22/07/2021

In its 2020 antitrust complaint against Google, the US Department of Justice took aim at one aspect of Google’s conduct that will be recognisable to anyone involved in competition law compliance training:

Google did learn one thing from Microsoft – to choose its words carefully to avoid antitrust scrutiny. Referring to a notorious line from the Microsoft case, Google’s Chief Economist wrote: “We should be careful about what we say in both public and private. ‘Cutting off the air supply’ and similar phrases should be avoided.”

Moreover, as has been publicly reported, Google’s employees received specific instructions on what language to use (and not use) in emails because “Words matter. Especially in antitrust law.” In particular, Google employees were instructed to avoid using terms such as “bundle,” “tie,” “crush,” “kill,” “hurt,” or “block” competition, and to avoid observing that Google has “market power” in any market.

The “notorious line” was reported by Intel vice president Steve McGeady, who testified in November 1998 that Microsoft’s Paul Maritz had explained to him that Microsoft would “cut off Netscape’s air supply” by giving its own Internet Explorer browser away for free.  Maritz denied saying any such thing, and it doesn’t appear in any of the courts’ opinions.  But it was a key part of the DOJ’s case and its framing of Microsoft’s conduct.

Google’s Chief Economist since 2002 has been Hal Varian, and his advice to avoid phrases like the one attributed to Microsoft seems hard to argue with now.  It’s not clear whether Varian was responsible for all of the documents called out by the DOJ – many of which were obtained and published by technology site The Markup – which read more like they were written by lawyers.

They advise readers to “assume every document will become public” and avoid any suggestion that products are “sticky”, users are “locked in”, or markets, products or resources exhibit a “network effect”, a “scale effect” or any kind of “leverage”:

Using the word “leverage” may make you sound like you went to business school, but it implies exploitation and an absence of consumer choice, which is not what we’re about.

As well as market power, they encourage readers to stay away from markets altogether:

For lawyers and economists, “market” is a loaded concept based on whether changing the price for one thing would lead people to buy a different thing ... For example, a field like advertising sweeps worldwide, and includes lots of different products and services. Defining a “market” more narrowly, or even defining one at all, can create issues if economists aren’t involved. There’s no problem with referring to a “market segment”, “sector”, “business” etc, but just try and avoid defining the “market” for what we offer.

They also counsel against other loaded terms and recommend more neutral ones: not “barriers to entry” but “challenges”; not “unmatched” or “unsurpassed” but “valuable”; and not “unique” but “new” or “alternative”.

All of these warnings would presumably be red flags for the DOJ – a second order of wrongdoing designed to conceal the first.

For its part, Google has said:

These are completely standard competition law compliance trainings that most large companies provide to their employees. We instruct employees to compete fairly and build great products, rather than focus or opine on competitors.

I hope I’m not going too far out on a limb to say that these do in fact look like many of the compliance training materials that I’ve read or written over the years, providing guidance on how to comply with the law and ensuring that pro-competitive conduct is not mischaracterised.   Compliance programs are generally seen by courts and regulators as a good thing; they can significantly reduce breaches and even demonstrate a culture of compliance that may be taken into account in determining penalties. For example, in the CSR case Justice French said that a penalty assessment should consider, among other factors:

Whether the company has a corporate culture conducive to compliance with the Act, as evidenced by educational programs and disciplinary or other corrective measures in response to an acknowledged contravention.

In 2019 the US Department of Justice amended its leniency policy to include consideration of “the adequacy and effectiveness of the corporation’s compliance program at the time of the offence, as well as at the time of the charging decision”.

It’s less common for a compliance program to be condemned as an effort to avoid antitrust scrutiny, as it is in the Google complaint. 

Of course all compliance programs are designed to avoid antitrust and other regulatory liability.  The main focus of these programs is on what a company does; and the Google documents dutifully affirm that the company should always compete on the merits and never exclude competitors or lock in customers.  But they also have to address what the company says, if for no other reason than that regulators and even courts pay so much attention to language, and there’s a risk that the wrong word will lead to enormous costs in money, time and reputation even where there is no wrongdoing.

Cut and thrust

Business people are an invaluable – and often the most valuable – resource for understanding what’s going on in an industry, a transaction, or a course of conduct.  But they tend to speak the language of business, not of law or economics.  And that language is naturally adversarial, even violent, in its tone and its imagery.  It can quickly evolve new and idiosyncratic meanings for existing terms.  And businesses are endlessly confident of their own success and desperate for every sale.  There are plenty of opportunities for things to be taken the wrong way.

In Australia, the High Court recognised in Queensland Wire that:

Competition by its very nature is deliberate and ruthless. Competitors jockey for sales, the more effective competitors injuring the less effective by taking sales away. Competitors almost always try to “injure” each other in this way. This competition has never been a tort … and these injuries are the inevitable consequence of the competition s.46 is designed to foster.

The US Seventh Circuit Court of Appeals in Ball Memorial Hospital v Mutual Hospital went further:

Competition is a ruthless process. A firm that reduces cost and expands sales injures rivals – sometimes fatally. The firm that slashes costs the most captures the greatest sales and inflicts the greatest injury. The deeper the injury to rivals, the greater the potential benefit. These injuries to rivals are byproducts of vigorous competition, and the antitrust laws are not balms for rivals’ wounds. The antitrust laws are for the benefit of competition not competitors.

That assessment was endorsed by the High Court in the Boral case, which also noted that:

Competition damages competitors. If the damage is sufficiently serious, competition may eliminate a competitor.

When the courts’ language is this gory with ruthless slashings, deep wounds and fatal injuries, it’s not surprising that businesses can be similarly colourful in their descriptions.

It’s also clear that firms and industries develop their own language to describe the world as they see it – sometimes by inventing their own words, but more often by adapting existing words to their own uses.  These may be buzzwords or terms of art that are taken from other disciplines or have been separately adopted and adapted by other disciplines – including business school, law and economics.

Looking at a familiar term, Geoffrey A Manne and E Marcellus Williamson note that:

“Market” for business purposes can mean: product, brand, segment, sector, customer base, customer group, customer type, channel of distribution, city, state, country, region, area of responsibility, or corporate division. Because of the multitudinous variations, how a business person uses the term “market” is meaningless for antitrust purposes.

