FRANCE STRATÉGIE & OECD WORKSHOP ON NEW INDUSTRIAL POLICY TOOLS Paris, 17 October 2022
Session 2: Can industrial policy be consistent with open and competitive markets? Benoît Cœuré President of the French Competition Authority
Logo Name
2|
Introduction Speaking as competition enforcer in a discussion on industrial policy, I risk being the one who spoils the party. I was reassured however by Philippe Aghion’s constructive comments on how industrial policy and competition policy can complement each other. The discussion on competition policy vs industrial policy is perennial, but today’s stakes are higher than ever. Industrial policy is in high demand in the face of technological change, climate change and the energy crisis. Competition policy is in equally high demand, with inflation running high and big players extending their footprints in our economies.
Will industrial policy and competition policy ever meet? There is more policy space than politicians generally think Too often, the policy discussion is about picking a winner between industrial policy and competition policy. This is not how it should be. We should be discussing complementarities and trade-offs between instruments. And the good news is: there are many more instruments than meets the eye. To mention but a few in the case of Europe, these instruments are antitrust, merger control, the State aid framework, the new package of digital regulation (including the Digital Markets Act, Digital Services Act, and Data Act), national rules on foreign direct investment, EU and national sectoral regulations, and the soon to be adopted EU regulation on foreign subsidies. To illustrate - there will be strong complementarity between the DMA and competition law. Indeed, looking ahead, antitrust enforcement can refocus on « frontier » players and activities, while the DMA will address gatekeepers’ practices which are known to be harmful. In the merger control space, the combination of DMA article 14 and article 22 of the merger regulation will empower the EU to identify and block killer acquisitions by designated gatekeepers. As for merger review and the foreign subsidy regulation, they serve different goals but are complementary by design. Indeed, the latter regulation sees an internal market distortion when « a foreign subsidy is liable to improve the competition position of an undertaking in the internal market and where, in doing so, that foreign subsidy actually and potentially negatively affects competition in the internal market ». Given this broad choice, policymakers should focus their efforts on picking the right instrument. This will have the added benefits of enhancing public accountability and forcing governments to clarify the objectives they pursue –which they often don’t.
Industrial policy and competition policy can support each other Enforcing competition policy absent industrial policy would overlook the fact that some level of market power is good for innovation and resilience, and that cooperation among firms and/or government intervention is sometimes needed to achieve economies of scale and manage externalities, e.g. climaterelated ones. Speaking about resilience, I was interested to read in the International Monetary Fund’s October 2022 World Economic Outlook that the bottom 40 % of US firms in terms of their mark-ups have fully passed higher production costs to their clients, while the comparable pass-through was only 60 % for the top 20 % - that is, firms with market power. That market power can enhance resilience to shocks should not come as a surprise. After all, under perfect competition, margins should be zero and the pass-through should be 100%.
|3 On the other hand, industrial policy absent competition policy unduly favours incumbents and discourages innovation. We do need competition on the merits. Those who claim that competition enforcers stifle innovation and look in the rear-view mirror are misguided. Dynamic efficiency is part and parcel of the competition toolkit and was used in recent EU merger bans.
So, which kind of industrial policy would I like to see? Let me suggest two simple rules: Rule #1: Target sectors rather than companies. This will minimise rent seeking, and conflicts of interest and it will increase chances of industrial success as well as options for consumers. BARDA-funded Covid vaccine grants benefited as many as 7 laboratories! In the European context, given narrow and/or concentrated domestic markets, this points to EU-wide rather than national policies and initiatives. Rule #2: Focus industrial policy on creating new markets, services and technologies. The Important Project of Common European Interest (IPCEI) on hydrogen is a case in point. When implementing Rule #2, I will feel more comfortable if a few guiding criteria are met: (i) government support should be upstream rather than downstream, (ii) it should benefit a large number of companies, and (iii) it should be proportional to what is needed, with a claw-back; (iv) risk should be shared, and (v) there should be positive spill over to the rest of the market.
And how can we improve competition policy? The Commission’s ban of the Siemens/Alstom merger is often quoted as an example of failed competition policy. This is the worst possible example! The time horizon underlying the Siemens/Alstom decision was very long. There was much talk at the time of Chinese competition being an immediate threat to European players, but the merger was blocked, and I don’t see many Chinese trains in the Gare Saint-Lazare. In my view, the real failure of competition enforcers is elsewhere: we have let global gatekeepers steadily build market power in a way they can now abuse to squeeze their clients and foreclose their rivals. After Siemens / Alstom, politicians have proposed to review the EU merger control framework to account for state subsidies and global competition, to extend the timeframe of merger reviews, and to give the European Council the final say on merger decision. All these proposals can be discussed but let me make a plea: enforcement should remain fact-based. Foreign entry can be slower that thought; efficiency gains promised pre-merger often don’t materialise. Too often, politicians buy into CEOs’ narratives which rarely come true. This is not to deny that there are things competition enforcers can do to better reflect the reality of fastchanging markets. To illustrate: we can account better for efficiency gains in the merger review process – keeping in mind however that it is incumbent on firms to show these gains. In the digital space, we can try to understand better how systemic gatekeepers can choke emerging innovative markets, particularly through conglomerate effects, and act accordingly. And we can wrap our mind around the economics of data in the non-digital space, e.g., in transportation or health, and how they impact competitive dynamics. With that, I am very confident that competition policy and industrial policy can be mutually supportive, as they should be.