

A lot of the principles, in particular the principles that I've talked about this evening, of competition in markets in QCMA are not things that come and go. Many of the principles are derived or can be found in Cournot's famous book in the 1830s; many of them are certainly in Marshall's Industry and Trade, which Rod was probably brought up to when he was an undergraduate student at Melbourne. This is not a criticism of Rod but is a bit of a criticism of the people who taught him. But, and indeed probably the latest principles that can be found I derived from Kaysen and Turner's very famous book on Antitrust Policy. Carl Kaysen was of course one of Maureen's supervisors at Harvard so it seems to me that these propositions are not new, weren't new at the time, they're quite old and really what the Tribunal was doing in QCMA was really laying down fairly standard propositions that any good economist would have adopted at the time and indeed any good economist would consent to now. So it seems to me that QCMA has lived for four years, you'd be a very brave person to predict how much longer they're going to continue, but it seems that the principles should indeed continue. Thank you very much.