Finally, both internal and external business communications include a lot of aspirational language or downright hype – they’re all selling products to customers, and this extends to pitching projects to managers or the board.  Justice Middleton recognised this tendency in the TPG/Vodafone case:

In that regard, out of court public statements of business people in connection with their enterprises come from a belief that their business is the greatest in the world, and without such a belief businesses would probably not prosper. Subject to the constraints of the law, public expressions of belief in one’s own business are to be expected and are part of marketing.

The result is that business may use the same language to describe either vigorous competition or exclusionary conduct; either a view of the world unique to that business or a recognised legal or economic concept; either a healthy belief in their product or a ruthless campaign of market domination.  Of course a regulator or a court will need to consider the language in its context and determine the most likely meaning; but there is surely value in choosing the right words in the first place.

Words and deeds

This isn’t a new issue.  In his 1968 article “The Naughty Words of Antitrust”, ex-FTC counsel John Loughlin found an increasing focus on statements and documents at the expense of economic analysis:

Careless words, written or oral, rather than elaborate economic presentations are often a decisive factor in antitrust cases, strongly influencing, first, the decision of the Government to file suit and, later, the decision of the courts on the merits …

Whereas corporate antitrust review was formerly concentrated on business practices, it must now critically be extended to a wide variety of intracorporate reports regarding competition, acquisition, new products, business development and planning, press releases, annual reports and other communications to the shareholders, and even statements to securities analysts.

In 2005, Manne and Williamson considered that this was still a concern:

In response to the burdens of principled antitrust enforcement, courts (and regulators) have sought to relieve the complexity of economic analysis by relying instead on business documents to prove antitrust violations. These documents are relatively easy to obtain, easy to digest, and replete with seemingly relevant information. Yet they are fundamentally flawed. They are written by business people, for business purposes, and their translation from business to law (and economics) is frequently untenable.

Much of the language that regulators seize on most enthusiastically – and compliance program warn against most sternly – is language that can suggest a purpose or intent.  But intent has a complicated position in competition or antitrust law.

In the US, intent can be a necessary but usually not a sufficient element of an antitrust breach.  For example, under section 2 of the Sherman Act, the offence of monopolisation requires the willful acquisition or maintenance of monopoly power, ostensibly involving both purpose and effect; while attempted monopolisation requires an intent to monopolise and also the dangerous probability of success.

Even so, in the Microsoft case itself the Court of Appeal offered a very limited role for intent:

[I]n considering whether the monopolist’s conduct on balance harms competition and is therefore condemned as exclusionary for purposes of § 2, our focus is upon the effect of that conduct, not upon the intent behind it.  Evidence of the intent behind the conduct of a monopolist is relevant only to the extent it helps us understand the likely effect of the monopolist's conduct.

In Australia, by contrast, the main competition law prohibitions can be satisfied by an anti-competitive purpose alone – even where that purpose has little chance of being achieved.  As the Full Federal Court said in Universal Music v ACCC:

The paragraph is satisfied if the relevant corporation has the requisite purpose, regardless of whether or not that purpose has been, or was or is likely to be, achieved. It may conceivably be satisfied even in a case where the Court finds the purpose could never in fact have been achieved; although that finding would be relevant in determining whether to infer the proscribed purpose.

However, in practice the Australian courts have been careful not to move too quickly from provocative language to an anti-competitive purpose.  For example, in Eastern Express v General Newspapers the Full Court found that, even if Fairfax executives had made comments to magazines suggesting they would drive a competitor from the real estate display advertising market:

the bellicose imagery employed in both interviews is more indicative of swagger, braggadocio and the presentation of a ‘strong’ image to readers of the magazine in question, than of the existence of a purpose proscribed by s. 46.

In Boral, the High Court looked beyond a number of “smoking gun” documents suggesting that Boral had increased capacity to drive at least one competitor from the market for concrete masonry products.  The Court warned of “the danger of confusing aggressive intent with anti-competitive behaviour.”

And in Pfizer, the Full Court examined repeated references to Pfizer’s “blocking” strategies against generic pharmaceutical manufacturers in its internal documents – particularly in early drafts – but similarly cautioned that “the use of particularly colourful language has to be understood in context” and found no anti-competitive purpose, including because Pfizer had no prospect of substantially hindering its generic competitors.

But while bellicose imagery and colourful language often has an innocent explanation, it may take a long and expensive trip to the final court of appeal to defend what could easily have been avoided.

Compliant compliance

It’ll be interesting to see whether the DC District Court or any appeal court say anything about the DOJ’s characterisation of Google’s compliance program – and whether the DOJ makes more of the issue at trial.  The complaint includes a few extracts from e-mails and draft presentations in support of allegations that Google had deliberately excluded competitors: aspirations of “world domination” and the desire to “protect [its] market share from erosion” and “protect Search exclusivity”.

These seem pretty tame in the scheme of things – barely a gun, let alone a smoking one.  Would the DOJ really argue that there would have been more incriminating documents if not for Google’s compliance program?  Or that the compliance documents themselves suggest an anti-competitive purpose?  Surely it will stick to arguing that Google’s conduct was anti-competitive in effect and infer any necessary intent from the circumstances.

But the DOJ’s complaint usefully emphasises that compliance documents themselves need to stand up to court and agency scrutiny – perhaps even more than other internal documents, since one of their key functions is to demonstrate a culture of compliance to a court or agency. 

I feel that the Google compliance documents referred to in the complaint are pitched about right.  The message is: these are our unobjectionable policies, let’s make sure we describe them accurately.  Or, as Nina Simone and later The Animals have summarised:

I’m just a soul whose intentions are good –
Oh Lord, please don’t let me be misunderstood.

Misunderstandings can arise when documents lose their nuance and take the blunter form of “Don’t say this; say that.”  That can sound like a prescription for euphemism or even deception – at least that’s how the DOJ seems to have interpreted it.  It may be better to stick with “Don’t say this if you really mean that”.  That’s a more difficult message to communicate, especially through slides, but can lead to a more profound compliance and may be less likely to be taken the wrong way.

Of course now we’re drifting into a language guide for language guides, and it’s easy to imagine a future regulator complaining that the next Google has hidden its true intentions in an Inception of compliance programs within compliance programs.  So let’s leave it there for now.

