Cristina Caffarra, 4 November 2024

The “loud calls for a joined-up approach linking competition, trade and industrial policy” have finally been heard in Europe, albeit with diffidence.[1]  For the few of us making these calls for some time,[2] this was a fundamental necessity in the wake of a huge shift in the political economy: post-financial crisis, post-Covid, as the faults of neoliberal trickle-down efficiency-led vision became clear and the wisdom of an “all of government” approach became obvious.  But no one cared or listened, least of all DG Competition which remained wedded to the “one instrument, one goal” mantra, unresponsive and persistent in its elitist posture – at once unadventurous, intellectually unambitious, and unwilling to engage. Sycophancy from the Brussels “antitrust Bubble” (lawyers, economists, journalists, companies and their lobbyists who’d do anything to get their case through) is the culture broth which produces statements such as “competition is just a side dish relative to the big issues of our times”.[3]  Unambitious, and wrong. With this attitude, it is no surprise competition enforcement in Europe has been indeed on the way to becoming a “side dish” – increasingly irrelevant to the economy at large.

But these same calls can no longer be ignored when the big vector for change coming from the political economy gets supercharged by the Draghi report[4], with a devastating diagnosis of Europe’s ailments, and the prescription that “industrial, competition and trade policies interact closely, and must be aligned as part of an overall strategy” (p. 9). President Von Der Leyen designed her new Commission indeed by distributing competences across Commissioners in a way which would have been unthinkable five years ago, in an effort to implement Draghi’s blueprint. Europeans don’t like to call it “all of government approach”, too American, but that’s what it is.

What can the “calls” mean in practice? We started exploring the question “in the first grown-up discussion” (Jan De Loecker) at CEPR’s Competition Research Policy Network online on 28 October.[5] This note summarises my thoughts, as informed by that very rich discussion.

1. Four foundational principles

No one is arguing for competition policy being upended. As Jan De Loecker (Leuven)  put it at the event, the common overarching goal should remain “how do we facilitate a pro-competitive environment that can meet global challenges without harming our consumers”. This is a goal we can all recognise. We are not talking about directly pursuing (say) reduction of inequality or decarbonisation through competition enforcement.  However, operating a “joined up approach” as suggested by Draghi (and welcome by many) involves establishing some foundational principles. 

First, it is essential there is a common understanding of the macro landscape. We cannot begin to analyse markets, even locally, without being cognizant of the broader reality staring at us in the face, how we got here, and what it may well mean. Sander Tordoir (CER) gave a strong summary of the major turnaround of the Chinese economy, the “second China shock” (overcapacity and low household consumption as the real estate bubble also burst), which created conditions for a massive manufacturing surplus (around 10% of China’s overall GDP[6]) being exported to the rest of the world. What does this mean for the competitiveness of Europe at home and abroad, our prospects for reshoring vs offshoring production, implications for labour markets, exports, profitability and investment? Europe is more reliant than ever on China for imports in key sectors, and the US for exports. And China is poised to take a dominant share of global markets in an expanding number of sectors.  As Tordoir said “if we do not pursue our own industrial strategy, we will import China’s imbalances, shrinking the EU’s manufacturing base, which is more productive than services, and innovation”. No one has the faintest idea about any of this in the competition world. As competition experts we do cookie-cutting market definition and competitive assessment with our little bag of tools, using backward-looking historic sparse data, oblivious to big transformative changes. Another one: what does anyone in competition know today about comparative technology advantages – for instance clean tech, in a world where greentech competition between blocks will be the name of the game? Any idea? Europe is actually strong in this area apart from solar panels and batteries, including green innovation and patents. Does anyone know?

Constantly updating knowledge of this broader reality should be mandatory for officials in competition as well as in industrial policy and trade. The analysis of markets for purposes of antitrust assessment cannot be ignorant of the macro landscape and its evolution, and of shifts in competitive advantage which affect competitive constraints. Yet it is.

Second foundational principle: understanding the interplay of policy tools in the pursuit of common goals. Tools are sometimes complements, sometimes substitutes and sometimes also in conflict, and all of that needs to be explicitly worked through. As Nils Redeker (Delors Institute) put it: “Industrial policy is about subsidies but in many instances, levers like trade policy, regulation, public procurement are much more important. This also means that if you are committed to supporting a certain sector for strategic reasons, you need to make sure that different policy levers are all pulling into the same direction”. While trade and competition policy have been operated as independent goods for as long as one can remember both in the US and Europe, it is now understood at least in the US that the “Washington consensus” of efficiency, free and open borders which made cheaper goods available to people as consumers also disadvantaged and disenfranchised them as workers – as production was offshored. And not only did this have significant distributional impact – critically, it also did not help competition at home: because (a point well made again by Jan De Loecker at the event) free trade that lowers input costs when there are barriers to entry into markets and economies of scale means incomplete pass through of those lower input costs by incumbents with scale advantages, and disproportionate margin increases which are not dissipated by entry – “what you’ve then done with trade policy is you just empowered the incumbent” which you then try to tame through competition enforcement; “not integrating the two policies will not deliver procompetitive effects”. 

