Skip to Main Navigation
Podcast October 27, 2021

Tell Me How: Big Tech Mergers - Part 2:  What Are Competition Authorities Looking At? 

View all episodes on our Tell Me How: The Infrastructure Podcast Series homepage

In this second episode on Big Tech mergers Prof. Katz and I begin our discussion on how competition authorities evaluate the benefits and costs of mergers and acquisitions, the types of indicators that matter and different views that policymakers have on the consequences of mergers.

This podcast series is produced by Fernando Di Laudo and Jonathan Davidar. 

Listen to this episode on your favorite platforms: Amazon MusicApple PodcastsGoogle PodcastsPodbean, and Spotify


Roumeen Islam: This is the World Bank’s Infrastructure Podcast. In the first of three episodes on big tech mergers, we discussed some defining characteristics of the market in which big tech companies operate and how mergers may affect competition in markets. In this second episode, we delve into policy designed to promote competition by supporting or limiting mergers between companies.

To give you an idea of how important mergers and acquisitions are in the industry, I searched for data and found some by the American Economic Liberties Project. This is what I found. Since 1998, Amazon has acquired 115 companies, about five a year. They were companies such as Leap Technologies, Health Navigator, and Art19. These varied in focus. A book reader, a digital health startup, and a podcasting platform don't seem to have much in common. Apple has bought about four a year since 1988. It bought Orion Network Systems, a networking software company, and it bought an asthma monitoring startup, Tueo Health. And Google bought Android - we're familiar with that, it's in many of our phones - and Play Space, a video screen sharing collaboration tool. It's obvious that big tech is buying companies for new business areas and also to make its current business operate better. Competition authorities assess how markets and eventually people and consumers, like us, are affected by these mergers and whether anything needs to be done about them. Let's find out how they do it.

Welcome back to our podcast, Michael.

Michael Katz: Thank you, it's a pleasure to be back.

Roumeen Islam: So, Michael, last time, I asked you about the pros and cons of mergers. And in your answer, you talked about the effects on competition and on consumer and worker welfare. Recently, many people, including the current chair of the United States Federal Trade Commission, have criticized use of a consumer welfare standard. Do you agree? And perhaps you could first explain what a consumer welfare standard is, and then let us know if you agree or not.

Michael Katz: Yeah, I'd be happy to do that. And it's absolutely important to talk about what we mean when we say ‘consumer welfare standard’ because I think we're in a situation, where different people in the public discourse mean different things by it.

Let me start by saying I support continued reliance on the consumer welfare standard. But properly understood. Now, some people, critics of the standard, say, is that the consumer welfare standard just looks at short term effects on prices. And it just asks are prices going down in the short run or up, say, as the result of the merger?

And I don't think that's the proper conception of the consumer welfare standard. In my view, and view of lots of economists, it also covers quality, innovation, and long-term effects. And in fact, I have a colleague, Richard Gilbert, and he and coauthors, and other people have documented that in the U.S. for example, the competition authorities pay considerable attention to the effects of innovation when they're evaluating mergers. And that's a long-term effect, it's not just about short-run prices.

Now, that said, although I am a proponent of the consumer welfare standard, it should be noted that, when we get into court, there's a tendency to focus on short-run price effects. I think the reason for that is those are the effects that it's probably easiest for us to predict when we're asking what's a merger going to do, what are the future effects going to be?

And so, there's a tendency to focus on that area. But I don't think that's inherent in the notion of a consumer welfare standard. I think a consumer welfare standard can be very forward-looking and it should be comprehensive.

So, I support that broad conception of a consumer welfare standard. What I don't support is going still broader, which is what some people would like to do and say - maybe broader may be the wrong term- but they want to do something different. They say: "Look, we're worried about size for its own sake. We just don't think firms should be allowed to be too big. We're worried about for political reasons or others. And also, that we want to prop up (they would put it differently, but I would say that's what they're saying) smaller competitors, even if that has to be done by stopping the bigger competitors from being more efficient.” And I think trying to broaden the scope of, especially, of a merger policy in that way and try to engage in these ways of shaping the market, I think, is a mistake. I think ultimately, they're going to be harmful for consumers and for the economy as a whole.

Roumeen Islam: I guess people want to focus on something concrete, like size, because it really is very difficult to measure or figure out what potential innovation is going to be in the future. So, if you're not looking at prices, then you might start thinking of some other theory that says [that] big firms, because they already have market power, get lazy and we should just look at size. What do you say to that?

Michael Katz: I think that you touch on one reason that people do it and that's an important reason, but there's also probably another reason. So one, as you say is, it can be very difficult, under current standards to challenge mergers and there are people who are concerned that it's too hard to challenge a merger and therefore turning to something like rules, bright-line rules based on size, could make it easier to challenge.

But then the other thing is I think there are people who believe that size itself gives rise to problems that we're just missing if we focus on what they think of as the consumer welfare standard. Did they think that ultimately, as they say, that size leads to inordinate political power and then distorts the political outcome?