In other news …

• In the Epic Games action against Apple and its app store rules in Australia, the Full Court has reversed the lower court’s decision to move the case to Northern California as preferred by Apple’s developer agreement. The ACCC intervened in the appeal, citing the public benefit in having competition law cases heard in Australia to provide clearer precedent and guidance to Australian courts, and the Full Court found there were strong reasons in public policy to hear the case here.  Apple has said it plans to appeal the decision, which would involve seeking leave from the High Court.
  The ACCC has found that mobile prices have increased since the merger of TPG and Vodafone, which the ACCC opposed but the Federal Court approved. Although price increases were accompanied with an increase in data allowances, the ACCC found that the most users didn’t need the extra data.  TPG and other carriers have challenged the ACCC’s assessment. 
   
• Back in May, twelve professional football clubs briefly formed the European Super League intending to replace or rival EUFA Champions League, bringing back memories of the slightly more successful Australian Super League – which survived a season and an epic Federal Court challenge under the Trade Practices Act 1974 before merging into the National Rugby League. The New York Times has a breakdown of what happened in Europe, commentators have debated whether the Super League would breach EU competition law, a Madrid court decided that it would, and the European Commission decided not to get involved.
   
• G+T has written the Australian chapter of the 10th edition Energy Regulation and Markets Review, which you can now read on the website. G+T also publishes an Australian Energy Regulatory Market update each month – but only to subscribers!  You can become one on our subscription page.
   
• Finally, the Federal Court has dismissed the ACCC’s case against NSW Ports, the private operator of Port Botany and Port Kembla. The ACCC had argued that the privatisation arrangements had an anti-competitive purpose and effect, since they effectively required the nearby Port of Newcastle to compensate NSW Ports if it ever expanded its container operations.  However, Justice Jagot found that NSW Ports had derivative crown immunity in its dealings with the NSW government, which had primary crown immunity because the State was not carrying on a business in privatising these assets.  Even without this immunity, Her Honour considered there was only a speculative hope – and not a real chance or possibility – that Newcastle would have entered the market to supply Container Port Services in NSW anyway, so the arrangements had no purpose or likely effect of lessening competition.  G+T acted for NSW Ports.

08/06/2021

The 2019 case of ASIC v Kobelt divided the High Court 4–3 and commentators in similar proportions, and prompted calls at the highest levels for reform of Australia’s treatment of unfair or unconscionable conduct.

Kobelt centred on the provision of “book up” credit to Aboriginal communities in remote and rural Australia. Mr Kobelt ran a general store in Mintabie in the far north of South Australia, and also sold used cars on credit. His book-up system required customers to give him their debit cards and PIN numbers; he would then withdraw their wages or payments as they were deposited, pay down their debts and provide them with goods from his store as needed across the payment cycle.

Record-keeping was certainly on the informal side, the effective cost of credit was high and the cars were generally clapped-out. Leaving your card at Mr Kobelt’s store made it harder to shop anywhere else. Many of Mr Kobelt’s customers had low financial literacy and were otherwise vulnerable. However, Mr Kobelt and his lawyers argued that his customers were largely happy with the system, they’d entered into it voluntarily, and they’d have trouble buying even clapped-out cars without it. Anthropological evidence suggested that the arrangement could also help customers to resist social pressure to share around payments as soon as they received them.

At first instance, Justice White found that Mr Kobelt’s system involved a form of paternalism which, while benevolent in some respects, kept customers in a situation of dependence and vulnerability. His Honour was mindful that the Court should not itself impose a paternalistic view of what was appropriate for Mr Kobelt’s customers, but he found that Mr Kobelt had exploited their clear vulnerability for his own benefit and to their detriment, which was unconscionable.

On appeal, Justices Besanko, Gilmour and Wigney all found that Mr Kobelt’s system was not unconscionable in the circumstances. Book-up had been widely used in rural and regional Australia, and Mr Kobelt’s customers understood at least the basic elements of the arrangements and were willing participants. Justice Wigney emphasised the need to respect their agency, and to avoid paternalistically imposing norms and values that did not sit with their culture and practices.

The High Court decision consists of five judgments. They are all strongly and convincingly argued. They all recognise the advantages and disadvantages of Mr Kobelt’s system and the need to balance his customers’ agency against their vulnerability, but they come to very different conclusions.

The joint judgment by Chief Justice Kiefel and Justice Bell, and separate judgments by Justice Gageler and Justice Keane, all agreed with the Full Court and found that Mr Kobelt’s conduct was not unconscionable. Justice Gageler said that he’d had some concerns in allowing special leave to appeal to the High Court, which have proved prescient:

Hard cases test and sometimes strain legal principle. They do not always lend themselves to elucidation of legal principle in a way that can be predicted to provide precedential guidance of the systemic usefulness generally to be expected from a decision of an ultimate court of appeal. Special leave to appeal having been granted, it is unsatisfactory but unsurprising to me that the Court should find itself closely divided on the resolution of the appeal.

A joint dissenting judgment by Justices Nettle and Gordon and a separate dissent by Justice Edelman agreed with the trial judge that Mr Kobelt’s conduct was unconscionable. The joint judgment took particular issue with Justice Wigney’s view that judicial intervention would be paternalistic:

It is not paternalistic to assess the vulnerability of Mr Kobelt's customers and whether that vulnerability was exploited. It is not paternalistic to take into account that the view of a vulnerable party of a transaction will be shaped by context and circumstance. Equally, it is not paternalistic to look at the transaction and the position of the parties objectively. It is to do no more than engage with the criteria of unconscionability.

Justice Edelman considered that the majority had applied the statutory concept of unconscionable conduct too narrowly – and had not freed themselves from the orbit of the modern equitable standard of unconscionability, despite the explicit intentions of the statute. His Honour thought that the very language of unconscionability could defeat any effort to broaden the concept:

If so, any lowering of the bar towards the nineteenth century equitable meaning synonymous with “unfairness” or “injustice” may only be possible if “unconscionable” is replaced with “unjust” or “unfair”.

In 2020 the entire township of Mintabie was closed and returned to the Anangu Pitjantjatjara Yankunytjatjara lands after a report found that it was an access point for drugs and alcohol into the region – as well as concerning credit arrangements and the sale of potentially unroadworthy second-hand vehicles. This must be one of the most comprehensive remedies since Tanis, and reminds us that the courts aren’t the only way to deal with these issues.

Is fair fair?

Following the Kobelt decision, Justice Maxwell of the Victorian Court of Appeal closed an oration to the Victorian Law Foundation by taking up Justice Edelman’s suggestion:

Are the courts still being too restrictive? Or is the problem with the standard? Adoption of fairness as a test might not be conducive to greater certainty. But it would certainly promote better understanding by all concerned – and, it might be hoped, higher standards of conduct – if we had a prohibition on conduct which was “in all the circumstances, unfair”.