Another example is digital trade: free trade in digital goods (originally intended to “oil the machine” around trade in physical goods, but now mostly signifying trade in data) has been a pillar of digital giants’ world domination; but this is in direct contrast with initiatives to curb their power at home through antitrust – again liberal trade policy runs counter to competition policy.

Tordoir (referencing Michael Pettis) also said at the event: “Tariffs, wage moderation policies, and competition policy all to a degree represent trade-offs and transfers from consumers to producers. Doing such transfers only makes sense in areas where we have a real shot of regaining competitiveness”.To expand: “we have to think about to what extent do we support our producers.  You don’t want to do so across the board, there may be areas where you just want to accept cheap imports because we have no industry like solar panels like Draghi says; but in other areas you may want to make sure that we regain or retain competitiveness”.

Which takes me to the third foundational principle: selective approach. Draghi identified 10 strategic sectors for discussion in his report, and for each of them he sets out an industrial strategy for Europe to recover ground. This requires agreement on the objectives, the combination of tools and the funding – not as a generic endeavour, but as a specific effort for specific sectors. Redeker: “the key question is not whether we will have industrial policy in Europe or not – but whether we will have a coherent joint EU-wide industrial strategy, or we’ll end up with 27 national industrial policies…(First,) we need to make a decision on what sectors we want industrial policy for, and why. Industrial policy comes with tradeoffs, and we need a good understanding of what we’re doing and why…but the EU is not there yet… only in the last term, there are more than 50 technologies and industries deemed to be strategic, ranging from heat pumps and solar panels to robotics and biomonitors…. This mixes together industries with very different objectives – on solar panels the objective is to diversify supply, reduce dependence on China, but this requires a very different tool set than supporting an infant industry like hydrogen where we want to build a dynamic competitive advantage in Europe.  Yet this is all mixed together in the Net Zero Industry Act”. 

Tordoir also mentioned the need for “shades of gray”: “the EU should shift away from a yes-or-no debate on industrial policy to a nuanced when-and-how discussion considering the characteristics of each industry, its prospects and its strategic value”. And “(it) should focus its subsidies on sectors where short-term assistance is needed to help infant European industries, such as hydrogen, achieve scale, or employment-rich sectors like cars, where we have a real shot at making the transition to electric vehicles (The EU is already a net exporter of EVs). It should also prioritise support to markets for goods, like wind turbines, in which a global oligopoly or duopoly is likely to arise, and in which a dependence on China would be risky. That way, the EU will help new businesses to grow while minimising handouts to those that do not need them”[7]).

Competition policy comes into this in two ways: one, as a fundamental value in the design of the strategy for each sector – not to cement incumbents into place and favour national champions, but to foster competitive structures. And two – and this is the most controversial part – competition policy has been historically a “cross sectional tool”, by which I mean the toolkit has developed as an industry-agnostic kit: market definition based around a prescribed hypothetical 5-10% price increase whatever the sector, market power proxied by current “share of revenue” whatever the sector, limited interest for other indicators of power (from data to capital intensity), vertical deals presumptively benign, etc. Will return on this below, but it is time to accept that the analysis of competitive conditions, asset re-combinations, incentive effects, may well require sectoral differences.

Fourth foundational principle: acknowledging and working through tradeoffs. There are inevitable tradeoffs when pursuing competition, investment and growth.  The commonly accepted wisdom is that competition is the “engine” for investment and growth. That said, in practice there are tradeoffs between short term and long term, price effects and investment effects, consumers and producers. This is something competition policy never fully acknowledges because it is premised still in the main on a neoliberal neoclassical view in which productive efficiency is the goal, and economic “tools” are designed (crudely) to “maximize” a notion of “consumer welfare”. There is a strong presumption that “competition delivers” – low prices AND innovation AND the right amount of competitive investment – but there is limited interrogation of whether this is true, for what sectors, type of investment, or innovation. The persistent dispute around telco mergers (also a hot topic following the Draghi report) is evidence for this.[8]

In summary, “we need to make sure that we don’t only look at the European market in isolation, but think of the European market within this globally changing landscape. And for markets that are tending towards oligopolies globally, we may, in fact, be better off as Europe creating larger firms of your own that can compete. Airbus is the classic example… that’s probably safer to do so in markets that are already Europeanized, like the goods market which is more integrated than services markets, so there’s less of a risk of creating a shark in a small pond. You can create a larger shark, and it will face competition from firms in other European countries and from the rest of the globe” (Tordoir).

