chevron-down Created with Sketch Beta.

Antitrust Magazine

Volume 36, Issue 1 | Fall 2021

Interview with Pierre Regibeau, Chief Economist, Directorate-General for Competition, European Commission, Brussels

Anna Lyle-Smythe

Summary

  • Interviewer: Anna Lyle-Smythe
  • Interviewee: Pierre Regibeau
Interview with Pierre Regibeau, Chief Economist, Directorate-General for Competition, European Commission, Brussels
Felix Fernandez Gonzalez via Getty Images

Jump to:

This interview was conducted for Antitrust on July 14, 2021, by Anna Lyle-Smythe, Editorial Board, Antitrust Magazine and Partner, Slaughter and May, and Benoit Durand, Partner, RBB Economics.

Antitrust: Hi Pierre, for the benefit of our readers who may not be familiar with the economist function at DG COMP, could you start by telling us a little bit about the role of the Chief Economist and the Chief Economist’s Team (CET).

Pierre Regibeau: As the Chief Economist, my main role is to directly advise Margrethe Vestager, the Competition Commissioner, and Olivier Guersent, the DG COMP Director General, on economic matters relating directly to cases or to broader policy issues. In order to provide this advice, of course I rely on a team of about thirty economists.

Now it’s still important to distinguish between two parts of our role.

One is advice. So even if we haven’t been closely involved in a case, when the case team comes to Margrethe and she has to make a decision, we will provide her with our advice.

But, more often than not, we also have members of the CET who are directly involved in the case and are essentially members of the case team, even though they still report to me rather than to the case manager.

These are really the two facets of our work.

Antitrust: The CET was created now over fifteen years ago. Do you think the role has changed, and if so how, over that time?

Pierre Regibeau: The role certainly has changed, if only because the size of the team has changed. At the beginning, there was fairly little room for being systematically involved in some of the details of the cases, so that is something that has developed over time.

That is important because the more you are part of the team, the more that you cooperate, the more you can have an impact on the assessment of a case.

That has been an evolution in this way. We first got involved a lot in mergers, and that relationship between the CET and the mergers team within DG COMP is still the strongest. But over the last few years we have been involved much more in antitrust and in State aid, and much stronger relationships are emerging there.

Then of course, also we have this more general policy advisory role. Especially under the new mandate of Margrethe Vestager, since she is involved not only in competition economics but has a number of other tasks, we have also to advise on a broader set of issues including issues relating to our relationship with the other Directorates-General (DGs).

Antitrust: So you are involved in pretty much all disciplines across the board—mergers, antitrust, State aid—and, as you said, you have only about thirty people working with you. Is there some sort of selection mechanism that is in place to choose which cases you are involved with?

Pierre Regibeau: Well, I just spent two hours setting the allocation of resources for antitrust cases!

Individual economists are of course not 100 percent free to choose but we do try to take people’s preferences into account, and especially we do try very hard not to over-­specialize. I really want absolutely everybody over the course of their careers to be comfortable with at least two of the three tools, and ideally with the three, so we try to plan people’s careers so that they get exposed to all three instruments.

It is not only good for their careers, but it is good for us because as economists of course we know that there are common principles that run across the arguments. If we want to make sure that there is economic consistency with the way we deal with State-aid cases and antitrust cases and merger cases, it is very important that people have had all of those experiences.

In terms of how we deal with this, quite frankly we don’t have a single person in the CET who is involved in only one case. That just does not exist. I’ve got people who are actually involved in eight cases, so that tells you a little bit about the pressures that we face.

Antitrust: Are there cases that you would want to be involved in but you cannot because of time constraints or resource constraints?

Pierre Regibeau: It is happening. As I told you, I just went through the allocation process for antitrust, and in my list I had important priority cases where essentially the case team gets full access to our resources. Then there is the next set of cases that we are following closely and we perform specific tasks, as required by the needs of the case. And then there are matters which we are just following closely. And then there are things where we follow from further away. And there are things we don’t follow at all, as that is not required.

So yes, choices at this stage have to be made. But—in the end—resources management is always about prioritization.

