The wheels are back on

"The goss", I said a while back, "is that the Commission stands a good chance of having Lodge overturned in the Court of Appeal, but stranger things have happened", Lodge being the High Court case involving price-fixing amongst Hamilton real estate agents.

The Commission had rather unexpectedly lost it - "...and then the wheels came off" - but now the Court of Appeal has put the wheels back on, tightened the nuts, and banged the hubcaps into place.

In the High Court, Justice Jagose had said that the estate agents had come to a collective agreement (their agencies would stop absorbing the cost of Trade Me real estate listings and instead start charging the house sellers for it), and had given effect to it (part of it, for example, was a collective withdrawal of listings on Trade Me, which duly occurred).  But - and this was the surprise bit - the arrangement didn't breach s30 of the Commerce Act. If the seller wanted a Trade Me listing, it was still open to individual real estate agents to offer a discount or indeed carry the cost themselves if they badly wanted the listing. If the ad cost was still negotiable, then it couldn't be said that the collective agreement had controlled the price.

Not so, said the Court of Appeal. The estate agents had collectively agreed to remove State of the World A (the agencies wore the cost) and replace it with State of the World B (the vendor probably paid, but maybe twisted the arm of the agent and got it for free). That's not on: at [89]

We agree with Mr Dixon’s [John Dixon QC, for the Commission] submission that all of those vendors after January 2014 who chose not to list on Trade Me when faced with having to pay for it, and indeed those who did pay the fee, lost the opportunity to be offered a price which had been set for an agency operating in response to working competitive market forces ... Plainly the agreement in principle to withdraw from agency-paid Trade Me advertising would affect price; if the vendor did not have a Trade Me advertisement it had lost an allowance or credit that had been previously provided. The price for the real estate agencies’ services was correspondingly more.

And the Court gave a homely but useful example: at [88]

By way of example, if the retailers of motor vehicles in a street all agreed on an asking price for a certain model, aware that this was the asking price only and the end price after negotiation could be quite different, that would have an anti-competitive effect in the way discussed in Plymouth Dealers’ Association [an American case from 1960]. The starting point for one side of a negotiation about price would undoubtedly affect the end price, even if it may be possible or even likely the end price would be different from the starting point.

I could see Justice Jagose's point, but the Court of Appeal has put us both right. It's clearly the better take, and I'd be mightily surprised if the Hamilton agents press on to the Supreme Court.

Getting back onto the safer ground of economics, there was a side issue of whether economic evidence about what the real estate agents were likely to have done in any event was admissible. The High Court had said no; the Court of Appeal agreed.

And there's a lesson there for businesses in the real world. It's a recurring theme: some sort of shock happens - in this case, Trade Me tried on an enormous rise in its listing fees, in the air cargo cases airlines got whacked with new security charges - and the question arises, what to do? Chances are, in a competitive market, you're going to have to pass them on to the ultimate customer. All of you in the real estate or air freight game may have independently come to the same conclusion.

But you mustn't get together and collectively agree to do it, even if the two outcomes don't look very different. "It would have happened anyway" will be no defence. At [112] the Court of Appeal agreed with the High Court which had said at [238] that

in principle, s 30 type conduct does not avail of a competition analysis. Constraints on price-setting are deemed in breach of s 27. That the same price may have arisen in the counterfactual (ie, absent constraint) does not respond to the presence of constraint in the factual. The proposed [economics] evidence was irrelevant

So there have been twists and turns in the story, but the end point for businesses is still as I put it after the High Court: "it remains highly dangerous to go near any discussion of price or pricing models with your competitors ... How much of a cost to absorb, and how much to pass on, needs to be your own independent decision".

Double or quits?

Last month Australia's High Court told Japanese company Yazaki that no, it couldn't appeal against the record A$46 million cartel fine it had copped in Australia's Full Federal Court in May.

Now that the legal race is run, we can sit back and learn some lessons from the whole affair. Mostly, unfortunately, they're lessons about what not to do, though on the positive side they are guides to avoiding future heffalump traps.