Now, let me say one thing on that probably is different in the past, is that the role of a company in particular, Facebook plays in the U.S., but also in other countries too, in terms of affecting public opinion and political discourse. I think that is an issue. It's something that worries a lot of people, but my own view is that's not an antitrust problem. It's a public policy problem we have to address, but I think that is something that's gotten people to say, “We need to do more than what traditional antitrust looks at because traditional antitrust would not say ‘oh look, Facebook has too much power, it shapes public opinion too much’.” And I agree with people that's a problem, but I don't think it's an antitrust problem.

So, coming back to your question,

There's also one last thing, that the people we’re pushing away from the consumer welfare standard, again in a lot of cases, because they mean something very narrow by the consumer welfare standard. So, they say, “Look, it's important to prop up smaller competitors, because ultimately, they're the ones who are going to innovate, or, in the long run, they're going to keep prices down.” And I would say that's true, that's something we want to take into account, but we can’t take it into account under the consumer welfare standard.

Roumeen Islam: All right. So, let's move on to thinking about how we are going to assess whether a merger will harm or hurt competition and consumers, and you're trying to predict the future. So, now we're back to thinking about whether or not this is going to be an innovation enhancing merger or not. So, how do you do that?

Michael Katz: The short answer is you struggle because as you've identified, when you do merger analysis, it's largely a predictive exercise. You're asking yourself, “If the firms are allowed to merge, what will happen? And how does that compare with what would happen if they're not allowed to merge?”

So, you have to predict the future. And in fact, you'll hear economic experts when they testify at trials, talking about the, but-for world, now you have the actual world and the but-for world. Cause you're doing this comparison. You're saying what happens with them without the merger. What you try to do at a very high level, is you have to have a model of how the firms that are proposing to merge, compete with one another and compete with other firms.

And so, you develop a model, it might be a formal algebraic or numerical model, or it might just be a conceptual framework, but you try to identify and understand how they compete with one another and then how that's going to be affected when they merge and they no longer compete with one another, but will continue to compete with other firms.

And that's necessarily an emphasized process and one that has uncertainty, which is actually one of the reasons why in the U.S., the primary merger statute says that a merger is illegal, if the effect of the merger may be to substantially lessen competition or tend to create a monopoly.

And the key words there is that it may be substantially to lessen competition or that it may tend to create a monopoly. It doesn't say you have to have certain proof or that it's definitely going to happen. It's really about there being a substantial threat and this is the kind of thing that lawyers of course  spend a lot of time figuring out exactly what those words mean and how strong the threat has to be and how certain you are. But what's clear is we have to allow for some degree of uncertainty because it is a predictive exercise.

Roumeen Islam: Can mergers be undone according to the U.S. Competition Law? Can you go and say we made a mistake, I'm going to undo you now?

Michael Katz: Yes, in theory, they certainly can. This is one of the issues that's coming up in the United States now with Facebook, with Instagram and WhatsApp and this question of whether they can be unwound. The FTC, the U.S. Federal Trade Commission did in fact, consider unwinding a hospital merger several years afterwards, too. So, it certainly is a possibility.

Roumeen Islam: What do competition authorities need to show in order to stop mergers?

Michael Katz: Let me talk about the process in the U. S. cause that's the one with which I'm most familiar. So, ultimately if the parties want to go ahead with a merger, but antitrust authorities - let's talk about the federal agencies in the U.S. Department of Justice or the U.S. Federal Trade Commission - the antitrust enforcers want to block it. What they have to do is they have to convince a court that the merger would harm competition and consumers. And they have to do that by going through a several step process with the court.

Roumeen Islam: So, what's the first step here?

Michael Katz: So, the first step, if you're going to try to challenge a merger is you try to establish a prima facie case of harm.

And what that means is you go through some sort of basic steps and you hope that, if you're the plaintiff, that what comes out of that is you've convinced the court that yeah, it looks like there will be harm from the merger. And the way you typically would do that is you do, what's known as defining a relevant market.

So, what you're doing there is you're saying, what is it that the merged firms do? What market are they competing in? What products are they offering (these are the ones that we're concerned about in this case) we're asking what products may suffer from a loss of competition here if the firms merged?

So, that's the relevant market. And then once you've defined the boundaries of relevant market, you can calculate market shares and how concentrated the market is.

Roumeen Islam: Let me just stop you here Michael, because this is all very vague for someone who doesn’t actually know what that means. So, give me an example where what the relevant market is, is not clear.

Michael Katz: I would say actually, in a lot of cases, it's not clear because of the way we do the process. Let me give an example of some markets I've been involved in. So, take, especially at a merger of two companies that were debit card networks. Now you would say, okay, the debit card networks compete with each other to offer debit card network services.

I'll say just loosely, yeah, fine debit cards compete with each other. So, we should think of debit cards as being in the relevant market. But of course, there are lots of other ways to pay. So, you might use a general-purpose credit card, like one that's branded with MasterCard or Visa. Or you could use cash or if you're American you might use a check. I think a lot of people in other countries may not, as they're under 30 and they may not know what a check is, but we still use checks a lot in the United States.

Anyway, there are a lot of different ways to pay. So, then the question is, if I'm thinking about a debit network merger, should I also take into account competitive pressures from credit cards?

