ANTITRUST MAGAZINE ONLINE: So sometimes not being involved in the day-to-day work on the case gives you a perspective. Is that kind of role similar across different types of matters, like mergers and trade practice matters and cartels, or even on the legislative side?
LILLA CSORGO: It shouldn’t come as a surprise that mergers make up a big part of what I and the economists generally do. That doesn’t mean that questions of trade practices, like monopolization, aren’t a huge part of the job as well, and that’s really where a lot of the gnarly economic issues come up.
There traditionally has been less of a role in cartels, but I have long been an advocate that economic analysis can be an important contributor to cartel matters in regard to circumstantial evidence. In my experience, there are guns, but they are often not smoking, so there is a role for economics to evaluate market outcomes, to help assess whether they are consistent with competition or not.
And then, you also just briefly mentioned legislation and regulations, and I’m absolutely involved in those as well. It’s always really exciting when amendments are on the table, but there’s also a lot of pressure to get it right because, as we all know, if you get it wrong, the ramifications of that are typically measured in decades.
ANTITRUST MAGAZINE ONLINE: Is there a role for economics in the decision of which remedy option to pursue, like whether to pursue something as a cartel or on the civil side, for example?
LILLA CSORGO: There is, I think, in the sense that you want to have effective deterrence. When I say “effective deterrence,” you must also take into consideration overdeterrence. So it may well be that a civil remedy is more appropriate in any one case. I think economics certainly has a role in trying to assess whether you want to go a civil or criminal route.
ANTITRUST MAGAZINE ONLINE: You talked about your role in the oversight perspective, but it may also be helpful to have economists involved in the investigative team on the day-to-day review. Do you have a perspective on that?
LILLA CSORGO: Yes. I don’t think many agencies have the luxury to have economists that carry out their own investigations and so bring a wholly independent view. The U.S. agencies are probably the only ones that are big enough to allow for that.
At other agencies you really want to have the economists integrated right at an early stage so that they can develop the appropriate theory of harm at that early stage, to inform what information should be requested, and then to analyze that information as it comes in through the door so that you can evolve the theory of harm and just more generally what your thoughts are on any particular investigation.
ANTITRUST MAGAZINE ONLINE: You’ve worked at different agencies, in some economies that are relatively larger or smaller than others. Have you perceived a different role for economists or for competition law enforcement more generally in the relatively larger economies as opposed to the relatively smaller jurisdictions?
LILLA CSORGO: The first thing I want to point out there is that there is a difference in what people will think of as small and large. I often hear that come up in the Canadian context, that Canada is a relatively smaller economy. From the perspective of having worked in truly smaller economies, Canada isn’t small. Canada falls into one of the ten largest economies in the world, while meanwhile places like Hong Kong and New Zealand are respectively about fortieth for Hong Kong and fiftieth for New Zealand.
It wasn’t until I was in New Zealand, where there is also the added element of geographic isolation, that I really, truly understood what it was to be a small, isolated economy and to have a visceral understanding of economies of scale.
It inevitably leads to more concentrated markets, and so there is that much more of a need to have strong monopolization laws. So in that sense it’s a real relief to see that in New Zealand they finally have changed their law restricting businesses with substantial market power from taking advantage of that power for an anticompetitive purpose . Hopefully, that legislation, which will come into effect in April, will really make a difference and make the monopolization law more effective.
I would also say that there is also that much more reason to have an effective market study law, one that allows you to compel information, because it’s not just the agency but the government itself that would like to understand whether the competitive outcomes that you observe are somewhat inevitable, that they are a feature of the economy, or if there is some independent impediment that is potentially outside of competition law or is not easily addressed by competition law, but is otherwise open to hopefully positive intervention by the government.
The only thing that I would note in addition is that, in the case of Hong Kong, they don’t have a merger law, and I don’t think that any economy is open enough to justify not having a merger law.
ANTITRUST MAGAZINE ONLINE: The converse of the recognition of more concentrated industries, like in New Zealand for example, seems to be whether efficiencies should be given special weight in those kinds of jurisdictions where efficient scale is going to inevitably lead to a larger share of the market. That consideration has led to the distinct approach in Canada. Do you have any observations on that?
LILLA CSORGO: Well, this is of course a big question in Canada, and it’s the one that’s up for grabs in the upcoming consultations with the government. I absolutely think that the merger efficiencies defense should be done away with.
First off, in the context of how it was originally justified on the basis of national champions, what the efficiencies defense allows you to get at is fixed cost savings. But fixed cost savings aren’t ones that are passed through by way of price, so I can’t say that I understand even the original justification that somehow this would have given Canadian firms a leg up in the international marketplace.