• What implications for industrial policy and trade?

The obvious implication of a “joined up approach” from the perspective of industrial policy and trade is that both need to explicitly embed competition values.  This means that – as much as competition policy needs to be schooled on macro developments – industrial policy and trade need to be informed by competition expertise.

For industrial policy, the first order issue remains there is no coordinated effort at EU level, and there is the big question of funding instruments (Coeure’ Redeker). This is the major obstacle any EU-level industrial policy would need to overcome.  But as first principle, “joining up with competition” must mean not indulging national fantasies for local champions but designing the combination of levers taking into account the need to foster/preserve competition. As noted by Redeker “competition policy matters to ensure policy coherence. (…) We know from research that industrial policy works best if you target competitive sectors. There’s both theoretical reasons for that, and also empirical research on cases like South Korea showing that when industrial policy worked, it worked because it targeted sectors where there was fierce competition and firms were really competing. (…) Second, we should really prioritize firm-neutral instruments, and that’s regulation and public procurement. (…) I would like experts on competition policy to be in the room where these kinds of instruments are designed in order to make sure that they are working on a competitive basis”.

In trade, Europe remains a “WTO faithful” committed to WTO principles and the neoliberal myth that open economies and globalization can deliver competitive markets. As mentioned this consensus has come under much pressure as it is recognised that the pursuit of efficiencies has led to incumbents being able to exploit lower input prices and scale economies to become more profitable and keep others out. In the US, alongside tariffs to protect domestic workers and re-onshore production there has been explicit recognition that trade policy in the original Bretton Wood conference 80 years ago had “antimonopoly” as an important guiding principle to protect the future from the antidemocratic tendencies of oligarchs (following Germany in the 1930s). Trade Ambassador Tai[9] has talked about protecting jobs and “growing the middle class from the bottom up and the middle out” as explicit joint objectives of the Biden administration’s trade and industrial policies.  Competition is regarded as a key component of this – involving rejection of national champions and consistency with antimonopoly enforcement.

• But now let’s get specific on competition policy

Now let’s turn to competition policy practice and enforcement.  What would a more “joined up” approach mean here? How do we map these principles into practice?   European antitrust agencies have been extremely protective of their patch. There has been little appetite or effort to engage with, let alone drive, a reflection on goals, values and methods.  In the US, in the early days of the Biden Administration noted legal scholar Tim Wu (advisor to the White House) circulated a memo on “The Grand Unified Theory of Antitrust Revival (And a Plan for Action)”, which was the basis for the 2021 Biden Executive Order stating boldly that fighting corporate concentration was a goal of the Administration.  In contrast, Europe has contemplated no revival and no rethink. The end of the neoliberal consensus is not something anyone mentioned or reflected upon in competition circles (I tried, no takers). Instead of working on waking up the “sleeping giants”, how Wu called the underutilised enforcement powers of the federal antitrust agencies, we have had extremely marginal rewrites of guidelines – Market Definition, Art 102. A total waste of time. These produce excitement in the Brussels Antitrust Bubble and academic legal circles, but truly no have no impact on the ground.

So we were already to start with in a situation in which – Draghi or not – antitrust enforcement in Brussels and much of Europe has been weak, obsolete and ineffectual.  There’s a number of major problems that – Draghi or not – remain to be addressed. To name but a few:

Inapt “theories of harm”:   We are stuck between tremendous lack of imagination, on the one hand, and lack of courage on the other.  In mergers, “theories of harm” are always about narrow effects on specific markets and involve either loss of competition between existing assets, or “foreclosure” of rivals arising from integration of assets in some sort of vertical or conglomerate relationship. That’s pretty much it, with frequent contortions to articulate “leveraging mechanisms” including little toy models with silly “calibrations” based on three data points. We are just beginning to recognise power arises in “ecosystems”, not just from being big in a narrow market – but we don’t know how to deal with it. The Booking/etraveli prohibition was mostly a political statement (let’s block a non-Big Tech old-style conglomerate digital merger – who remembers one-stop-shopping in Guinness/GranMet?) but not a blueprint for anything.  Conversely we waste time making up a preposterous foreclosure story (of rival robot cleaners from Amazon’s marketplace) instead of articulating a story on data collection in Amazon / iRobot.  Theories of harm in conduct cases are also majorly wonky: the Amazon and Apple abuse of dominance cases for instance were framed bizarrely as “exclusionary” cases when they were all about extraction and exploitation. Such a waste of time.