The ACCC had already been considering a prohibition against unfair trading or unfair business practices through its participation in the Australian Consumer Law Review, and Chair Rod Sims swiftly endorsed Justice Maxwell’s conclusion, acknowledging the influence of the Kobelt decision and commentary on the ACCC:

Sims says Maxwell’s speech and Edelman’s dissent “have helped clarify my personal thinking and I think the thinking of other people in the ACCC.”

“I don’t think this is a matter of criticising judges at all, I think it’s a matter of now we’ve clearly got an issue,” he says. “It’s up to the legislature to be clear about what they mean, and if what they mean is unfairness, which I think they do, then we should change the test to that.”

The ACCC has gone on to recommend the introduction of an unfair practices prohibition in its Digital Platforms Inquiry, its Customer Loyalty Schemes Review, its Perishable Agricultural Goods Inquiry, both Interim Reports (so far) of its Digital Platform Services Inquiry and a number of speeches.

While the ACCC has not yet formally advocated for any particular wording, Chair Rod Sims has described some of the parameters of the proposed prohibition:

A general prohibition against unfair conduct would, on the face of it, be sweeping. Sims acknowledges that it’s a lower hurdle than unconscionable conduct, but he emphasises that conduct would only fall under the new law if it was both outside the business’s normal practices and caused substantial detriment to the consumer or small business.

A further potential limitation may appear in Mr Sims’s appeal to business to contribute to the formulation of the new prohibition:

“Help us work up how it should actually work. But accept the notion that you shouldn't be treating consumers unfairly when that results in significant detriment to them, and when you're not doing something reasonably necessary to run your business.”

Solace of Quantum

At first instance, the recent case of ACCC v Quantum Housing Group appeared to confirm both Justice Gageler’s fear that Kobelt might not provide precedential guidance of systemic usefulness and Rod Sims’s view that the courts were still setting too high a bar for unconscionable conduct.

In that case, the parties agreed that Quantum had engaged in both unconscionable and misleading conduct when it pressured its property investors to acquire services from its related property managers to maintain their status within the National Rental Affordability Scheme. They made joint submissions on remedies including a penalty of $700,000.

Justice Colvin accepted the proposed penalties and found that Quantum had engaged in misleading conduct – but could not find that it had engaged in unconscionable conduct, based on his reading of the High Court’s decision in Kobelt. His Honour traversed each of the five judgments and concluded that:

The majority view supports the adoption of a standard that requires exploitation of disadvantage by a party in a stronger position by conduct that is well outside the bounds of what is generally seen to be moral, right or acceptable commercial behaviour.

On the facts, his Honour decided these were sophisticated property investors who could look after their own interests, understood their dealings with Quantum and had suffered no financial loss. They were not vulnerable or in a position of disadvantage that would be expose them to being exploited or victimised, and Quantum’s conduct did not fall within the statutory notion of unconscionable conduct.

The ACCC appealed the decision on the issue of unconscionable conduct. Quantum did not participate in the appeal, so the Court requested the appointment of counsel to act as contradictor to the ACCC’s case. And so it was that two of our most distinguished alumnae faced off in a battle of pure principle.

In March 2021, Chief Justice Allsop and Justices Besanko and McKerracher allowed the appeal in a joint judgment. They found that statutory unconscionable conduct did not require the exploitation of any vulnerability, disability or disadvantage. On the correct view, Quantum’s conduct “sufficiently departed from acceptable business standards as to be against or offend conscience” even if it had not taken advantage of any vulnerability of the investors:

Surely to predate on vulnerable consumers or small business people is unconscionable. But why is it not also unconscionable to act in a way that is systematically dishonest, entirely in bad faith in undermining a bargain, involving misrepresentation, commercial bullying or pressure and sharp practice, using a superior bargaining position, behaving contrary to an industry code, using significant market power in a way to extract an undisclosed benefit that will harm others who are commercially related to the counterparty?

ACCC Chair Rod Sims has welcomed the Full Court’s decision as a “vital clarification” of the law:

This is huge. This has very far-reaching implications for big business in terms of what they can do, small business in terms of the protections they’re entitled to, and consumers in terms of the protections they’re entitled to.

Telstra’s $50 million penalty

In May 2021, the Federal Court accepted Telstra’s admission that it had engaged in statutory unconscionable conduct when sales staff at its stores improperly sold mobile contracts and devices to vulnerable Aboriginal and Torres Strait Islander customers. It imposed a penalty of $50 million, as agreed by Telstra and the ACCC – the second-largest penalty under the Australian Consumer Law and the largest for unconscionable conduct.

Justice Mortimer accepted that it was appropriate to characterise the conduct in question as unconscionable, adopting Chief Justice Allsop’s explanation in ANZ v Paciocco that the prohibition of unconscionable conduct involves:

recognition of the deep and abiding requirement of honesty in behaviour; a rejection of trickery or sharp practice; fairness when dealing with consumers; the central importance of the faithful performance of bargains and promises freely made; the protection of those whose vulnerability as to the protection of their own interests places them in a position that calls for a just legal system to respond for their protection, especially from those who would victimise, predate or take advantage; a recognition that inequality of bargaining power can (but not always) be used in a way that is contrary to fair dealing or conscience; the importance of a reasonable degree of certainty in commercial transactions; the reversibility of enrichments unjustly received; the importance of behaviour in a business and consumer context that exhibits good faith and fair dealing; and the conduct of an equitable and certain judicial system that is not a harbour for idiosyncratic or personal moral judgment and exercise of power and discretion based thereon.

Her Honour didn’t refer to either Quantum or Kobelt, but her adoption of the Chief Justice’s extended description in Paciocco suggests an implicit rejection of the narrow view of statutory unconscionable conduct and an embrace of broader ideas of fairness and fair dealing, good faith and honesty, and the condemnation of trickery, sharp practice and injustice – the kind of ideas that proposed reforms are intended to promote.

Where to now?

It’s too early to say how these recent cases might affect calls for a new prohibition of unfair or unjust conduct – or for any change to the statutory unconscionable conduct provisions. The High Court decision in Kobelt remains the current authority on the standard for statutory unconscionability and may remain divisive until another case reaches the High Court.