When it comes to the usual arguments that surplus is surplus and that it doesn’t matter what the source of that surplus is, whether it’s consumer surplus or producer surplus, we shouldn’t lose sight of the fact that that producer surplus is not costless—it is coming at the cost of allocative inefficiency. Meanwhile, there are costless ways of achieving production efficiencies. That is through competition, and so what we really should be doing is encouraging more competition rather than less, and we should be looking for other means of increasing that competition, including, say, in the case of Canada, doing away with inter-provincial trade barriers. Efficiencies shouldn’t be coming at the end of the day off the backs of consumers, where those consumers include business consumers.
ANTITRUST MAGAZINE ONLINE: Is there still some role for efficiencies in a merger analysis?
LILLA CSORGO: Yes, absolutely there is. As I said, fixed cost savings aren’t the ones that are going to be passed through in some way. Variable savings, on the other hand, are ones that can potentially benefit consumers. It could well also be that we have a merger that could bring about efficiencies that don’t actually correspond with an increase in market power, and those of course should be no-brainers; those should be mergers that we would want to encourage.
ANTITRUST MAGAZINE ONLINE: Switching gears a bit, to what extent have you observed economists at the different international agencies collaborating or sharing thoughts or theories on a particular merger or trade practice case?
LILLA CSORGO: I would say that this works best between Australia and New Zealand. In my experience, it’s really a fantastic, close relationship where you even have commissioners that are cross-appointed across the two agencies. That allows for really free, open, and candid discussion about matters that affect both jurisdictions.
I think that also helps with the relationship building more generally, so even if it’s not in a context of a particular investigation, because you have those relationships you can more easily just call each other up to discuss a particularly problematic issue that the other jurisdiction might have insights on.
In other jurisdictions the conversations—for reasons of confidentiality, so for good reason—can be a little bit more cryptic and consequently from my perspective just harder to follow and get the most out of.
But at no time are they useless. They always have some sort of benefit, but I do wonder, in the context of discussions of more general theories of economic harm, whether those conversations couldn’t be a little bit more candid.
ANTITRUST MAGAZINE ONLINE: To what extent have you observed that the agency economists are interacting directly with the parties’ economists, or are they typically more working internally as part of the team, and would you see benefits from more interaction with parties’ economists on matters?
LILLA CSORGO: They should absolutely be speaking with each other, and the more frank and candid and free-flowing those discussions are, the better.
On the parties’ side, I very much would like to see that the parties’ economists are not acting as advocates. And then, from the agency side, I would like to see that at that stage of the discussion it isn’t a question of “us and them,” that it’s still very much an investigation, which means that the agency is seeking to figure out the best way forward. It may well be that at the end of the day that the parties’ economists are right.
But I would also like to see that there is as much transparency as possible in the work that the economists do, that they are forthcoming about the assumptions that were made, the data that was used, the limitations of that data, because of course any modeling exercise is a simplification. That doesn’t mean that it’s not potentially quite useful, but to be able to properly ascertain its usefulness we need to understand the assumptions on which it is built.
There can be at times—and I would say that this is true for all parties once we get into litigation or a more contentious relationship—that there is this desire to somehow obfuscate, which I think is really problematic, so I would like to encourage transparency by all parties as early as possible and to allow for those frank discussions that hopefully at the end of the day allows us to better arrive at the right outcome for consumers.
ANTITRUST MAGAZINE ONLINE: Oftentimes agency economists have more information than the merging parties, to the extent that they can subpoena third parties and collect information from other interested parties. But they also cannot share much of that information. Do you have any observations on that?
LILLA CSORGO: Well, unfortunately, that asymmetry is inevitable in those early stages. Of course, we do want to protect people’s confidential information, so when I say “unfortunately” it’s unfortunately in the sense that it doesn’t serve this transparency that I’ve been otherwise advocating.
ANTITRUST MAGAZINE ONLINE: Maybe we could touch on a couple of what might be called emerging issues and get your thoughts. For example, there is clearly an increasing focus in the labor market on wage-fixing and no-poach agreements in many jurisdictions. Do you see any distinctive issues with regard to assessing buyer power and the use of competition law to address labor market issues?
LILLA CSORGO: I have to say that it’s so much easier if it’s per se illegal because of course then you don’t have to worry about the output effects. In the recent amendments in Canada, where the labor agreements are per se, then the economics is relatively straightforward.