Just cannot deal with incipiency: Antitrust remains terrible at dealing with incipiency. Because the concern is “creating or strengthening dominance”, the requirement is there needs to be material increment from something we can measure with something else we can measure. But the whole history of lack of any enforcement in digital mergers demonstrates our total inability to deal with a potential issue which is now small but can grow very fast.  The cases are countless – in Microsoft / Activision, the improbable theory of harm was console foreclosure, instead of cloud gaming advantage.  Agencies have more or less accepted by now the idea of “killer acquisitions” (or their “reverse” version) in digital[10] (see Adobe /Figma, though the deal was withdrawn). But the whole of AI lack of enforcement to date is in the same category – instead of pursuing “aggressively weaponizing scaled assets to gain first-mover advantage” as our theory of harm, we wring hands and admire the problem, then conclude “there is no creation of market power in any relevant market to date”. Really?

Persistence of the “efficiency” myth: Notwithstanding the comprehensive takedown of “efficiency” as a pursuable goal, it remains pervasive in our language and posture. Dubious practices are defended on grounds they are “efficiency enhancing”, state aid is allowed if it demonstrably “enhances efficiency” in dealing with “market failures”. We are still stuck there.

Again and to be clear, all of these issues (and more) require urgent fixes but they pre-date Draghi. Then in came Draghi, and he layered on top the issue that Europe has serious problems of growth, productivity, innovation, AND our firms are sub scale in terms of size, they don’t invest enough, innovate enough and are just not large enough to compete internationally.  But instead of engaging constructively with this hypothesis, the response of the antitrust Bubble has been to close ranks even further and persist in the cry that “competition matters!” “this is the slippery slope to national champions!”. The Bubble and its well-meaning academic supporters prefer to travel the road of ”competition matters, don’t give up on competition” (which no one is seriously suggesting aside from loony fringes) instead of thinking seriously about what policy R&D one needs to initiate.

So what does Draghi’s exhortation that we join competition with industrial policy and trade mean for how we should think of antitrust enforcement in practice – on top of the existing issues?

I believe the need for “size and scale” is important and real for Europe, but of course being laggards depends on a host of structural issues and not on merger control in most industries, or relaxing enforcement.  Indeed Draghi’s suggestion that the telco sector in Europe is overly fragmented and could do with some consolidation has attracted almost universal pushback.[11]  The response has been one cannot “magic up” a European market where there isn’t one, just to fabricate lower market shares and pass a merger. True. That said, while I do agree that both merger control and enforcement against abuse are almost comatose in Europe already, and cannot be relaxed further (4,000 deals approved in Brussels in the last 10 years, only 9 prohibitions overall! 10 Article 102 decision in the last 5 years, with no impact on the ground!), I also think we should be a lot more selective in our approach – in terms of how we identify markets and industries in which to intervene – and more creative in our conditionalities. 

Specifically, and linking to the discussion above as well as Benoît Cœuré “questions for reflection”, I believe we should adopt a more sector-specific approach in which we coordinate competition enforcement more closely with industrial strategies at the sector level, and trade developments. The impact of asset reallocations through mergers depends on how a sector or market is exposed to global changes in competitive advantages and macro stance turnarounds. Yet this is almost never discussed in cases. There is formulaic market definition analysis which uses a cookie-cutting approach (simple price tests, shock analyses, some internal documents on perceived competitive threats), there is a perfunctory discussion of “barriers to entry” – yes in principle the analysis is “case by case”, but the understanding of the broader landscape is very shallow. Even less is the understanding of investment and innovation dynamics. And critically, we don’t know how to place deals in context with the kind of macro reality, technology and competitive advantages discussed in the first Section. I have been involved in countless deals in which all that mattered was market share increments in 172 individual national product markets even though the industry was global and China was looming all over.

This is anathema in terms of today’s competition rules, when case review is mostly initiated as a routine exercise with mechanistic thresholds based on observables (revenue) applied across the board. But this means resources are scattered across “undeserving” cases. We need a selective focus: designate sectors, firms and markets as more or less strategic and view priorities through this lens. If we saddle an agency with a lot of routine work, this goes at the expense of resources going to critical cases. Choose sectors which deserve close scrutiny in terms of their “strategic” importance.  This is not a “free for all for other mergers”. Use presumptions to filter cases and reverse the burden of proof, as suggested (for a long time) by Tommaso Valletti.  Of course, firms which want to consolidate via mergers will be quick in labelling themselves as strategic sectors, so care is needed.