Antitrust: In terms of CET allocation to cases, who is making the request? Is it you, Pierre, saying, “We want to be on that case,” or is it the case team saying, “We want you?”

Pierre Regibeau: There is a regular exchange between the CET and case managers. Usually, it is the case team that flags the cases that require support by the CET. But, this request can also come from our side, if we hear of a case we think we ought to be on.

Antitrust: One of the things you mentioned earlier, Pierre, was your role in liaising with other DGs at the Commission. What about interagency cooperation at the economist level?

Pierre Regibeau: One has to draw a distinction between our relationship with other European agencies and agencies outside of Europe.

With European agencies, we have actually a European network which involves both lawyers and economists. A part of it is a kind of Chief Economist’s Teams network. So we have close contact, we talk to each other, we have two conferences a year where we have people presenting papers but also we discuss cases. We do get requests for advice from the smaller economics teams from national authorities. So the cooperation there is very frequent and very good.

Outside of the EU, there is a communication and discussion on some cases. Usually, this coordination does not happen directly at the level of the economists; it would be at the level of the case handler or the Head of Unit or even the Director reviewing the cases. From time to time, if there is a specific economic issue to be discussed, a member of my team or a member of my team who isn’t on the case team would attend, but I would normally not get involved myself.

There are some exceptions. As you know, there is a discussion group with the U.S. Federal Trade Commission on pharma. For this the Chief Economist’s Team is a bit more directly involved and that is something that I follow a bit more closely personally.

The relationship with the BRICS Competition Authorities (Brazil, Russia, India, China and South Africa), for example, is less developed. I clearly regret this as, before joining the Commission, I was an active member of the network of BRICS Competition Authorities.

Antitrust: Switching gears a little bit and moving to the role of economics in enforcement, notably in relation to Articles 101 and 102. How well aligned are current enforcement practices with competition economics principles, and how do you think that has changed over time?

Pierre Regibeau: I think one can obsess a little too much about the distinction between effects-based and non-effects-based. What you want as the Chief Economist is to make sure that the theory of harm makes sense and that everything is based on evidence.

Actually, I think that people probably overestimate somewhat the importance of all the empirical work and simulation that we do and underestimate the crucial role of keeping everybody’s nose on a consistent theory of harm, a well-articulated one. So it starts very early in the case: “Why are we doing this case? What are the arguments? Therefore, given the arguments, what is the evidence that you need?”

We should also think early on in cases about where we are going to land, in particular in terms of remedies, if we don’t have a good remedy, frankly, given that we have constrained resources, and this is not necessarily a high-priority case. This conceptualization, I think, is absolutely crucial from the beginning to the end.

Then, of course, you need different kinds of evidence. The first one is evidence to support the assumption of your theory of harm, and that should certainly come before we go looking for effects.

After that, frankly, for cases where you will be able to find effects, are cases where empirically it is going to be very difficult and you are only going to be able to argue for likely effects.

So this thing about, “Oh, you really need the fingerprints to see effect or not”—I think that is not really the main issue.

Antitrust: There has been talk about the fact that enforcement of competition law has been slow, has been difficult and challenging, and a way to be more effective is to move towards ex-ante regulation, as with the Digital Markets Act (DMA) being proposed by the Commission. What are your views on the current balance between “ex-post enforcement” and “ex-ante regulation”?

Pierre Regibeau: When you look at the DMA, I think there are two aspects. One is to see it as an ex-ante regulation corresponding to and accompanying antitrust. Another one is to see it as an ex-ante regulation dealing with a somewhat different problem.

Let me start off with the second aspect. The fact is that the DMA is not per se aimed at dominant firms. There are no references to “markets,” to “dominance,” to “market share.” The DMA is going after firms that are big, important enough so they significantly affect the European Economic Area (EEA). These firms also tend to be “tentacular,” in the sense that they influence many different sectors.

That is the part which is different, that is dealing with aspects of problems that we would not deal with anyway because we could not establish dominance. You know we don’t care about size per se and about the overall EEA economy; we care about antitrust markets and competition.