First up in the catalogue of mistakes was Yazaki's disastrous decision to keep on fighting both conviction and penalty through the courts. Yazaki was bang to rights from the off: Yazaki and a bunch of other companies had already been convicted and fined in the US and Europe for price-fixing, bid rigging and market allocation in the market for "wire harnesses", which are things that link up the electrical components in a car. Yazaki had, for example, worn most (€125 million) of the €142 million penalty the European Commission had imposed in July 2013 for a cartel that had been ripping off Honda, Nissan, Toyota and Renault.

Sure, Yazaki had every right to claim it hadn't been ripping off Toyota in Australia (though it had). And to make the ACCC prove its case (it did). And to dispute the penalty imposed (originally A$9.5 million). But the proverbial snowball in hell had a better chance of beating the rap. Apart from the evidence in other jurisdictions of Yazaki being up to its eyeballs in wire harness rorts, Sumitomo Electric, Yazaki's comrade in cartel crime, had ratted Yazaki out both overseas and in Australia under the cartel leniency regime. Yazaki was always going to get its collar felt.

The sensible thing to have done was fold your hand as gracefully as you could. And you'd like to think Yazaki, or any New Zealand company in the same pickle, would send out for some independent economic and legal advice on the realistic chances of beating the rap and on the likely penalty (very low, and rather high, would have been the answers). But no: Yazaki appealed the original conviction in the Federal Court and the A$9.5 million penalty; the ACCC cross-appealed; and Yazaki scored one of the bigger corporate own goals of recent times, and got its fine quintupled to A$46 million.

Good. It's about time something disrupted the cynical calculus of appeals. Recall that in both Australia and New Zealand one way of setting cartel penalties is three times the profit gained from the cartel, the idea being that it acts as a disincentive to absorb cartel fines as a cost of doing business (provided the perceived chance of being detected and convicted is higher than one in three). But there's nothing equivalent deterring companies from clogging up the courts with doomed nuisance appeals. A $10 million fine? A 30% chance of getting it reduced to $5 million? A legal bill of $1 million? See you in the Court of Appeal, is the line of reasoning. This time it's fallen flat on its face, and I'm not sorry.

And it's not as if the courts are in any good place to absorb the dead weight of these nuisances. The ACCC first laid charges in December 2012. The Federal Court's liability judgement came out in November 2015 The original penalty judgement appeared in May 2017. The Full Federal Court  ramped up the fine in May 2018. The High Court finally drew two lines under the affair in October. That's nearly six years from go to whoa. I'd guess a similar timetable would have played out here in New Zealand. It's too slow. If 'hard core' cartelists like Yazaki think twice in the future before clogging up already slow courts, it'll be a good outcome.

Another lesson is that if you are in a cartel, or discover you've been in one, do not walk, do not amble, do not pause to think you're well hidden in the undergrowth and they'll never find you, RUN for the protection of the cartel leniency programme. So far Yazaki has copped well in excess of $1 billion globally, while Sumitomo Electric is home free, at least with the criminal enforcers (private class actions can't be fended off in the same way).

That was always good advice, but even more so now given the reasoning behind the Full Federal Court's whacking up the fine, and which I'd expect the New Zealand courts will be looking at. For one thing, it took the view that there were more contraventions than the Federal Court had identified (five, rather than the primary court's two, namely an overarching scheme plus a giving effect). For another, it said Yazaki's whole corporate income in Australia would go into the calculation of the maximum penalty (it ended up being a penalty based on turnover rather than on three times profits).

Yazaki wriggled: its argument was that the rest of its Australian business (which didn't even know about the cartel fix being in) shouldn't be swept into the maximum penalty calculation for what happened in one little corner. The Full Federal Court wasn't having a bar of it: at [198]

We also do not accept the respondents’ [Yazaki's] submission that the construction for which they contend is to be preferred by considering the position of a tiny subsidiary of a major Australian corporation, carrying on some very incidental and discrete part of the corporate group’s business, and engaging in cartel conduct, in circumstances where no other company in the corporate group had any knowledge of what was going on. If the consequence is that all of the values of the supplies made by every company in the corporate group were brought in, irrespective of their connection with the conduct of the contravening subsidiary corporation, we would not regard that as an absurd consequence, that is, a consequence obviously unintended by the Parliament.