Otherwise, the area of buyer power is one that is quite fraught. Starting with terminology, there is just a lot of confusion as to what is meant by buyer power.
Does it include countervailing power so that the price is above marginal cost and if you have a depression in the price that actually results in an increase in output, which traditionally has been treated as a good thing; or is it restricted to monopsony power so that prices are at marginal cost and there is a risk of them going below cost, so a depression in price results in a decrease in output; or is it an umbrella term that covers both of those effects and you need to probe further as to what is meant in any particular case when it is used?
So if we are, say, again in the merger context and there is this question of buyer power, I think it is facile—and this is certainly what has been done historically—to say that it is just a mirror image of what happens on the seller side.
First off, on the seller side you are typically not faced with the problem of trying to determine whether prices are above marginal cost. You can go in with quite a degree of confidence that they are, or otherwise it’s not a merger that the competition authority would be looking at in the first place. So we have this question as to how prices relate to cost on the input side.
There are also often issues related to the elasticity of supply, and this is particularly a concern in Canada because of the preoccupation with deadweight loss as a result of the efficiencies defense. So you could well be in a situation where the input in question is a byproduct, a waste product, of some upstream firm’s production, and it could be that prices are depressed so much so that they are no longer being paid for the input but rather they have to pay to have the input taken away. That raises all sorts of questions about the related deadweight loss.
So, as I said, it’s actually quite a challenging area and I think the agencies have not served the competition community well by saying simply that it’s a mirror of what happens on the seller side.
ANTITRUST MAGAZINE ONLINE: Could some of those issues get buried in a per se prohibition on buy-side agreements?
LILLA CSORGO: They could be, although I think there are other reasons as to why you would want to have a per se provision in regard to labor markets.
ANTITRUST MAGAZINE ONLINE: Right, although the Canadian proposals now may extend beyond labor for per se buy-side prohibitions.
LILLA CSORGO: Yes. And it’s not something that I think should be done without a lot of thought as to how you would want to capture problematic buyer-side issues. It’s a difficult area.
ANTITRUST MAGAZINE ONLINE: The Canadian courts have tended to require the Competition Bureau to quantify alleged anticompetitive effects to the extent possible. On the other hand, amendments to the Canadian Competition Act last year specifically provide that the Competition Tribunal may have regard to impacts of mergers or competitor collaborations on nonprice competition elements, such as quality, choice, or consumer privacy. Do you have any thoughts on how adjudicators might best approach anticompetitive effects that may be challenging to quantify?
LILLA CSORGO: Before I answer that part of the question, I want to note that I think that it was a misstep in the Canadian jurisprudence that there has been what I would say is an undue emphasis on quantification. I would go so far as to say that there is this naiveness about the extent to which that is somehow divorced from qualitative assessment. Quantitative assessment is very much informed by other facts, or should be informed by other facts, and those facts are often ones that are arrived at through some sort of qualitative assessment. They can’t be divorced. Good analysis really relies on all possible sources of information, be that the views of market participants, industry experts, documents, and so forth.
Having said that, I think there is a danger of the jurisprudence further contaminating the qualitative aspects that you mentioned as now being formally part of what should be assessed under the law. At the end of the day, you can always put some sort of number on any aspect of competition, including quality or variety or privacy. I mean, how much do people value privacy—that is something that, in theory at least, can have a number applied to it—and if it does, it should be recognized that that number is going to be informed by non-quantitative pieces of information.
ANTITRUST MAGAZINE ONLINE: Another hot topic, if one could call it that, is more than one jurisdiction expanding or proposing to expand the scope of conduct covered by abuse-of-dominance or monopolization provisions and increasing the fines and penalties for such a competition law contravention.
Do you have any thoughts on that development generally or, in particular, one point that you mentioned earlier about the risk of overdeterrence or what may be described as Type 2 errors in prohibiting potentially efficient conduct?
LILLA CSORGO: I’m only going to comment on the question of fines. The truth is that at the end of the day there are so few abuse cases that even if there was an increase in fines, it would be very unlikely that that would have an overdeterrent effect.
We need to consider the probability of being actually caught; and once that is taken into consideration, then those expected fines are really quite low. We don’t want to be in a situation where it is rational, profit-maximizing behavior for a firm to engage in that conduct because it is willing to take the risk, even if fines are high, of getting caught.
ANTITRUST MAGAZINE ONLINE: Turning to another current issue, while you were at the Canadian Competition Bureau last year, the Bureau hosted a Competition and Green Growth Summit. One issue emerging from that Summit was potentially a tension between environmental policies, or even competitor arrangements that may be part of an environmental policy, where the objective is to seek to reduce consumption of certain products, and a competition policy that generally aims to restrict restraints on output. Do you see tension there, or how might that be looked at from an economic perspective?