Now, of course, there is also a complementarity, and from that point of view I see—from my selfish point of view—a useful role for DMA acts as a kind of relief pitcher.

What is a little less obvious is when you start having in the DMA rules that don’t have this kind of origin and might be based on a feeling that those behaviors are bad, but maybe not rely on as much evidence that economists like you or I would like.

I think that’s one of the reasons also why in the DMA you’ve got blacklists and greylists. They do realize that when you move to the greylist, we need to kind of tailor those broader requirements to the specific cases.

While I like this approach, I think that if you tailor ex-ante regulation too much, it loses much of its effectiveness and simplicity. So personally—this IS a personal point of view, and I know some people at the Commission would not agree with me—I would be in favor of having a relatively limited number of rules initially under the DMA that would be rules that are simple, ex ante, applied broadly, maybe with some tailoring across-the-board, and are based on sufficient evidence that those behaviors are actually bad.

And of course, you don’t need to get stuck with those early rules. The CMA includes investigatory powers to find, maybe faster than we can, other harmful behaviors, and then they can add them to the list. We can also eventually feed them other behaviors emerging from our antitrust cases and so on.

And I don’t think the DMA in the current version does a terrible job of that. Still, I am a little bit afraid of what might happen through the approval process, and especially when you see some members of the European Parliament, who appear to wish for an immediate robust and all-encompassing regulation. I really hope that they will understand that we need to walk before we can run.

Antitrust: Moving on, because there are a lot of initiatives at the moment by the European Commission which we’d love to discuss with you. One of those initiatives is the revision of the Guidance on market definition. That document was issued in 1997 and hasn’t been touched since. The Commission recently published a working document explaining that it is going to be revised. We have seen some agencies suggesting that we no longer need to worry about market definition. As an economist, how do you see the role of market definition, in particular having in mind that there are a lot of challenges, notably coming from the digital sector, on how do you deal with ecosystems and data etc.?

Pierre Regibeau: The first thing, as you would expect, as an economist my answer would be to say, “I don’t need market definition.” What I look at are competitive constraints—in mergers whether these are likely to decrease or not, and in antitrust whether there is something that prevents legitimate competitive constraints to be exercised.

On the other hand, we still need market definition to translate this meaningfully into existing law and to translate this into things that the courts can understand.

We’ve got to live with it, so if we live with it, we might as well try to adapt it.

Now I would say that I think the basic logic of market definition is still pretty good and does not need to be changed. For example, many people have complained about geographic market definition. Frankly, I think that if you just look at who exercises an actual or potential competitive constraint in a given market, and you consider the kind of time horizon which is appropriate for an industry, there is nothing that needs to be changed there.

In this respect, it is important, though, of course to not get stuck with what was, rather unfortunately, written in the Guidelines about the two-year horizon. This two-year horizon we never applied. We have always considered that in different industries the proper horizon was different, and that is definitely something that we have to do.

Now, there are of course other kinds of newer issues that have to do with platforms, and with digital, for example. In platforms, of course, you know the issue is: Do you have a market for platforms or do you have platforms involved in a series of markets?

But there are deeper issues. First, there is the fact that many digital platforms can discriminate across consumers quite effectively. As economists we know that, once you have quasi-perfect discrimination, the whole traditional approach to market power kind of falls down because there is no longer a link between market share and the hike of the price, at least directly.

This has potential consequences for the use of market shares as a proxy for market power, for example. Is it still useful in one way or another? Market share across what? That’s an issue.

More importantly, also not only can those firms discriminate between consumers, charge different prices—many don’t, but they could—but they certainly can also choose what kind of products to show you. This choice depends on the information that that company has on the consumer’s preferences. And then, when you think of a merger between two companies that have different sets of products, somewhat overlapping, and also have different sets of information about consumers. If you think a little bit about that, to really understand a merger you not only have to see the overlap between the products that those companies can offer, but also the overlap between the information they have about their consumers.

These are much deeper issues, which again we are not going to address this time, but that tells you that yes, for the future, given that things change, maybe there should be another revision of the market definition ten years from now.