Second, we need to work through trade offs much more explicitly. We tend to be focused on short-term effects, that’s the antitrust bias, but we need much more substantive work to understand the nature of trade offs over time – with investment, with sustainability.

Third, we need to consider what conditionalities may be acceptable, and what mix of instruments. Do we want to be “lighter touch” and leave more discipline to ex-post enforcement? Draghi suggest this may be better than stopping some mergers outright.  Again, hard to make generic cases. We allow too many mergers through, we are a merger factory, merger enforcement is comatose today. Do we want more mergers for which we then have to unscramble eggs? And conduct enforcement is notoriously slow and ineffectual.  Again we would need serious work on a selective sectoral approach to consider the mix of instrument in different sectors. Benoît Cœuré: “the question is about the kind of safeguards that we want to see in place if we want to make it work. And maybe some of us will say there are no possible safeguards and it’s never going to be possible. But we have to look in good faith at these safeguards and see if it can if it can work”.  Exactly. By the same token, can we accept undertakings on investments as a quid-pro-quo for allowing a merger through? This has so far been anathema to the competition community, and Tomaso Duso explained clearly why, but perhaps needs thinking through. Let’s put together serious multidisciplinary working groups, and be creative.

As stated by Benoît Cœuré at the event, “competition enforcers can ignore industrial policy, but industrial policy will not ignore them”. We need “a change of paradigm in the design of European economic policy: moving away from the traditional silo approach of competition and trade and industrial policy towards a unified approach”, “stakes are extremely high, and competition enforcers should play their part to meet the challenges”, instead of “sticking to a strict interpretation of their mandate, for fear of a return to what they see as the old competition policy driven by politics and by national interest”. And, boom: “competition policy is industrial policy” (echoing Tim Wu); “we have an existential task in making industrial policy work and be pro-competitive; if we’re not part of that discussion, we won’t like the results”.  Could not be clearer.  I am not a one-woman think-tank, and no one has ready-made answers. Of course there is scope for major derailment into politicization – a point everyone agreed on (see Tommaso Valletti’s account of his own experience with state aid and IPCEIs). But that said, it’s unacceptable for the competition community to hide behind the usual excuse that “proposals for change are not specific enough”. Get to work, not in silos. Because doing nothing IS a policy choice, but not one we can afford.

---

[1]  Zach Myers, 24 October 2024 Competition policy must reflect Europe’s reality, not its aspirations | Centre for European Reform

[2] Cristina Caffarra and Burcu Kilic, 7 March 2024, Re-joining trade with antitrust | CEPR; Cristina Caffarra and Nathaniel Lane, 24 April 2024 Not a ‘side dish’: New industrial policy and competition | CEPR; Cristina Caffarra, 18 September 2024 Draghi’s real message on European competition enforcement: “Not delivering on innovation and growth” | CEPR; and Cristina Caffarra, 30 September 2024 New Commission Mandate: Why “Modernize” Competition Policy in Europe?

[3] The hotel meeting that worries the world’s biggest corporations – POLITICO

[4] Mario Draghi, 9 September 2024, EU competitiveness: Looking ahead – European Commission

[5] New European Commission, New Mission: “Modernizing Competition Policy”- How? | CEPR Webinar, 28 October 2024. Speakers: Benoît Cœuré, Cristina Caffarra, Nils Redeker, Sander Tordoir, Jan de Loecker, Tomaso Duso, Tommaso Valletti, Alexandra Geese.

[6] Brad Setser, “The surprising resilience of globalization: an examination of claims of economic fragmentation”, in Strengthening America’s Economic Dynamism, edited by Melissa S. Kearney and Luke Pardue, Aspen Institute, 2024.

[7] John Springford and Sander Tordoir, June 2023, “Europe can withstand American and Chinese subsidies for green tech”, Centre for European Reform.

[8] Tomaso Duso, Massimo Motta and Tommaso Valletti, 17 September 2024, Draghi is right on many issues, but he is wrong on telecoms | CEPR

[9] Conversation with US Trade Ambassador Katherine Tai, and Simon Johnson MIT, chaired by Cristina Caffarra, online 22 July 2024 Competition Policy RPN – Shifting the Trade Paradigm | CEPR

[10] Cristina Caffarra, Greg Crawford, Tommaso Valletti, 11 May 2020, ‘How tech rolls’: Potential competition and ‘reverse’ killer acquisitions | CEPR

[11]             See Duso, Motta, Valletti, op. cit.