LILLA CSORGO: There is absolutely the potential for that tension. The competition laws of New Zealand and Australia are well placed to potentially address that tension because there are authorization processes in those countries that allow for competitor collaborations that are to the net public benefit.
We are, I think, going to see more of that kind of potential collaboration across firms, and we want to have a competition law that is well positioned to deal with one-off, short-lived relationships. If they are not one-off, relatively shorter relationships, then generally—maybe not always, but generally—it is better to have a regulation that more explicitly deals with that and can address the matter in a longer-term way.
Just as a simple example, there are so many jurisdictions that are doing away with single-use plastics, and that can in some way be thought of as a reduction in competition—some consumers might prefer one retailer over another because they don’t have to show up with their own containers—but, rather than going through some sort of authorization process that deals with some subset of retailers that have entered into an agreement, it is more efficient to have a regulation that deals with it wholesale.
ANTITRUST MAGAZINE ONLINE: Do you have any thoughts on suggestions in some jurisdictions that competition law should be part of the response in addressing perceived high inflation in food prices or other prices and assisting in economic recovery in the post-pandemic era?
LILLA CSORGO: I don’t think it has a formal role in that, but I think that attention should be paid to complaints that firms may be using this as an opportunity to tacitly coordinate and to pass through more of those costs than would otherwise occur in a competitive environment. Of course, once you have successfully managed to tacitly coordinate on inflation-related cost increases, then it’s just so much easier to continue in that relationship post an inflation environment.
It may be questioned whether an agency has sufficient information to justify a cartel investigation, or whether it’s a situation where you want to make use of market study powers so that the agency can determine whether there has been some sort of shift, some sort of event, that is inexplicable absent that kind of coordination. Of course, that would require the agency to be able to compel information, and so you would require a law that would allow for that.
ANTITRUST MAGAZINE ONLINE: Are there any other emerging or hot economic issues that you’d like to flag for our readers that they should be focusing on in the years to come?
LILLA CSORGO: There has been increasingly this renewed emphasis and discussion of competition and the competitive process, particularly coming out of the United States, and I personally am very happy to see that kind of renewed emphasis.
But I do want to make sure that simply mentioning that is not where the conversation ends, that it has related questions as to: how do we, for example, consider efficiencies in that type of setting; and how do we assess whether a reduction in competitiveness is too much of a reduction? Historically, rightly or wrongly, we have had a focus on price, and price is a relatively accessible measure, and if we are moving away from that, then what are we—in practice, not just in words—moving to?
ANTITRUST MAGAZINE ONLINE: In addition to serving as head of economics at some enforcement agencies, you have also been an adjudicator, at least in Canada and now in New Zealand as well. Do you have any tips for how private parties and their economists can best present economic evidence to the agencies and the adjudicators?
LILLA CSORGO: I’m going to come back to my theme of transparency and making it absolutely clear what your model is assuming and to be up-front as to what the possible limitations are of that, and to explain why the model is nonetheless appropriate in the circumstances of that particular case; and if there are limitations, to indicate what those limitations are.
The other thing that I would encourage more generally is that, in the context of merger simulations, there really is not a lot of emphasis on checking the robustness of the simulation results. There should be. It should play more of a role.
If we have an econometric model, there are so many tests that are already built into that methodology to see the extent to which the results are actually significant. There are techniques that have been developed over time that are part and parcel of the econometricians’ toolkit for carrying out robustness checks. We don’t see that on the merger simulation side nearly to the same extent and certainly not in the same way. It’s not as if those tools are absent. They should be more regularly used.
ANTITRUST MAGAZINE ONLINE: This may be too technical a question, but what types of tools are available to check the robustness of merger simulations? Is it correct that every merger simulation predicts some price increase?
LILLA CSORGO: Yes, that’s certainly true. By way of example—and it probably is the main example—typically a model is calibrated on a limited amount of data. So how much does the calibration change if you shift that data in a small way? Also, it is predicting a price effect based on that calibration, and so if we have historical information, historical data, historical prices, that are also meaningful in the context of the merger, does that calibration reflect that historical situation? Those are two ways of potentially testing for the robustness.
There is also the scope for ex-post analysis, whether that’s by a competition authority or by the academics who potentially have access to relevant information. There is benefit in looking at the extent to which a merger simulation accurately predicted what actually happened.
