Epochal shifts in the economy, accelerating over the last five years of ‘polycrisis’, have also profoundly affected the structure of markets and the relationship between consumers, workers and producers, as well as the bargaining power and distribution of rents between them. There is a deep and urgent question about how antitrust enforcement can carry on ‘as usual’ in the face of these events. For instance, if we broadly agree (and many do) we have been too lenient for two decades and let too many meritless deals through, doesn’t this justify a more lenient error cost approach? Even if multiple deals were approved at the time because there was genuine uncertainty about their effects and concern about committing ‘Type 1’ errors (overenforcement), now that much of the uncertainty has resolved and we know many have turned out to be problematic (erecting moats and strengthening ecosystems), doesn’t this justify a tougher line? If a few more deals are opposed than ten years ago, so what? The narrative that startups need exit strategies and if they cannot sell to large incumbents they would lose all incentives to innovate is just a defence old chestnut for what is a much more variegated landscape (see Visa/Plaid and recently Adobe/Figma, both high-profile deals abandoned because of regulators’ opposition but where ‘walking away’ turned out well for both sides).
More generally, foregoing poorly documented ‘efficiencies’ claimed from a deal is not a disaster – when put to the test, these are often fragile. Leaving mergers aside: in conduct, consulting economists are trained to come up with ‘efficient/pro-competitive’ explanations for every conceivable practice – from tying and bundling to exclusivity payments, to loyalty discounts to discrimination. We know how to model these and show – surprise! – they increase consumer welfare. Economic analysis was useful to establish 20 years ago that a conduct may be benign in some situations, but in the case of entrenched quasi-monopolists in gatekeeping positions we should not credit esoteric efficiency claims and should not allow economists to dream them up.
Yet it is striking how much pushback there is from antitrust economists to the idea of adapting our enforcement practice to the current times. A first problem is the belief that there is a set of precepts – somehow all contained in ‘IO economics’ – which can give us a pure, objective, scientific and a-political assessment of market conduct. Aside from the enormous hubris and conceit this view entails, it is a stretch on several levels. The economics we practice today in antitrust is not pure, neutral and a-political. It is founded on neoliberal principles that became established as part of a deliberate conservative project started by President Reagan in the 1980s, involving William Baxter (first DOJ’s AAG and later FTC Commissioner) and indeed a contingent of Chicago academics (from Robert Bork to Richard Posner), then exported to Europe in the late 1990s under Commissioner Monti, after the infamous prohibition of GE/Honeywell. This orientation (pushed hard as the ‘more economic approach’ by consultants who saw also major business opportunities) became established in Europe in the early 2000s. And while things have evolved since (after Chicago we have seen post-Chicago, the empirical IO revolution, etc.) the precepts of ‘efficiency’ and ‘consumer welfare’ remain the guiding lights today for economists inside the agencies and consultants making submissions.
Is this bad? Well, it is an illusion to claim these are neutral principles, pure and a-political. In the main, they tend to support a pro-incumbent assessment. Take the ‘as efficient competitor test’, an established landmark test applied widely in conduct cases that basically says a new entrant which is not ‘as efficient’ as the incumbent should not be in the market – in essence, it deserves to die. Take the plethora of ‘efficiency justifications’ for conduct, from exclusive dealing to exclusivity discounts to tying and bundling to price undercutting: competition is ‘for the market’; the supplier needs exclusivity to make ‘relationship-specific investments’ into the customers which would not be justified unless he had all of the demand; profits in an ‘aftermarket’ are dissipated in the ’beforemarket’; there are economies of scope in supply or demand justifying bundle discounts; and so on. Some of this may well be true in a few cases, but there is also today a proliferation of these exculpatory narratives, as a set of adaptable narratives that economists pull out on request. Take the way we have badgered regulators that discrimination is output-enhancing in the first place, and therefore not anticompetitive. Take the way we badgered them about needing a presumption in favour of vertical and conglomerate deals – because of the strength of the ‘one monopoly profit’ logic (‘I don’t need to foreclose rivals downstream, I can take my profit elsewhere’) and of the ‘elimination of double marginalisation’ (mostly a chimera). Take the reluctance to engage with exploitation because it cannot be easily identified, which means every case (from Amazon Buy Box to Apple Spotify) pivots first into exclusion theories that are mostly inapt (the real issue in these cases being rent extraction, not exclusion). Much of the ‘analysis’ of conduct tends to reflect a DomCo defence playbook.
A further problem is that there is an obstinate siloing of competences, premised on a sense of superiority and detachment that ‘we know what we are doing over here, these other policy tools are nothing to do with us, we have a clear solitary superior mission’. This is another dimension of the conceit of the profession: IO is ‘the’ dominant economic discipline relevant to antitrust, and nothing else really gets to look in sideways. Business scholars are mostly seen as fluffy because they don’t ‘do maths’ or talk about ‘value creation’, which is not the point. Economic historians are irrelevant. Macroeconomics – who cares. Industrial policy – what does it have to do with us? Trade – not even close. Data protection? This is for the Data Protection Agencies over there.
What would be good to see?