Antitrust: I would like to bring up a different point about the history of DG COMP’s practices. Companies that are looking at mergers will often focus on the synergies and benefits of their deal and yet we still haven’t had a case that has actually been cleared on the basis of efficiencies. Is that because mergers are not generating efficiencies, or is the standard of proof just too high?

Pierre Regibeau: It’s a very complex question. Efficiency is something I’ve said from the beginning when I took the task that I care a lot about.

Now, what are the difficulties?

The first is that it takes time to really investigate efficiencies. If as a lawyer you are really serious about playing the efficiencies card, that is something you have got to start early on, before notification; otherwise, we are not going to have time.

Unfortunately—I’m not saying this for all lawyers—there is still a bit of a feeling that if you come up and you start with efficiency, you are saying, “Oh, I have a bad case,” like it’s acknowledging that “Yes, otherwise the price increases would be very significant.”

I think this is a mindset we have to get away from because, as an economist, if you don’t have an efficiency expert, I do expect some increase in prices. So that ship has kind of sailed. If you really think you have strong efficiencies, start with the strong efficiencies.

Now, it’s true that we have never officially cleared a case or written in a decision “Oh, it was bad but because of the efficiencies we do it,” but there are cases where the final decision was certainly influenced by the belief that there were non-negotiable efficiencies.

Do we expect mergers to really create lots of efficiencies? Apparently, some judges in Luxembourg do believe that, that there should be a presumption that they create efficiencies.

Is it true? As an economist, I look at the empirical evidence, and if you look at the empirical evidence both on merger profitability and on efficiencies, with a bit of a difference between the United States and Europe—where mergers tend to be found to be a bit more profitable and a bit more efficiency-enhancing in the United States than in Europe—the answer is that: “Well, at most it’s 50/50.” In many cases actually you have inefficiencies of mergers, which we never bug people about. In my mind, this means that there should not be any presumption of efficiencies but that we should be open to evidence-based efficiency arguments.

So what this tells you is that overall probably it is not that often that we have strong efficiencies in mergers, but precisely because we cannot have a presumption that there are.

I am going to evoke very briefly, without telling you which one, but a case which involved a claim that one of the companies was doing an innovation that would make the whole production process less polluting. Of course, if you tell me that, first I have to see, “Well, let me see if this innovation is credible or not.” There were some uncertainties there, but we will not get into the details.

Then I would ask you also about merger specificity, which is another reason why sometimes efficiencies do not win the day: “Okay, fine. If you’ve got this great process, couldn’t you license it, for example, for a price?”

The nice thing in that case is that the type of innovation that was useful was not the kind of innovation for which you could get strong intellectual property protection, so licensing was not in the cards. In that case, I was quite open to the idea that the diffusion of this technology to the bigger merger partner was actually a merger-specific benefit.

This at least gives you an idea of all the conditions that need to be fulfilled for us to take this seriously even if we are not biased at all.

Antitrust: You mentioned pollution which brings to mind another really hot topic, namely the drive for positive action on climate change. You have made some statements about green efficiencies. What is the economic case for sustainability being a factor of competition enforcement, and do you think that needs a change in approach to be able to factor that into our current European merger system?

Pierre Regibeau: For one thing, sustainability and pollution are not necessarily the same.

Let’s start with pollution because it’s a simpler issue. We have had cases, like the washing machine case, where the efficiency was that they would pollute the water less.

So let’s think about it. I’m a consumer. I am willing to pay for a good-quality washing machine. Fine. I am also willing to pay for a washing machine that consumes less electricity because that’s a lower cost to me. Beyond that direct utility to me I get the subjective utility that I feel good because I buy this machine which is good for the greater environment. This is giving me a warm feeling. That is still something that is linked to the fact that you buy the machine yourself; therefore, again, that is your willingness to pay. This is nothing new.

Now there is a more nuanced benefit—I’m not making this up; these are the kinds of things that are discussed extensively in environmental economics—that even if I don’t buy a machine or if I buy one, I have an additional utility from the fact that there are good non-polluting machines around. In environmental economics they make the distinction that I get a utility from going to the park in town and I get a utility from knowing that there is a park there even if I never go. Now that’s different, because I get that utility whether or not I purchase the good, so that the firm cannot capture.