ANTITRUST MAGAZINE ONLINE: There is an impression that agencies tend not to do enough of those ex-post or retrospective merger analyses. But it seems so helpful to ask the question—“We approved this merger some number of years ago. Let’s find out what has happened.” Have you seen much advocacy within agencies to do that kind of analysis? Certainly, resource and budget constraints are potentially a concern within any agency.
LILLA CSORGO: Yes, certainly ex-post analysis is something that has been on the radar at competition agencies for my entire career, so that’s thirty years now. The fact that you don’t hear more about them is not due to a lack of willingness. There is a willingness, and certainly attempts have been made—sometimes quite successfully, other times less so. A lot of it comes down to data limitations.
Market studies are not normally discussed in this context, but if there is an ability to compel information ex post the merger, that puts you in a so much better situation to actually be able to assess what did in fact happen to prices.
In New Zealand, when I was there, we did try to do ex-post analyses, and our focus ended up not being so much whether we got it right or wrong, but whether there were particular aspects of the investigation that we neglected, failed to take into account, or overemphasized. The easy one, for example, is if the agency was relying on entry and entry didn’t take place, that’s something that’s relatively easily assessed. But it may also be helpful to consider other factors, for example, if the agency was relying on imports and there has been fluctuation in those imports—it is useful to understand the source of that fluctuation so that you can better assess the next merger.
ANTITRUST MAGAZINE ONLINE: As you said earlier, economists and economics also have an important role to play in conduct cases. One area economists can certainly contribute to is screening and detecting certain acts such as collusion. Do you have examples where agencies proactively try to screen and detect cartels?
LILLA CSORGO: Yes. I would divide it into three: (1) agencies that are lagging behind on that; (2) agencies that are relying on widely publicly available information and doing very high-level assessments of that data to try to see if there is something amiss; and (3)—and I think this is really the most effective—is when agencies are working in conjunction with other parties, such as other government agencies that have access to good procurement data, and analyzing that data in a very close way. That final one is the one that has the greatest potential, but it means that these government departments need to work together. I find it amazing the extent to which that can be problematic.
ANTITRUST MAGAZINE ONLINE: Do you have any other things you’d like to add or topics we may not have asked you about that you’d like to address?
LILLA CSORGO: There are two things.
One of them goes back to me hammering on about transparency. One thing that New Zealand does particularly well that agencies around the world, to the extent that they don’t already do it, could really benefit from is that in the merger context, but sometimes also outside the merger context, they produce fulsome investigation reports that are available to the public.
The benefits of that are manifold. Not only does it make the thought process of the agency transparent to everyone, but the fact that it’s subject to that kind of public scrutiny, and the mere fact of writing something down, really forces an agency to be truly coherent, cogent and logical in its assessment. There is so much benefit in that. I think it really speaks to an agency’s accountability that those kind of reports are made available to the public.
Second, I want to mention something about supplementary information requests (SIRs), which are the Canadian equivalent of second requests under the U.S. Hart-Scott-Rodino Act premerger notification process.
We talked about the emphasis on quantification in the Canadian context. Compared to when I was last at the Canadian Competition Bureau, one of the big things that has changed is the introduction of the SIRs.
I find that process—I’m trying to land on exactly the right word without getting everyone’s back up—I think that there is real benefit in a process that allows for greater iteration. But the problem I would say with the SIR process—and I don’t know the extent to which this extends to second requests in the United States because I don’t have any direct experience of that—but the issuance of a SIR happens at a stage where the agency knows enough to know that there is a problem, but it doesn’t know enough to be able to know exactly where and how those problems might arise, and so there is this tendency, that is perfectly understandable, to be very expansive.
Two problems can arise. One is that you end up getting all this information—and of course gathering information is costly, receiving it is costly, and just housing it is costly, and there is a pressure to do something with that information, which is also potentially costly—but it may be information that you don’t actually need.
Part and parcel of that situation is, because you have asked for the information at such an early stage, it may well be, despite this incentive to be expansive, that you still end up missing the mark. That you find out upon further investigation that the real problem lies elsewhere. But by then the process in Canada, although it doesn’t preclude accessing further information, doesn’t tend to lend itself to getting that information.
Even if you more or less have the right information, it’s not often that you wouldn’t want to have some sort of clarification and follow-up. The process of the SIR doesn’t tend to allow for that. So I am in favor of something that is more iterative.
ANTITRUST MAGAZINE ONLINE: We want to thank you, Dr. Csorgo, for taking the time and providing such a thoughtful and comprehensive perspective on a wide range of topics.