However, the Full Federal Court in Quantum has clarified that the majority in Kobelt did not require the exploitation of a special category of disadvantage or indeed any disadvantage, clearly distinguishing the test of statutory unconscionability from the narrow equitable standard. Lower courts are likely to apply the broader standard at least for the foreseeable future.

At the same time, the principle that the statutory standard should be broader than the equitable standard is clear from the legislation and has been repeatedly affirmed by the courts – while in practice the courts have continued to apply the standard in a way that many commentators consider too narrow.

It may be that, as Justice Maxwell suggests, new language may be required to significantly expand the concept. But even with new language, the considerations outlined by Chief Justice Allsop may still be the ones that determine whether conduct is unjust or unfair, and in what circumstances the court should intervene.

We’ll keep an eye on any future unconscionability cases and any proposals for new legislation that may emerge this year.

In other news …

• A reduced 11-member jury made short work of Australia’s first contested criminal cartel prosecution, finding mobility-aid provider Country Care and its managing director not guilty of all charges. After almost three months of hearings and two days of jury directions, the jury only took four hours to reach its verdict. Because juries don’t provide reasons, we can’t know whether this one was swayed by arguments that the alleged cartel parties were really in a joint venture, or had its doubts about the immunity witnesses; but it’s clear that juries are adding a whole new dimension to the already complex cartel framework.
• The trial in the US case of Epic v Apple has concluded after three weeks of testimony over Apple’s exclusive app store and in-app purchase system. In parallel Australian proceedings, the Federal Court granted Apple a temporary stay with a view to relocating the case to the US; Epic has appealed the stay and the ACCC has been granted leave to intervene as a non-party in the appeal, where it will argue that disputes involving Australia’s competition laws should be heard and determined by Australian courts. In the UK, the Competition Appeal Tribunal has found that the US is the appropriate forum for the dispute. Epic has also taken similar action against Google in the usual jurisdictions.
• My colleague Susan Jones has written the chapter on pharmaceuticals and medical products for the GCR Asia-Pacific Antitrust Review 2021 and it’s now available on our website. It’s a great way to catch up on recent developments in competition in this critical area.
• Victoria has passed its Zero and Low Emission Vehicle Distance-based Charge Bill 2021 which will impose a charge of up to 2.5 cents per kilometre for electric (or hydrogen) vehicles; other states have announced or are considering similar charges. In 2015 the Harper Review recommended direct charges on road use to replace indirect charges such as registration fees and the fuel excise, and received many angry submissions about the Constitution.
• Social media doesn’t seem to be particularly harmful to the mental health of teenagers, according to a recent study eloquently summarised in G+T’s “One Thing” newsletter. (The jury’s out on the rest of us.)

07/06/2021

Recently the ACCC released a joint statement with the UK Competition and Markets Authority and Germany’s Bundeskartellamt reaffirming the importance of effective merger control.  The statement warns agencies, courts and tribunals to remain vigilant against anti-competitive mergers – even where faced with uncertainty about the future, strong advocacy from the merging firms and their advisers, and a fractured or hesitant response from suppliers, customers or competitors.

At the launch of the statement, ACCC Chair Rod Sims suggested that these challenges may be heightened in Australia, given the role and the recent approach of the courts in our system.  He foreshadowed that the ACCC would suggest some changes to the merger review framework around the middle of 2021 – an important step in a process that has been underway for some years now.

The ACCC has a decent track record in court when it comes to consumer law, cartels and other collusive or unilateral conduct – it wins some and loses some, as you’d expect from an active enforcement agency whose priorities include taking test cases to clarify the boundaries of the law.  The courts have recognised this role in considering costs orders, most recently in Colgate-Palmolive:

The public responsibilities of regulators mean that they must, on occasion, pursue hard cases; cases based on circumstantial or highly contentious evidence or perhaps unsettled areas of law. It follows that it might reasonably be expected that they will lose some cases, sometimes emphatically. A regulator who chooses only to pursue easy cases and easy targets might perhaps have an excellent strike rate in terms of winning cases, but it would not be doing its job.

When it comes to merger cases, the ACCC’s informal clearance process gives it real influence over the deals that go to completion and the conditions that are attached – plenty of mergers are abandoned after a red light in a statement of issues, or cleared on the basis of undertakings.  But in recent years, the few merger cases to be resolved by the courts have gone against the ACCC.

In 2020 the courts cleared Vodafone’s acquisition of TPG and Pacific National’s asset purchase from Aurizon, both of which the ACCC had opposed.  Reflecting on these cases, Rod Sims located the key difficulty facing the ACCC in the need to prove future events to particular standards of certainty that are not always easy to pin down:

The heart of the problem is that, in merger cases, the ACCC is required to prove on the balance of probabilities what is likely to happen in the future if the merger does not proceed. That is, in essence, the merger regime seeks to compare what will happen with and without the merger and determine if the difference can represent a substantial lessening of competition.  [Emphasis to indicate various standards of certainty added.]

That summary neatly illustrates the complexities that have built up around the merger test, raising thorny issues of proof and prediction, probability and possibility that begin – but certainly don’t end – with the interpretation of the word “likely” in the statutory merger test.  Digging into these issues can help us understand the changes the ACCC may propose and the impact they might have.

A likely story

The word “likely” apparently comes from early Scandinavian forms such as the Old Icelandic líkligr: probable.  It appears about 350 times in the Competition and Consumer Act 2010, usually in some conjugation of “X, or is likely to X”.

That construction first appeared in the original section 45 of the Trade Practices Act 1974 – “the restraint has or is likely to have a significant effect on competition” – and has since spread throughout the legislation, including the section 50 merger test, which at first only had “the acquisition is likely to have the effect” but now has “the acquisition would have the effect, or be likely to have the effect of substantially lessening competition”.

This dual or two-limbed structure has shaped the way the courts have interpreted “likely” and the degree of certainty or probability it is thought to require.  Since Justice Deane in Tillmanns Butcheries in 1979, many of the authorities have reasoned that:

• whether the acquisition would have an effect must be determined on the ordinary civil standard of a balance of probabilities, or whether the conduct was more likely than not to have that effect; so
• to avoid redundancy, the question of whether the acquisition would be likely to have an effect must require some lesser degree of certainty.

