We are due to get a new set of guidelines on abuse of dominance, designed to take stock on a couple of decades of case practice and court precedents. It’s all part of a full repaint going on over at DG Comp, sprucing up all the various guidelines to suit the zeitgeist for a 2020s regulator.

The once-in-a-century financial crises continues to cast a long shadow. We know that new guidelines will mean more intervention, more flexibility and discretion than before. The guidelines will grapple with how to loosen the strict tests for an ‘as efficient competitor’ and for refusal to supply, and also make the most of opportunities for categories of abuse based around the concept of self-preferencing.

Notwithstanding the pleadings this week in the European Court of Justice, there is perhaps an inevitability to that. We continue to live through an era where it has become accepted through repetition (perhaps benefiting from the so-called illusory truth effect) that markets are working less well than before, and becoming less competitive. Still, why not take the opportunity anyway to step back and ask: what should abuse of dominance law actually be for?

Well, as I see it there are broadly two camps out there. In the blue corner we have camp Neutrality. This is broadly where I would put the regulators, including the European Commission. Broadly speaking, in Neutralists (or should that be Neutralitist? Not sure.) want to see dominant firms treat other firms in the same way as they treat themselves in all matters – with all business being conducted on an arms-length basis whether it is internal or external. In the red corner is camp Efficiency. In camp efficiency a dominant firm can do whatever the hell it likes unless its conduct is excluding an as-efficient competitor. No surprise to find lots of economists in the red corner, and lots of love for a strictly enforced AEC test. The Efficientists (Efficienarios??) were whooping when the European Commission was strung up for its failure to do a proper AEC test in the Intel case.

They are both wrong.

First, the idea of neutrality everywhere is problematic. It’s one of these things that only sounds like a good idea until you properly think about it – although it’s a dangerous one because on its face it sounds so fair and open and jolly nice. But it’s quite possibly shaping up to be one of the largest economic policy errors of our times. As I’ve said before, they didn’t give Oliver Hart a Nobel prize in economics for being good at crossword puzzles. He established the critical insight that complementary assets and capabilities are usually better under common ownership, and that firms often organise their structure around these clusters of complementary activities. Open arms-length relationships are beset by a host of contracting problems which cannot easily be solved. Put another way, self-preferencing is at the very core of what it means to be a firm in a market economy. Competition cannot meaningfully proceed if every firm treats every other firm as though they were all part of the same big happy family.

Second, the idea that we can benchmark abusive conduct using the as efficient competitor test is equally wrong. Take the example of exclusivity rebates that has spent a decade or more weaving its way through the European courts. The AEC test says I’m allowed to juice up my exclusivity rebates so high (though no higher) that an as-efficient competitor makes zero economic profit (after all, they are not foreclosed). And all whilst having structured my prices in a way that means I continue to extract very substantial excess rents as a dominant firm. What potential entrant in its right mind would take a chance on the market if I can make sure that their best-case-scenario is to earn zero? This is not rocket science people.

What should we really be trying to do?

Here is my answer. We should be trying to protect dynamic competition. This position sits a bit between the red and blue corners. Here in the purple corner, us Dynamists (Dynamites??) think the purpose of  abuse of dominance law is and should be to protect the dynamic process of competition over the medium and long term. It builds from that central article of faith that dominance itself is not a problem. Or as Peter Mandelson famously said to Silicon Valley executives back in the late 90s “we are intensely relaxed about people getting filthy rich”.

Here’s how dynamic competition works in a nutshell (Friedrich Hayek eat your heart out):

– nobody really knows what people want (especially not civil servants)

– different people try to figure out what that actually is

• if it’s cheap to spend time figuring that out, lots of people try
• if it’s expensive, then only a few people will try

– someone really nails the answer, and BANG you have the next iPhone or Google or whatever

– everyone loves it, and uses it – because everything else is shit

– the firm who figured it out makes loads of dosh and becomes dominant

– other people think ‘I too want loads of dosh’ so they spend time trying to figure out something new that people want even more

– and the cycle repeats

Whether or not you end up with competitive market structures (in the static sense)) at any given moment really depends. If two or more people have good ideas at the same time, then you’ll get that. If they don’t, then you will have periods – long or short, who knows – where market dominance exists. What really matters for dynamic competition is that people spend as much time as possible trying to figure out what people really want. As Hayek put it, if we already knew what people really wanted, then we wouldn’t need all the cost, waste and noise of competitive markets in the first place. It would be far better to just give it to them. Remember: real socialism has never been tried!

So the role of abuse of dominance law really comes in to protect that most crucial part of the process: “…and the cycle repeats”. That means keeping in place very strong incentives to keep people figuring out better ways of giving us what we really want. Crudely, those incentives are represented by that bit above that says “I too want loads of dosh”. It’s no good for competition if either dominant firms (or regulators, for that matter) structure the market to take that dosh away. The cycle will not repeat. And when the cycle of innovation and dynamic competition stops, you get poor very quickly whatever market structure you are living with.

The ideology of both the red and blue camps put the process of dynamic competition at risk in important ways by potentially weakening economic incentives. For example:

– An as-efficient-competitor test framework allows dominant firms to do all sorts of things to hide their profits from being challenged by new entrants. Those profits can be as high you like, but they all need to be on the table for competition. It’s no good if the best outcome an entrant can expect is to simply earn non-negative profits on the contestable part of demand. Meanwhile the dominant firm gets to rake in monopoly profits on the rest. If I manage to come up with a competitive product against a dominant firm, I should get a decent taste of whatever profits are left in the market. So it’s really a question of how profits are shared with the entrant, not about whether those profits are non-zero. The AEC test gets nowhere near testing that question in the right way.

– A similar problem arises with the modern regulatory obsession with neutrality everywhere. Neutral arms-length dealing, access and interoperability all hand the car keys to new entrants without them having to go through the crucial step of thinking “what would be an even better car”? Static market structure is improved at the expense of dynamic competition. Hand over the keys too easily and of course no one is going to go and get their own car instead. Even worse, dominant firms are also told not to bother to go and find new ways to give customers what they really want – because all that needs to be opened up (or even better left completely) to rivals. Much better then for everyone – competitors included – to just stick to what they have.

What abuse of dominance rules best protect dynamic competition?

My suggestion would be that the right underlying economic principle is to limit strategies that reduce the share of a dominant firm’s profits that can be captured by a successful entrant. This goes to heart of protecting the incentives that power dynamic competition. This principle captures a number of familiar classes of abusive behaviour. For example:

– margin squeeze: which aims to move profits from the contestable retail market up into the non-contestable wholesale market;

– predatory pricing: which aims to move profits over time from periods where entrants are active to period where entrants have exited;

– exclusivity/loyalty rebates: which aims to move profits from the contestable volume to the non-contestable volume.

I think there is a strong commonality between this underlying economic principle and the suggestion we have made that a good alternative framework is to look at situations where dominant firms focus competition the non-replicable dimensions of competition. By employing these strategies, firms can ensure rivals spend more or all of their time competing where it is hardest and where the available profits are lowest (all whilst preserving higher profits elsewhere). For example, the practice of tying focuses competition away from the tied product and on to the dominant product. Competition becomes about the non-replicable part of the bundle, meaning that rivals can’t succeed in the standalone market without giving up all their profit to compensate customers for the bundle discount. In a similar way to a loyalty rebate, profits are transferred from a contestable to non-contestable portion of demand, limiting the profit upside to a successful entrant.

More on this in our consultation response to the European Commission available here.