My issue is not that enforcement in Europe is lax (it is however impossibly slow), but that the lens we collectively use in this space is too narrow given the state of the world we are in. The mantra has been historically that ‘as long as we keep markets competitive, we are doing our jobs’, to which we recently added ‘we don’t know what helping build resilience means’. We are in a deep hole, and we are witnessing major policy shifts and major rethinks in all areas of economic policy. What would a more open approach look like here? At a minimum, in my view it would require:
- A tough line on mergers. This is already happening (the European Commission and the UK have been more aggressive than in the past, alongside the US agencies). But the runway for landing cases should be a narrow one. We have already allowed multiple deals to roll markets and eliminate production capacity from Europe – from chemicals to pharma to agrochemicals to others. The excellent New Merger Guidelines just issued by the US FTC/DOJ (Federal Trade Commission 2023) contain a (rebuttable) presumption of illegality at given levels of concentration, and more generally have rewritten the rules of engagements for Parties in merger analysis to reflect much more closely how competition ‘presents itself’ in the current world: no clear distinction between horizontal/ vertical deals, more serial deals, more platform deals, more conglomerate and vertical deals, more partial ownership, more deals involving worse conditions for labour. This was not the case 15 years ago; it is the case now. These Guidelines make a real effort to reflect a new political economy; our equivalent in Europe are obsolete.
- A more explicit stance against concentrated markets (let alone monopolies). Directly recognising concentration is a problem, because market power engenders dysfunctions (and begets inequality). The traditional EU competition law claim that ‘being dominant isn’t an issue, abusing that dominance is’ is also obsolete. Concentration is an issue per se, because significant market power is an issue. Nothing should be allowed that increases or preserves that concentration, regardless of conduct. Structural separation should not be seen as an unthinkable last resort we could never really conceive of.
- Recognising and developing ‘ecosystems’ analysis. Much of the competitive interaction occurs between ecosystems, no longer in individual markets. We are missing a key dimension of competition for not being able to articulate how concerns might arise from the combination of assets and capabilities, short of a single ‘leveraging mechanism’ in a given market.
- Enforcing against exploitation, not just exclusion. The nature of concerns especially with large digital platforms is typically exploitation, and rent extraction. Yet a lot of cases (Apple Spotify, Amazon Buy Box, etc) were initially framed as exclusionary cases. What a waste of time. Of course this is because there is court precedent of exclusionary abuses (from Microsoft down) that we do not have for exploitative abuses. New Art. 102 Guidelines are expected to be released for consultation in the Summer 2024, but it is also expected they will not address exploitation. If so, this is problematic. The FTC in the US can pursue ‘unfair method of competition’ and although we are also beginning to see ‘fairness’ as a concern appearing in a few European Commission cases, it is not mainstream.
- Enforcing against discrimination. We have spent two decades telling regulators that discrimination is not a competition issue because it mostly expands output and is therefore pro-competitive. And that it could only possibly be a problem when it leads to rival foreclosure. This is problematic. We do worry about discrimination on multiple grounds short of foreclosure, including inequality. In the US, the FTC has revived discrimination as an area of focus for its enforcement/consumer protection role, and this is an important development.
- Taking up labour market issues. We are relatively oblivious to these in Europe, and we should not.
- Taking up data protection violations explicitly. One of the major learnings of the last five years surely has been that it is market power violations that enable data protection violations at scale, and vice versa it is data protection violations at scale that contribute to creating and preserving market power. Yet we continue only reluctantly to acknowledge the role data plays in market power and its abuse.
- Rethinking the approach to state aid to embrace modern industrial policy, away from the narrow obsession with market failures and efficiency as the only worthwhile pursuit and in the direction of a broader-lens, ‘European’ industrial policy which builds on a grown-up analysis of our shortcomings and needs. This would involve also being involved in actively promoting the creation of sovereign funds for investment in infrastructure.
Note this is not, as some have falsely claimed, an effort to ‘expel economics’ from enforcement. Far from it. It is a call to be less dogmatic that ‘we know best’ and ‘we have wisdom’, and look at what is before us. More generally, competition regulators should clamour to be part of the current Great Reordering, not be held back by the posse of lawyers, consultants, economists invested in the status quo – as if wisdom had been fully achieved sometime 20 years ago and now just needs to be applied, like Moses Tablets. The strong reaction in the US to the Draft New Merger Guidelines issued in July 2023 was mostly based around ‘we got it right in 2010, that reflected the right economic thinking, why change’. Because the new version better fits the world as we see it. The political economy, indeed.
Agencies typically say ‘our job is (just) to preserve/restore competition’. Good as that is, they can be a bigger and better part of the solution to the issues that face Europe. The Spanish Presidency Resilient EU 2030 ‘non paper’ calls for a mission ‘not just to prevent Europe’s decline’; it acknowledges ‘the changes experienced in the international order in recent years’, and lists the ‘multiple strategic vulnerabilities the EU should address in this decade’ (including reindustrialisation and rebuilding production capacities, as ‘the days of unchecked offshoring and blind reliance on imports are over’, enabling local technologies and hosting meaningful debates on emerging challenges) (Spain’s National Office of Foresight and Strategy 2023). Tellingly, the document hardly mentions competition policy, and only in terms of the role of state aid as preserving the ‘level playing field’ (p.31). To the extent this is because the role of competition regulators is perceived as technocratic and narrow, it is not a good thing. There is potentially much to be achieved with reorienting competition policy in ways that can be complementary to these broader, existential goals.
References
Federal Trade Commission (2023), Merger Guidelines 2023.
Spain’s National Office of Foresight and Strategy (2023), Resilient EU 2030.