On that basis Justice Deane concluded that “likely” required only “a real chance or possibility”.  In the AGL case in 2003, Justice French supported Justice Deane’s conclusion for similar reasons:

In any event as a matter of construction if “likely” simply meant more probable than not, it would be difficult to distinguish the application of that limb of the formula from the application of the first limb which, having regard to the onus of proof applicable in proceedings under Pt IV, could be established on the balance of probabilities.

Not everyone is persuaded by this argument.  Former Justice Heerey wrote in 2011:

The short answer is that an awkwardness of drafting should not lead to a reading which commits the greater sin of ignoring the ordinary meaning of words.

Things got more complicated with the Metcash case, where at first instance Justice Emmett found that, although the ACCC only needed to show that there was a real chance of substantially lessening competition, it first needed to establish the future state of the market, both with and without the merger, on the balance of probabilities.

While that two-stage process has not been followed, Justice Emmett had crucially separated the elements of section 50 from the standard of proof required to establish them – issues that had arguably been conflated in some of the cases that reasoned that “likely” must mean less than “more probable than not”.  Justice Middleton seized on this distinction in Vodafone/TPG:

One must distinguish between the requirements of the substantive law (s 50) and the principles or rules of evidence. The content of s 50 is not addressing the evidentiary burden.

The joint judgment of Justices Middleton and O’Bryan in the Pacific National/Aurizon appeal also questioned the basis for the “real chance” test, but came to a pragmatic landing:

In our view the ordinary meaning of the word “likely” in everyday language is “probable” (in the sense of more probable than not) and that meaning is consistent with dictionary definitions … However, the word “likely” has been construed to mean a likelihood that is less than probable for 40 years … if the meaning of the word “likely” was being considered for the first time, we would have been inclined to adopt the meaning probable, but there is insufficient reason to change course at this point in time.

But their Honours resolved the issue by concluding that:

The matter that must be proved on the balance of probabilities is that the impugned acquisition would have, or be likely to have, the effect of substantially lessening competition in any market.

That is, the ACCC may not need to separately prove, on the balance of probabilities, what is likely to happen if the merger doesn’t proceed – that will be a critical element that will contribute to the single evaluative judgment to be made by the court, but it doesn’t have its own standard of proof.

The ACCC does have to prove, on the balance of probabilities, that the acquisition would have a real chance of substantially lessening competition.  That might sound like an enigma or kōan – if a tree is likely to fall in the forest – but it appears to be the test for now, despite any lingering judicial unease about its history.  And even that test has proved difficult for the ACCC in the recent cases.

What changes might the ACCC propose?

To address these difficulties, the ACCC has pointed to a number of features of merger assessment frameworks overseas that are likely to inform its proposals for reform in Australia.  These include a rebuttable presumption that would put the burden on the merger parties to show that certain deals would not substantially lessen competition; measures to address the acquisition of nascent or potential competitors; and a more mandatory and suspensory merger clearance framework.

A rebuttable presumption

The ACCC has floated some form of rebuttable presumption in merger assessments since at least 2016:

Do we need to consider something similar to the approach adopted by US courts where once markets are defined and the merger is likely to result in a significant increase in concentration, there exists a “rebuttable presumption” that the merger should not proceed absent evidence to the contrary?

It's not yet clear how a “significant increase in concentration” might be defined.  Under the US Horizontal Merger Guidelines the agencies will raise a rebuttable presumption where a merger would increase the Herfindahl–Hirschman Index (HHI) of market concentration by more than 200 points to 2500 or more – most mergers from 5 to 4 competitors would meet that threshold.  The ACCC’s 2008 Merger Guidelines also refer to the HHI, but the ACCC hasn’t mentioned it in a public competition assessment or press release since 2010 – perhaps because in Australia’s smaller markets many mergers would fall foul of these thresholds.

A rebuttable presumption may not have made any difference in cases like AGL and Vodafone/TPG, where the acquiring party applied to the court and took on the burden of showing that there was no real chance that the merger would substantially lessen competition.  It might tip the balance in cases where the ACCC initiates the action, as it did in Pacific National/Aurizon, where the ACCC had the burden of proving that the acquisition was likely to lessen competition – though even in that case the Full Court found that such a prospect was unlikely. 

In the United States, the Competition and Antitrust Law Enforcement Reform Act 2021 introduced by Senator Amy Klobuchar in February 2021 would prohibit acquisitions that “create an appreciable risk of materially lessening competition”.  An “appreciable risk” sounds like the same ballpark as a “real chance” or a “real commercial likelihood”, and it’s not clear that a “material” lessening of competition is less than a “substantial” one – though that was what Senator Nick Xenophon intended with the Trade Practices Amendment (Material Lessening of Competition-Richmond Amendment) Bill 2009.

The US bill would raise a rebuttable presumption of such a risk not only where there is a significant increase in market concentration – formalising the existing practice of the agencies – but also where one party has more than 50% of any market and at least a “reasonable probability” of competing with the other; or where the acquisition exceeds $5 billion, or $50 million if the acquirer is valued at over $100 billion.  These additional circumstances target strategic acquisitions of nascent or potential competitors, which have proved elusive to merger control.

Nascent or potential competitors

The ACCC has noted that other jurisdictions also use tools such as abuse of dominance or monopolisation provisions to address these nascent mergers.  For example, in December 2020 the US Fair Trade Commission took action against Facebook alleging monopolisation and unfair trading in relation to its acquisitions of WhatsApp and Instagram, which the FTC had previously cleared. 

In Australia, it might be possible for the ACCC to use section 46 in relation to an acquisition that has the purpose, if not the effect or likely effect, of substantially lessening competition.  The ACCC has always said that section 46 is concerned with – and limited to – exclusionary conduct, but it could argue that certain mergers are exclusionary.

In the United Kingdom, the Furman Report recommended a “balance of harms” approach that would take into account the scale as well as the likelihood of any harms and benefits resulting from a merger, in order to address acquisitions that might have a low chance of a very high impact on competition. 

Although the Competition and Markets Authority has expressed concerns about that approach, its Digital Markets Task Force has recently recommended a specific merger control regime for digital platforms with strategic market status.  The regime would include an obligation for those platforms to inform the CMA of all acquisitions, with mandatory notification and suspension of transactions that meet threshold tests of materiality and connection to the UK.  It would also shift the standard of proof from the current balance of probabilities to a “realistic prospect” that a transaction would substantially lessen competition.  It has not recommended a reversal of the burden of proof on the basis that it would be difficult for the merging parties to meet this burden in the vast majority of cases.

