Howzat!?

All over New Zealand there are High Court judges cowering in corners, whimpering "please don't let it be me, please don't let it be me..."

Because someone's going to get lumbered with the Sky TV / Vodafone appeal against the Commerce Commission*, while someone else is going to draw the short straw for the NZME / Fairfax appeal. You can imagine the goss over the G&Ts in the beaks' bar: "There'll be economists. And counterfactuals. And TSPs and two-sided platforms and a UFB. Why me?"

The Skyfone one shouldn't be too bad. It's essentially a "yes it is", "no it isn't" kind of argument: how could the Commission have drawn the conclusions it did from the evidence it considered? However it goes, I doubt that it's going set any big precedents for how the Commission interprets the Commerce Act.

There is one line of argument that might say something about how the Commission should go about assessing the "likelihood" of counterfactuals. For those fortunate enough not to follow these kinds of cases, the Commission assesses the effect of things like a Sky TV / Vodafone merger by comparing it with what would have happened in any event otherwise (the 'counterfactual'). Sometimes that's not clear: there might be several different but plausible things that might happen. And in that case, as para 2.32 of the Commission's  Merger and Acquisitions Guidelines says, my italics,

If competition would be substantially lessened in the scenario with the merger compared to any one of those likely states of competition without the merger, then the merger will have a likely effect of substantially lessening competition.

The Skyfone appeal argues (at 2(I)(iv) on pp7-8) that the Commission

Failed to stand back and consider whether each of the preconditions for the Commission’s theory of harm would all be likely to occur – which would have led the Commission to conclude that the cumulative probability of these preconditions occurring is remote.

That's a plausible runner about the likelihood of the counterfactual world, and we'll see how it is received. Overall though I'd say the appeal will face heavy going: as a general rule arguing that an expert tribunal has ballsed up its day job isn't the easiest thing to push uphill. But everyone's entitled to have a go, and in Sky TV's case, it's a completely rational thing to do. Sky TV lost some $220 million in market value after the merger decline was announced (share price from $4.35 to $3.78, times 389.1 million shares, equals $221.8 million). That's worth spending quite a bit, even on a long shot, to regain.

The NZME / Fairfax one will be trickier. It does have the same "yes it is", "no it isn't" components that Skyfone has, particularly around the issue of what markets are affected, but it could also lead to new case law on the Commerce Act, and it also questions the Commission's processes.

On the factual bunfight, the Commission said, for example in para 628 of its final decision, that there were three affected markets: online New Zealand news, Sunday newspapers, and community newspapers. The appeal says that each of these should have specifically been analysed as a two-sided market (readers on one side of each platform, advertisers on the other), and that on a proper view of the evidence there would be no loss of competition in any of them. The appeal also runs an alternative line that there might be a single two-sided market for everything: I wouldn't bet the house on that one, guys.

The appeal raises some wider issues. The big one is what should be counted as benefits or detriments, and how. The Commission had said that the big downside would be the loss of editorial diversity, a 'loss of plurality' from two big media groups merging. I had no trouble with that: as I argued here, a loss of product range options can clearly be seen as a consumer detriment. And even if it couldn't, I'm with the Commission when it said (at X39, my emphasis) that "it is clear from previous legal cases and common sense that we can and should take all the consequences of the merger - positive and negative - into account". But it didn't help (although what it said was in my view absolutely right) that the Commission looked as if it was philosophising on what is or is not an essential bit of the social infrastructure of a democracy.

That's been seized on in the appeal, where para 16 argues that the Commission must stick to the economics, and that (as case law puts it) it mustn't go in for "pure speculation" or "mere intuition". Para 17 goes on to say that even if the Commission is allowed to count things like plurality, it didn't do it right, and that other benefits (mostly cost savings) outweighed any plurality losses.

It doesn't throw in (as I would have) that the Commission didn't have any go at quantifying what the plurality loss might look like: the Commission said at para 1654 that "Such effects are unable to be quantified". And it would have been a heroic exercise, yes, we know that. But absent some numbers, an appeal could argue, "The Commission says the loss is significant. It says that implicitly it must be more than $200 million, the top end of the potential benefits. But how does it know? For all the Commission knows, the loss could be big, like the Commission says - $150 million maybe - but that would be less than the benefits. So how does it know the loss is more than $200 million?"

There are also some questions around the Commission's processes. Some of them look unfounded: "granting extensive anonymity and confidentiality to a large number of third parties submitting in opposition" doesn't fly for me, on either the facts or the logic. But there's possibly a more substantive point where the appeal says, you sprung a big surprise on us very late in the day about what counterfactual you had in mind. From the outside it's hard to tell exactly what's going on there, though there was a late flurry of correspondence between the Commission and the applicants well after the Commission's conference last December, which was somewhat unusual. Nobody - including the Commission - wants a process that doesn't tickle out all the evidence and give it a fair go, so the court running a health check on how it operates may be quite a good thing.

There's also an "appearance of predetermination" argument. As an article by Nick Grant in the NBR surmises ('NZME, Fairfax 'go full noise' with StuffMe appeal') it might be a complaint about an ill-judged 'StuffME' hashtag the Commission briefly used: not as professional as it should have been, but not a hanging matter, either. I doubt if it was enough to taint the whole proceedings.

We can't see everything - there's a wholly blacked-out bit in paragraph 19(c), for example, in the 'Natural justice and fairness' section of the NZME / Fairfax appeal - but on what we can see, what are the umpires likely to do with these appeals?

I'll be surprised if I see the Commission trudging back to the pavilion.

*Update July 1 - one of the judges is off the hook. The Skyfone appeal has been flagged away.