One final learning is that, yet again, the primary victims of this cartel were other businesses: industrial inputs like wire harnesses are a common target for cartel operators. When it comes to debating cartel enforcement policy, it's fine for the business community to have reservations about criminalising cartels and sending ringleaders to prison, and to worry about pro-competitive collaborative activity being mistakenly characterised as anti-competitive cartels. But that's as far as concern or business solidarity should go: more businesses should realise they're the prime prey for these conspiracies, and should be supporting throwing the book at the Yazakis of this world.

Two sides to the story

You might think the ugly process of nominating US Supreme Court judges and the predictably  5 - 4 conservative/liberal votes in many of the Court's decisions wouldn't matter a hoot to us humble toilers in the vineyard of New Zealand competition and regulation. But you'd be wrong.

The US Supreme Court, in the Amex case, said something important in June about two-sided markets: formally it's Ohio. v American Express Co. and even has its own Wikipedia entry if you'd like a quick recap of what went on. Two-sided and multiple-sided markets and platforms are everywhere these days - I'm writing this post on one - and even if US jurisprudence doesn't always get a lot of traction in our Anglosphere courts, sooner or later the latest American anti-trust thinking tends to find its way through to us too, not least because the same economic experts front up here.

So Amex is relevant to us, and indeed we've traversed this territory ourselves in the past. I was involved at the time, so I'll just refer to the public material on the credit cards 'interchange case' of 2006-09, where an agreed settlement was reached pretty much on the steps of the High Court (the announcement of the Commerce Commission's proceedings in 2006 is here, one of the settlement announcements in 2009 is here)

This latest US case was about whether credit card companies can include 'anti steering' provisions in their contracts with retailers. 'Anti steering' means that a retailer, if it's signed up with Amex, say, can't nudge ('steer') the Amex-card-using shopper to some other means of payment. The retailer will be wont to steer the shopper to use the card system that costs the retailer least. That's usually not Amex: it charges relatively high fees to retailers to fund a relatively generous rewards programme for its cardholders.

Amex would argue that its card brought its (typically higher spending, upmarket) customer into the shop in the first place, and that at least one of the thoughts going through the buyer's mind when they spend up big in the store is the payoff from the Amex rewards programme. It's a swizz, on this reasoning, for the retailer to benefit from the Amex-initiated deal but put it through the till on someone else's card network.

Before reading Amex, I had sympathised with the plaintiffs (the Feds and 17 American states initially, but down to just 11 states at the Supreme Court). Language along the lines of "you mustn't mention there are competing alternatives to this card" doesn't sound good at all, and both Visa and MasterCard had agreed to stop doing it, with only Amex ploughing on all the way through the American courts. My first instincts would have been along the lines of the minority in the Supreme Court, which said

If American Express’ merchant fees are so high that merchants successfully induce their customers to use other cards, American Express can remedy that problem by lowering those fees or by spending more on cardholder rewards so that cardholders decline such requests. What it may not do is demand contractual protection from price competition (p26)

But to my considerable surprise (I didn't ever expect to find myself agreeing with the conservative majority of the current Supreme Court) the 5 - 4 decision in favour of Amex looked the right call.

The majority found that the credit card market is a two-sided market, which you would think is beyond much doubt. The District Court first hearing the Amex case had, however, oddly found separate single markets for retailer and shopper card services. The District Court got put right by the US Court of Appeal, and the Supreme Court affirmed it.