A mandatory and suspensory regime

The ACCC has pointed out the differences between Australia’s informal clearance regime and the processes that apply internationally:

When we look to our international counterparts, the Australian system is out of place next to the formal, mandatory and suspensory regimes in many jurisdictions … For example, there is no compulsory upfront information requirements, and transactions are not suspended pending ACCC clearance. This latter point can cause us large problems when companies can threaten to complete the acquisition at various points of our investigation, as sometimes happens.

The ACCC could modify its own policies and processes to encourage this behaviour, for example by regularly asking for an undertaking not to complete the transaction before it will consider granting informal clearance – as it does in relation to formal merger authorisations. 

But to be most effective, this is likely to require legislation.  That would be a significant change to Australia’s merger regime following the changes to the formal clearance and authorisation processes recommended by the Dawson and Harper reviews, and could mean the end of the informal clearance process as we know it.

It’s also not clear to what extent the change would increase the ACCC’s influence on merger outcomes in practice.  The number of completed mergers investigated by the ACCC has dropped dramatically over the last decade, now averaging between one and two each year since 2015 – suggesting that most of the mergers likely to interest the ACCC are now being notified ahead of time.

Even where the ACCC has investigated a completed merger, it hasn’t gone on to take action against such a merger in the modern era.  The ACCC now treats completed mergers as enforcement investigations rather than informal merger decisions, and stopped listing them on its public register in August 2020.  But the last completed deal that remains on the register – Google’s acquisition of Fitbit completed in January 2021 – may prove an exception as the ACCC considers its options.

It’s possible that the ACCC will recommend changes that go further than the options it has publicly raised, in light of recent and proposed changes overseas.  We’ll keep you up to date with any further developments in the ACCC’s thinking or the meaning of “likely”. 

In other news …

• The ACCC has released a report into the authorisations it has granted to collaborations designed to address the COVID-19 pandemic.  The ACCC received 33 applications for urgent interim authorisation between March and May last year, and granted most of them within 48 hours.  It went on to grant final authorisation for all but five applications, which had already been withdrawn.  It has recently issued conditional interim authorisations to replace a number of the COVID-19 cooperation authorisations that were about to expire, including those between major supermarkets, 7-Eleven and its franchisees, and private health insurers – the last of which then withdrew its application for renewal.  Two other COVID-19 authorisations have also expired without being renewed, and the rest will expire in the coming months unless they are renewed.
• The second interim report of the ACCC’s Digital Platform Services Inquiry 2020–2025, focusing on mobile app marketplaces, was just released.  G+T has a great update on the ACCC’s findings and proposed measures.  The ACCC’s third interim report will focus on market dynamics and consumer choice screens in search services and web browsers, and has released in Issues Paper; submissions were due by 15 April 2021.
• The Commonwealth Government ran out of time to pass the Treasury Laws Amendment (2021 Measures No.1) Bill 2021 to extend relief measures permitting virtual meetings and electronic execution of documents before those measures expired.  ASIC has adopted an interim “no action” position in relation to virtual meetings but electronic executions have returned to their somewhat uncertain pre-COVID state.
• Airlines still face turbulence/headwinds/etc in Australia as overseas, with Virgin Australia emerging from voluntary administration, REX expanding into major domestic routes and the ACCC tasked with monitoring the industry until 2023.  Our colleagues Louise Klamka and Johnathon Geagea have a detailed update.
• Criminal penalties apply to cartel conduct in New Zealand as of 8 April 2021, almost ten years after the first bill to criminalise cartel conduct was introduced.  The Commerce Commission has some arresting videos illustrating the kind of conduct that may now result in jail time. 
• First-person view (FPV) drones captured spectacular footage of a bowling alley in Minnesota and the Fagradalsfjall volcano in Iceland.  In Australia, drones are regulated by the Civil Aviation Safety Authority (CASA) and FPV drones are subject to additional regulation since they don’t provide the traditional line-of-sight to the drone that is usually required. 

Don’t forget, you can read our last edition of the News Media Bargaining Code.

31/03/2021

The News Media and Digital Platforms Mandatory Bargaining Code has finally passed into law, news has returned to our Facebook feeds and it looks like Google search will remain accessible to Australians (other search engines are available). The negotiations around the Code have tested the bargaining power of the digital platforms, the news media and even the Australian government, who can all now say that they’ve got what they wanted, even if what they wanted has changed in some subtle but important ways.

The Code had its genesis in the reforms to the media ownership rules, which cross-bencher Nick Xenophon tied to measures to support public interest journalism, including an ACCC inquiry into the impact of the digital platforms on the news media. Senator Xenophon had previously said:

I have a particular interest in changing our competition laws so that we have an ability for every media organisation to be able to take on those content aggregators and search engines that are cannibalising their content.

The ACCC’s Digital Platforms Inquiry covered many aspects of both digital platforms and the news media, but the two collided in the ACCC’s conclusion that the platforms had disproportionate bargaining power when it came to the links they provided to articles on news media websites – what the ACCC described as “news referral services”. This was a very ACCC way of looking at the issue, more nuanced than the idea that the platforms were cannibalising news content, but not necessarily how the digital platforms had viewed the services they were providing.

Everyone recognised that the news media benefited from referral links and had chosen to allow the digital platforms to post links and accompanying snippets. But the digital platforms also benefited from those referrals, which made their service more attractive to their users. The news media argued that the digital platforms were getting the better end of the deal, perhaps by a long shot – and certainly the platforms were more profitable, and their share of digital advertising kept increasing at the news media’s expense.

The ACCC didn’t weigh in on the relative value of these arrangements but considered that there was an imbalance in the bargaining power between the platforms because:

Google and Facebook each appear to be more important to the major news media businesses than any one news media business is to Google or Facebook.

To address this imbalance, the ACCC recommended that each of the digital platforms should develop its own voluntary code of conduct, which should include objective criteria for negotiations with media businesses and minimum commitments to:

Where the digital platform obtains value directly or indirectly from content produced by news media businesses, fairly negotiate with news media businesses as to how that revenue should be shared, or how the news media businesses should be compensated.

When progress on those voluntary codes slowed, the Government directed the ACCC to develop a mandatory code for the industry. Under the ACCC’s draft Code, the Treasurer would identify digital platforms that had significant bargaining power in relation to Australian news providers, and those platforms would participate in a process of final-offer arbitration to determine the amount they would pay news businesses for making their content available. Following submissions from the news media, the draft Code as introduced to Parliament confirmed that “making available” would include links to or extracts of news content, as well as substantive content itself.