This kind of utility we could actually count as an additional improvement in quality, which cannot be priced and cannot be recovered, and then yes I would be open to using some of the techniques that are used in environmental economics to essentially get these kind of benefits into the analysis.

Finally, there are the more direct externalities. There are the things that go to people who are completely outside of the market, people who will never buy a washing machine, and they will still benefit from cleaner water. Economically there is no big principle that tells us that you should not take those into account.

But realistically, if you don’t want to open a can of worms, and also if you want to convince those who read Article 101(3), as saying that those efficiencies have to be within the same market, then that externality I can only count within the same market. So I can count the fact that other people who buy the machine will also get cleaner water because I buy a machine myself. That we can take into account. That’s fine.

Going farther than that, I think, at least in the first step, is going to be hard for us. So that’s for pollution.

But if you look at sustainability, that is a bit different. The better example to give you of that is that of “sustainable development”: “Okay, we want to make it sustainable.” If people like sustainability, they will pay for it, so the same kind of allocation of benefit applies as in the pollution example. But producers are also saying, “If I am the only one to do it, it is not possible,” so we have to be able to agree not only to work on it together with others, but also to agree on things like, “Oh, we will all stop selling the non-sustainable product.”

How should we assess such a situation? Working on it together, that’s fine. We have Horizontal Guidelines, and the farther away the coordinated activity is from pricing, the more lenient we are. So if you want to work on it together—either because there are economies of scale, economies of scope, saving in the development level—that is fine if these benefits are demonstrated and cooperation is strictly limited to the required activities. We are also open to new arguments, and I think those might be reflected in the Horizontal Guidelines. For example, in the absence of synergies, we would expect producers to do it alone and we have their own label certifying that this is ‘sustainable,’ but if there are too many labels in the economy, nobody knows what they mean, so you would lose the significance of the label. In such cases we could be open to the quasi-standardization argument that you can coordinate with other producers to restrict the number of labels and agree and the product properties that they certify.

On the other hand, once you’ve got the labels, why would you really need to take total customers’ choice further and say, “We have to agree to stop offering the lower quality?” Why? There might be situations where you can argue that, but it is not totally obvious to me.

I mean you can have for every kind of issue these subtleties, and I think that at the CET we are quite open to broadening the principles a little bit according to what I have told you and then to consider every case on the merits.

Antitrust: Going to another topic that has made a lot of waves lately, there have been calls that mergers involving “Big Tech” should be reviewed under a different standard. In fact, in the United Kingdom, as you know, a more cautious standard would apply for assessing mergers involving firms with “Strategic Market Status.” There has also been discussion about the fact that DG COMP needs to be tougher or change its views on non-horizontal mergers (i.e. conglomerate mergers) when they involve Big Tech. What is your take on this?

Pierre Regibeau: I will limit myself to Big Tech. Actually I myself spent a little time looking at the acquisitions, identifying all the acquisitions by the GAFAM, and also I spent time going back through the Web to what the press said at the time, also what the industry said, to see whether the target was essentially referred to as a start-up or how big the target is, whether the target was a leader in its field, and whether some of its technology was actually used by the platform.

The picture that you get, with some differences across the platforms, is that yes most acquisitions are really start-ups. Some of the acquisitions are “killer,” in the sense that they disappear as such, but most acquisitions are not killer, significant parts of them—specialized people or technologies—and were folded into other products of the ecosystem of the platform.

We know that there are a few cases—and I could identify some myself—where probably there might be a bit of a killer acquisition flavor, and actually, frankly, if you read the press at the time, you would have gotten that at the time.

But now that we are a bit more aware of this and the consequences of some early acquisitions, actually I think that the ones that are fairly obvious killer acquisitions, even if they come early, we can probably handle, especially now that we have the possibility of reviewing mergers that fall even below the thresholds of the national authorities. So from that point of view I feel fairly comfortable.

Now the ecosystem thing—one has to be very careful because one has to realize that in those markets there are going to be different types of competition. There is going to be competition between ecosystems and competition between more single or fewer product type of firms, and it is our duty to make sure that both can coexist.

