Another great competition workshop

Over the weekend in Wellington the Competition Law and Policy Institute of New Zealand put on its 27th annual workshop, and as usual it was a fascinating get-together (I wrote up last year's one here).

The big keynote session was Dr Jorge Padilla, head of Compass Lexecon Europe, on "Reform of misuse of market power", exactly the right choice given our current reconsideration of s36 of the Commerce Act. And Jorge had the gumption to prepare the helicopter-view, framework-setting piece the rest of us should have put in to MBIE's s36 review. Sure, we have mouths to feed and payrolls to meet, and perforce went straight into the undergrowth details of counterfactuals and case law and purpose versus effects, and none of us is paid to take the big picture view of a good MBIE policy analyst or a conference headline act, but with hindsight I, for one, certainly wish I'd taken more time in my own submissions to set the broad context first.

Never mind: Jorge filled the gap admirably, and I'm sure his high-level analysis will find its way into the MBIE hopper. He cast it in the form of an 'error/cost' model - no, come back! I'm going to explain! - and it made a lot of sense.

First, as background, on Twitter a while back I found a wonderful way to remember 'Type 1' and 'Type 2' errors: remember the story of the boy who cried 'Wolf!'. First time round, the villagers thought there was a wolf, but there wasn't. Second time round, they thought there wasn't a wolf, but there was. And the trick is that they committed Type 1 and Type 2 errors in that order. Type 1 is seeing something that isn't really there, convicting an innocent person, wrongly seeing anti-competitive abuse. Type 2 is not picking up something that you should have, acquitting a guilty person, letting an abuse slide as acceptable commercial practice.

Jorge's big idea was that not only legislation like s36, but also the shape of the competition authority and its governance, should all be designed as the solution to the problem of minimising the expected cost of errors, taking into account both the likelihood of type 1 versus type 2 errors, and the costs of the two kinds of errors. This will tend to steer you to the right legal test along the spectrum of per se legality at one end (eg above-cost 'predation'), through rebuttable presumption of legality (where you're most worried about Type 1 mistakes), through rule of reason (the error risks are more balanced), through rebuttable presumption of illegality (where Type 2 is front of mind), to per se illegality (eg cartels). He put 'refusals to deal', for example, in the box where on balance you'd be more concerned about Type 1 errors (wrongly pinging a practice that actually makes pro-competitive sense), so you'd want to apply a regime of presumptively but not absolutely legal. Below-cost predatory pricing would on similar reasoning be presumptively illegal.

Someone asked him about purpose-based or effects-based approaches to s36: he went for effects based, so let's hope that view, too, finds its way into our policy redevelopment.

The other big set-piece was Prof Carol Beaton-Wells, Director of Studies, Competition and Consumer Law at the University of Melbourne, on "Cartel Leniency and Trade-offs". This was one of those presentations that changes your mind about something that had previously seemed beyond all reasonable doubt. In this case, the previously unquestioned idea was the effectiveness of the leniency programmes that competition authorities run to destabilise and uncover cartels, and which typically offer immunity from prosecution by the authorities to the first (and only the first) cartel member who comes in and rats out the others (and continues to cooperate, for example by providing incriminatory evidence).

For me, and for the other Commissioners who were there at the time it came in, the Commerce Commission's leniency programme was the bee's knees, and indeed the Commission's guidelines (here as a webpage, here as a pdf) say "International and New Zealand experience indicates that an effective leniency programme is the single most effective tool available to detect cartels".

So Caron's presentation was an absolute eye-opener. Among other things, she pointed to data showing that the number of cartels being rumbled hasn't increased, and may have gone down, after the introduction of leniency; that the cartels being discovered are overwhelmingly ones that had already fallen over ("digging up dead bodies", as she put it) rather than the ongoing ones you might well be more concerned about; and so on, all the way down to the practicalities of problems with getting juries to go along with the evidence given by the dobbers-in, which juries tend to mistrust as self-serving or underhand.

For now, I'm at the gobsmacked stage of reaction, and if you're the same, we'll all have to go and work our way through Caron's latest publications on the (in)effectiveness of leniency policies. While you're at it, have a read of Robert C Marshall's and Leslie M Marx's The Economics of Collusion: Cartels and Bidding Rings, which I've recommended before and which is pretty much indispensable as a starting point for recent thinking about cartels.

It also, I think, helps explain what may have happened post-leniency: as Marshall and Marx show, cartel members and cartel hunters are engaged in an ongoing "arms race". Leniency may have given the hunters an edge, but cartels may have responded to the riskier environment by tightening confidentiality and devising stronger bonds of mutual obligation. That may also have happened post-criminalisation in Australia, where there was a long gap between criminalisation and the emergence of the first criminal prosecution. One possibility, as I suggested in 'Cartels? What cartels?', is that today's cartels are emulating submarines: running quiet, running deep.

And those were just the highlights of the workshop. We also had sessions on the s36 review, vertical price fixing, judicial perspectives on competition law cases, the collaborative activities exemption in the proposed revamp of s30, a state-of-play report on the current telco sector review, and a primer on the (newish) unfair contract terms legislation. Great stuff, jam packed into a very valuable day and a half, and special thanks to Chapman Tripp for providing the Wellington facilities. Well done to John Land and all the CLPINZ team on pulling this year's one together, and clear the date in your diary for next year's.