A new “link tax”?

To the platforms, this sounded like a version of the “link tax” that was proposed for the EU Copyright Directive – though ultimately modified to exclude “acts of hyperlinking” and “individual words or very short extracts” – and older battles over hyperlinking that have flared since the early days of the web. Some of the architects of the internet such as Vint Cerf (who now works at Google) and Tim Berners-Lee have argued to the Senate Committee that any requirement to pay for links is incompatible with “the free and open internet” and risks “breaching a fundamental principle of the Internet.”

One concern is that, if large digital platforms are required to pay for linking to news content, eventually everyone may have to pay for linking to anything. Even the Digital Platform Inquiry resisted the suggestion of a licensing arrangement for digital platforms to pay news businesses for the use of headlines and snippets, including because:

It is unclear why digital platforms should compensate media businesses for use of content while not offering compensation to other content creators and websites.

Google’s solution has been to make deals with news publishers under which it pays for substantive news content and related services rather than links and headlines. For example, Google has recently announced agreements with News Corp, Nine, Seven West, The Guardian, the ABC and other Australian news sources who will participate in Google News Showcase, a feature of its mobile apps and websites. It stresses that:

Payments are made for the publishers’ curatorial expertise, for beyond-the-paywall access and to curate content for story panels. Publishers receive monthly fees, which are intended to provide payments over three years. News Showcase is not a pay-per-click model …

Facebook had also made similar deals with news publishers in the United Kingdom and the United States and announced negotiations in France and Germany for its Facebook News feature, similarly noting that it will “pay publishers for content that is not already on the platform” (likely to mean paywalled content) and making clear that it is not paying for clicks or links:

If publishers were paid every time they posted an article, or every time someone clicked on one, it would incentivise clickbait, sensationalism and volume over high quality journalism. It would also favour bigger publishers with the resources to pump out masses of stories.

That approach was partly crystallised in the first set of amendments made by the House of Representatives, which made it clear that the final offers to be arbitrated must take the form of a lump sum amount for two years – not a pay-per-click model – though the amount is still expressed to remunerate the publishers for links and snippets as well as reproducing the content itself.

Asked whether deals of the kind that Google and Facebook had made or contemplated would avoid the need for the Code to apply, Treasurer Josh Frydenberg said:

I don’t want to pre-empt decisions around the designation process. What I have said consistently, privately and publicly, is that if commercial deals are struck, that changes the equation. And I’m not talking just about commercial deals with Channel Seven or Channel Nine or News Limited; we’re looking for a broader range of deals, including with the regional players and the smaller players, and, as I understand it, that is what is happening.

Crisis and denouement

That wasn’t enough for Facebook, which had not yet finalised deals with Australian publishers and shut down access to local and international news for its Australian users. This was in line with the choice that ACCC Chair Rod Sims had set out in July 2020:

They could stop showing news media on their platforms completely – that is, no local or international media – but short of that, they are compelled.

Facebook’s algorithm was a little too enthusiastic in its attempt to identify every Australian and international news site – even briefly blocking access to Gilbert + Tobin’s popular careers page – but it also highlighted some of the bugs in the draft Code. For example, the takedown had to apply to all sources of news content, but “news content” wasn’t defined. The Senate has since clarified the definition; it has also refined the anti-differentiation provisions and added two months of mediation before the parties wind up in final-offer arbitration.

Most significantly, the final Code also now requires the Treasurer to give 30 days’ notice before designating a digital platform or service, and to consider whether the platform has significantly contributed to the sustainability of the Australian news industry by making deals with news businesses. That will give the digital platforms a clearer path to avoid designation, reduce the risk that they will need to pay for links or snippets, and avert the consequences that those payments might have for the digital platforms or the free and open Internet.

Facebook has said that looks forward to agreeing new deals with Australian publishers and has recently signed an agreement with News Corp in Australia and letters of intent with Seven West, Nine and others. Google is likely to continue making its deals.

A win-win-win?

Paying for curation, expertise or paywalled content, rather than paying for snippets or links, may not make much difference from an economic perspective, and it doesn’t directly address the imbalance in bargaining power in news referral services that concerned the ACCC. Rod Sims has previously questioned the sufficiency of the digital platforms’ recent arrangements and rejected Google’s suggestion that only its Showcase product might be designated under the Code:

If the arbitration only involved Showcase, the assessment under the code would fail to account for the value Google derives from users being able to access news media business content via its core Google Search platform, which is the source of the bargaining power imbalance this code was designed to address.

But the distinction is critical for the digital platforms, not least because it shifts the framing from an antagonistic footing to a cooperative one that may better reflect the complex relationship between the news media and the digital platforms – and may at least as directly address the original concerns that gave rise to the Digital Platforms Inquiry and eventually to the News Media Bargaining Code. And if the Code can correct any imbalance in bargaining power not through its terms or mechanism but through its very existence, it will still have done its job. As Rod Sims recently said:

The purpose of the code is to give them the potential for arbitration, which helps their bargaining position and therefore helps them reach fair commercial deals. The code is to make sure that news media businesses can work towards fair commercial deals with FB and Google, it’s all about supporting journalism – that’s the complete objective here.

We’ll keep you updated on the impact of the Code and whether the digital platforms can sign up enough news businesses – particularly smaller and regional news businesses to avoid designation of some or all of their services.

In other news …

• ACCC Chair Rod Sims unveiled the agency’s enforcement priorities for 2021 and reflected on the impact of the COVID-19 pandemic. G+T has a useful summary here.
• Senator Amy Klobuchar introduced a far-reaching antitrust bill which would increase enforcement resources, change the merger test to “an appreciable risk of materially lessening competition” and shifting the burden of proof in certain mergers.
• The 2021 Bannerman Competition Lecture was delivered by the Competition and Markets Authority’s Dr Andrea Coscelli. It’s an insightful perspective on the issues facing regulators and enforcement agencies around the world.
• My colleagues Luke Woodward and Andrew Low have a conversation starter on regulators’ aversion to behavioural undertakings in mergers where they are much more comfortable with similar commitments in other contexts.
• NASA’s Perseverance rover made a spectacular landing on Mars with a hidden message in its parachute.

Please subscribe, and let us know what you think.