Now you might say, “Wait a minute, MasterCard and Visa are doing both, but you have companies like Discover or American Express that are on the credit card side.” So, you'd have to ask: "Okay, we have debit cards. Do we put credit cards in the market or not?" And, where things get to be a bit of a mess, is the way formal market definition works.

A product is either in the market or it's out. So, you have to ask yourself in this formal process the courts use: are credit cards in the market or not? And that turns out to be very contentious. And the same issue comes up with this cash in or not. And from the point of view of an economist and doing sound policy, that's unfortunate because I think that the facts are pretty clear, and they are the following:

Debit cards compete more strongly with debit cards than with anything else, but debit cards also compete to a degree with credit cards, but to a lesser degree. Now, in thinking about analyzing competition, you might say, okay, that's fine, so now let's look at the implications of those facts.

But when you do market definition, you don't get to stop there. You have to say, okay, debit cards are in the market, but [for] credit cards, you don't get to say, well, they are sort of in, and sort of out. You have to say they're either in or out. And it could well be the case that you say I think they're in. And then they say, “okay, then the merger is fine because there's so much competition from credit cards. These debit networks only have a small share of the market.”

But if you were to say, “No, no credit cards are not close enough substitutes, yeah, there's some competition, but not enough,” then you might conclude, “Wait a minute, if we look just at debit cards, these net debit networks are too big to allow them to merge, they have too high a market share.”

So, what all of this is saying is the following: that you're taking, what can be a complex and sort of nuanced world, and you're trying to fit it into this box. You say, look, here's the market, things are either in or out. How you make that decision can then affect whether you think the merging firms have high shares or low shares, because the broader the market, typically the lower the shares, the merging firms are going to look, and then that's going to affect whether or not you've made your prima facie case.

Now, if that all sounds vague and needlessly complicated, some of it may be because my answer may not be as articulate as it should be, but a large part of that is that, in my view - and the view of many people - the formal market definition process has really gone to be in many cases, critical, but also, it's more of an obstacle to good analysis than an aid. Because as I said, I think fundamentally the problem is that it doesn't recognize this nuanced element, that you can have things that compete to different degrees.

And so, to say everything's either zero or one, it's just distorting reality. To come back to your original question, the first step is I define relevant market, and it's often, as you say, a really critical part of the case. It's often described as the most important part of the case in terms of which side is going to win, even if it's not the most important part in terms of actually understanding competition and what all this means for consumers.

As I say, you define the relevant market. Once it's defined you calculate market shares and what's known as market concentration. And then if you're the government, you try to show that those market shares are high enough for the merging parties and that there's now a presumption (what's known as the structural presumption), a presumption that the merger is bad.

And if you can do that, you say, look, we've shown [that] the market shares are high enough that there's a presumption that this will harm competition and consumers. Then we say that the government has established its prime facie case of harm.

Roumeen Islam: Okay. So, all I conclude from this is that if there are several substitutes for whatever it is that's merging and the elasticity of substitution is large enough, you shouldn't worry about it.

And then if there are many other types of firms coming in that are going to offer even more types of debit functions or credit functions, then you have even less cause to worry. And that would be the innovation part. That's what I take away from this.

Michael Katz: So, what you have said, I think is absolutely right, in terms of thinking about how the merger is really going to affect things and trying to understand competition. Unfortunately, though, the legal process in the U.S. (and similar things happen in Europe) is this sort of obsession with defining the boundaries of the market. And as I say, that then can take on a life of its own.

Roumeen Islam: So, the main problem is that 100% in or out, because there is hardly any market for which there isn't a substitute in any good, basically.

Michael Katz: It's both. At some level, you could say everything competes with everything because people have a finite amount of money or amount of time and you laugh but there have been defense attorneys who have come close to wanting to claim that. Look, it would not be inconceivable to me to see someone who was defending a social network for saying, “Look, we compete against television. We compete against people playing tennis on a recreational basis.”

Those are all things people do to relax in their time and people have made arguments like that. So, it all then comes down to where do you draw the line? And as you said to an economist, you think about what the elasticities in demand look like or the cross elasticities. Often in cases you don't actually have good estimates of what the elasticities are. And also, there's not an agreement on what is the threshold elasticity we should use.

Roumeen Islam: Yes. I see, Michael. So, we're going to end today's episode here, but before we do, I just wanted to clarify for our listeners that, when we speak of a high elasticity of demand or cross elasticity of substitution between products, what we're really saying is that price increases for one product will make us easily switch away to another.

So basically, the products, though different, they compete with each other. And with that clarification, let me thank you for joining us today, Michael.

Michael Katz: Thank you. This really has been a pleasure.

Roumeen Islam: Well listeners, what did we learn today?

Firstly, The more substitutes there are, the lower the chance of a firm exercising monopoly power.

Secondly, Thirdly, and as a result of this uncertainty, technical assessments and quantitative data only go so far. Authorities need to exercise substantial judgment in evaluating each case. We will have more on this next week. Thank you. And bye for now.

You can find more information about the podcast on If you've got questions or comments, we'd love to hear from you. You can also find us on all popular podcasting platforms.

This episode was released in October 2021. Don't forget to subscribe and thanks for listening.