Following on from that market definition, the majority in Amex said that you mustn't draw anti-competitive conclusions from looking at high prices on one side of a two-sided market, a conclusion which is now a commonplace in competition economics but doesn't seem to have been considered by the Supreme Court before (the minority referred to a case from 1953, but that was long before the modern theory of two-sided markets):

Evidence of a price increase on one side of a two-sided transaction platform cannot by itself demonstrate an anticompetitive exercise of market power. To demonstrate anticompetitive effects on the two-sided credit-card market as a whole, the plaintiffs must prove that Amex’s antisteering provisions increased the cost of credit-card transactions above a competitive level, reduced the number of credit-card transactions, or otherwise stifled competition in the credit-card market ... They failed to do so (pp15-6)

The Amex majority pointed to a variety of evidence that showed no anti-competitive detriment: for example

the evidence that does exist cuts against the plaintiffs’ view that Amex’s antisteering provisions are the cause of any increases in merchant fees. Visa and MasterCard’s merchant fees have continued to increase, even at merchant locations where Amex is not accepted and, thus, Amex’s antisteering provisions do not apply ... This suggests that the cause of increased merchant fees is not Amex’s antisteering provisions, but rather increased competition for cardholders and a corresponding market wide adjustment in the relative price charged to merchants (pp16-7)

and they also cited a wide variety of other evidence (on pp18-9) showing ongoing vigorous competition between the card networks, which made the plaintiff's claim of harm rather difficult to sustain.

They also dealt to the "inherently anticompetitive" argument which I would have been attracted to - that it is inherently wrong to forbid retailers to mention the competition. Amex said that the anti-swizz justification ("you can't welcome our customer and then do a switcheroo against us") was okay, or in the majority's words

there is nothing inherently anticompetitive about Amex’s antisteering provisions. These agreements actually stem negative externalities in the credit-card market and promote interbrand competition ... This externality endangers the viability of the entire Amex network. And it undermines the investments that Amex has made to encourage increased cardholder spending, which discourages investments in rewards and ultimately harms both cardholders and merchants (p19)

If you're still sceptical about ultimately harming cardholders, it helps to think of Amex as an agent aggregating the collective purchasing power of its well-heeled membership to wrest what is effectively a larger discount from retailers.

The minority saw things completely differently, but struggled with their arguments. They persisted with the manifestly uphill notion that there are two separate markets:

the relationship between merchant-related card services and shopper-related card services is primarily that of complements, not substitutes (p11)

Since only substitutes are in the same market, there must must be two markets:

there is no justification for treating shopper-related services and merchant-related services as if they were part of a single market, at least not at step 1 of the "rule of reason" (p12)

The reference to "step 1" is to the 3-step process American jurisprudence follows (plaintiff says anti-competitive harm, defendant can rebut as pro-competitive, plaintiff can respond that it could be achieved less intrusively).

In any event, as a market definition, this looks a somewhat contrived description of what the majority better characterised as two sides of a single transaction. The minority also argued (which they needed to, since their first line of attack was weak), that market definition didn't matter, since evidence of higher prices was enough to establish anti-competitive effect:

The District Court’s findings of actual anticompetitive harm from the nondiscrimination provisions thus showed that, whatever the relevant market might be, American Express had enough power in that market to cause that harm. There is no reason to require a separate showing of market definition and market power under such circumstances. And so the majority’s extensive discussion of market definition is legally unnecessary (p14)

Not convincing at all: if a price increase is sufficient evidence of market power, every business in the country would be in the dock. But when a judgement opens with wrap-me-in-the-flag huffing ("For more than 120 years, the American economy has prospered by charting a middle path between pure laissez-faire and state capitalism") it's a good bet that wrap-me-in-the-logic is in short supply.

The judgment was met with outrage in some quarters. A piece from the Brookings Center on Regulation and Markets reacted with "Why the Supreme Court’s decision in Ohio v. AmEx will fatten the wealthy’s wallet (at the expense of the middle class)", for example, and the Open Markets Institute's Lina Khan wrote on Vox that "The Supreme Court just quietly gutted antitrust law".

I can understand the sentiment: 'anti steering' had initially looked wrong to me, too. But this is now the second time for me in recent months where something that initially looked deeply suspicious from a pro-competition point of view was judged okay (the other one was the Australian Pfizer case). For me, both cases correctly avoided a Type 1 error (wrongly finding anti-competitive detriment). But I'd also accept that in dynamic industries like pharma (Pfizer) and platforms (Amex) the safe path between Type 1 and Type 2 errors is getting harder and harder to spot.