Actually, in a given case when you have an acquisition, I by now tell my teams to look at the two aspects: First, look at it in terms of system competition, then look at it also in terms of competition in the very specific markets in which the prey operates, and then try to see how those two fit.

Of these two complementary approaches, I still think the main look should be the market where the prey operates. If there is no concern, then you move to system competition. At this next level, I don’t want to go through the route of the efficiency offense. If you bring a complement—it’s a clear complement, it’s not a complement that will become substitute types of things—and it makes your platform more attractive, well so be it.

I only get worried if the target was a rare target—i.e. if acquiring this target would prevent others, including other rival ecosystems, especially a smaller ecosystem, from building their offering by also acquiring such a company. Why? Because in that case we know as economists that the more-dominant firm would have an incentive to overbid, to outbid the others, because part of what it needs to pay is for the maintenance of a monopoly. So that’s where I look.

For example—not that I think it was a huge issue there—if you look at Google acquiring Fitbit, one of the things I looked at is whether Fitbit is unique or are there a bunch of other companies in this sector that are independent, could be acquired, and have attractive technology. Of course, we also made sure there was an interoperability remedy.

So we do look at this. We do look at this, but we do have to keep applying competition law principles. For example, if you’ve got a company entering a new market, on which basis am I going to say, “Oh no, you can’t because you could be so good at it because of your ecosystem that you are going to be very successful.” I don’t want to put myself in a position where I start telling people they cannot enter a market with their already established firms just because they are going to be very good at it.

I think that if you are really worried about this kind of phenomenon, then you are going to have to deal with it, but not through competition law.

Antitrust: You have recently talked very openly about the need for competition enforcers to embrace algorithms and hire data scientists to work alongside economists and lawyers. How do you think that is going to improve the work that you could be doing at DG COMP? Are there any other areas of expertise where you think you would benefit from extra help as an economist?

Pierre Regibeau: There are essentially two levels of expertise for data scientists.

There are data scientists who know how to deal with data and apply algorithms and so on. We are starting to get people with this expertise at the CET, as we hire younger people and some team members also retrain. So that aspect I’m not too worried about.

I am really keen on having data scientists and IT people on the team who really know what they are talking about. As you know, having suffered through IT in all organizations, this is a rare asset.

One very important issue for which such expertise is useful is in remedy designs. People say, “Okay, we are going to have a data silo.” Fine, I understand, like you, at a vague level what a data silo is. But how secure is it? How hard is it to ensure that the data are only used for specific uses and so on? How easy is it to get around? And really, how does this work at the level of machine-language programming? This is something that you need to know if you are going to have serious enforcement.

Also, people who are able to, if not read algorithms, which is very hard, but at least play with algorithms to get an idea of what they can do, that’s useful.

This is especially true, now that you are going to have true artificial intelligence algorithms—what I mean by true AI algorithms are ones that essentially go and try things, not things that look for patterns—so really the open-ended kind of learning—then, by just looking at the algorithm itself, it is very hard to figure out what is going to come out of it.

I need people who can play with those enough to give me a direction of where this is going to go in the market, how it might interact with other algorithms, and so on. That’s the level of expertise that at least I would like to see within DG COMP.

Antitrust: Another complementary discipline that has been talked about a lot in recent years is behavioral economics. How does your team tackle the role of behavioral economics?

Pierre Regibeau: We don’t start by saying we are going to do behavioral economics or not. Again, you look for a theory of harm, and as you look for a theory of harm, you go through an economic argument: You say, “Oh yes, but consumers should be able to see through this and undo this devilish strategy.” But when you look at the facts, they don’t.

And then, you’ve got to step back and say, “Well, clearly we tend to assume that consumers are a bit smarter than they truly are.” Then the question is, “How dumb are they?” However you want to be careful because there could be a tendency to assume that they are as dumb as you need them to be for your theory of harm to work, which is clearly not what you want to do.

So you say, “Okay, so in this domain people are dumb.” Now is there some literature based on experiments and so on, or are there experiments that we can run ourselves, so that we could assess a bit more precisely to which degree they are dumb?

So I am quite open to doing this on the consumer behavior side, but then to do it seriously. On the other hand, I am not at all open to doing it on the corporate side because then it is too easy to say: “Oh, you know, we are a nice firm; we never raise prices”; this sort of behavioral “defense” would mean the end of enforcement.

Antitrust: Moving to another regulatory reform—there are a lot of them these days from the Commission as you know—is the new instrument that is being proposed to address the potential distortive effect of foreign subsidies in the single market, so to level the playing field. There are several instruments being proposed, but one of the instruments will be to require a formal notification of mergers where there has been a foreign subsidy involved. What is the case for subsidized acquisitions causing consumer harm? Some people have queried whether actually this new formal notification regime may actually create additional cost, raise legal uncertainty, and deter foreign investors from acquiring EU targets. How would you respond to that criticism?

Pierre Regibeau: My approach to this whole discussion has been to make sure that what we do makes sense and also is compatible with the World Trade Organization.

For imposing essentially State-aid rules on the operation of foreign firms within the single market I have no problem because it is just treating those firms the same way we treat European firms, but how it is going to be enforced is another issue.

The acquisition tool is more delicate. For one thing, what each Member State doesn’t like is acquisition by other States—having, let’s say, Chinese firms willing to overpay for their own companies, especially if they are willing to invest locally, is not something that governments complain especially about. So we must understand that enforcing this new tool might run into some opposition in practice.

What about the competition law rationale for the acquisition tool? When we talk to lawyers within DG COMP, they say, “It’s a distortion in the market for acquisitions.” As such, as a sentence, there is nothing I can object to, and I’ll leave it at that.

Now there is a little bit of deeper thinking that maybe we should do. Suppose that, rather than say, “Oh, you cannot have State aid,” you identify State aid and you ask, “But was it justified? Would this be market conform or did they overpay? Is there a legitimate return from this acquisition that gives the government that gave you this aid the normal market return?”

Suppose we did that. I would bet you that in most cases actually there would be a return. Why? Because what is the return for third country companies? Not only is it the money they make here, but the money they make by reimporting technologies into their home market.

Now, that should not in an ideal world enable them to outbid European bidders or other third country bidders, but those do not always have the same access to the third country home market. Therefore, the third country firm will have a higher valuation because of the lack of access to their home market and they would tend to outbid our companies as well.

If you really think very carefully about it then, that is a trade issue at heart.

Antitrust: Pierre, changing pace to wrap up our really interesting discussion—thank you again for your time—we sit here now nearly 18 months since the Covid-19 crisis started to take hold in Europe. It has clearly had a massive impact on everything businesses are doing in Europe. What role do you see for competition agencies in supporting the economic recovery?

Pierre Regibeau: Well, frankly I don’t see much of a role in terms of modifying the way we apply antitrust or mergers. I think that it is “steady as you go” there, taking into account the changing condition, if you take into account the failing-firm defense, you take into account the specific conditions of the market, and the changing expectations you are fine.

But there clearly is a big role for State aid. We have tried through the Temporary Framework to allow States to provide support, but to make sure that, except for direct loss support, this support would not necessarily go too much to firms that were already in trouble before. For firms already in trouble, there is the restructuring route. We have also tried to ensure that, if significant aid goes to large companies, there are at least some competition-preserving commitments.

We have to make sure that the same happens in the recovery framework. We must also try to ensure that recovery State aid makes sense. Should we keep supporting sectors that no longer have that much of a future? Should we now maybe have support that is a bit more conditional on investment, especially investment in innovation? This kind of stuff.

That’s why one of the things that we look at is not only how to modify the Temporary Framework. That’s something we are working very hard on. I think that is actually very, very important.

Antitrust: Thank you so much, and thank you again for your time. It has been a really interesting conversation and given us some fascinating insights into how the Chief Economist’s Team really works, and your thoughts on some of the current hot topics in antitrust.

Pierre Regibeau: You are most welcome. Have a good afternoon. Bye-bye.

    Authors