The scale of the problem that "big tech" creates is much smaller than the scale of the policy energy that goes towards the sector

Are Big Tech firms monopolies? My long-read Q&A with Nicolas Petit by James Pethokoukis of AEI.

"Big Tech firms like Amazon, Google, and Apple are some of the most popular companies among American consumers. However, among policymakers in Washington, they’ve become much less popular for many reasons. Chief among these reasons is the concern that these companies have become too powerful and need to be broken up, or otherwise heavily regulated. Recently, I spoke with Nicolas Petit about whether these concerns are well-founded, or whether Big Tech’s critics underestimate the consumer value these firms provide.

Nicolas is a professor of competition law at both the European University Institute and the College of Europe in Bruges. He is the author of the recently released book, Big Tech and the Digital Economy: The Moligopoly Scenario.

What follows is a lightly edited transcript of our conversation, including brief portions that were cut from the original podcast. You can download the episode here, and don’t forget to subscribe to my podcast on Apple Podcasts or Stitcher. Tell your friends, leave a review.

Pethokoukis: In the book, you focus on Apple, Microsoft, Amazon, Google, Facebook, and Netflix. Four of these companies are worth over a trillion dollars each, which is pretty big. Now, some people like to say that ‘every billionaire is a policy failure,’ and I get the sense they would say that every firm as big as these Big Tech companies is a policy failure as well. I assume you disagree with this?

Petit: Yeah, I do, and my book is basically trying to debunk this idea that there’s a problem with bigness in itself. And the distinctive feature of Big Tech is not that they are monopolies or gatekeepers or systemic firms. I think that the distinctive feature is that they’re highly flexible firms and they operate in these environments of deep uncertainty, where a lot of products and service recombination can arise due to the modular nature of data. And so I think what defines them first and foremost is that they’ve been able to overcome this deep uncertainty, but also, there’s much efficiency that goes with the size of these firms on the supply and the demand side, and therefore we should get accustomed to a world in which the future of the economy is in large-scale organizations.

I know, as a side note, that none of these companies seems to me to operate in a risk-safety-critical area, and none of them seems to me to be “too big to fail” like a bank. So frankly, I don’t really buy all that obsession with bigness. There might be problems, but they’re not in the bigness of these firms.

How did they get that big? I think some people think that while these companies used to be young and scrappy — that a long time ago, that sort of entrepreneurial innovative model that made them successful — somehow they’ve abandoned that model and now they’ve gotten big by buying up potential competitors when they’re small and by somehow influencing Washington now that their growth model is really crushing competition. That, rather than innovating and providing great services and great products to people.

Look, I think the reason behind bigness is really that they are providing a compelling value proposition to users. And the best way to think about that is probably to do a thought experiment like, “How would we live in a pandemic 20 years ago without Facebook, Google, and Netflix?” It would be unbearable. I mean, I myself live in Italy today and my family is in France and Belgium, and it would be impossible to communicate with my relatives in the pandemic. And the efficiency that we enjoy by virtue of these network effects is the source of tremendous service. The problem, of course, is that people like stories about things going wrong, and after the financial crisis we needed new culprits for the evils of the time. So I think a lot of energy has gone into looking at these firms as the source of a systemic problem. Again, I’m not saying that there are no problems with Big Tech, but I’m saying that the scale of the problem that they create is much smaller than the scale of the policy energy that goes towards the sector.

Are these companies monopolies?

The answer is “no,” and it’s on the cover of the book: I say they are “moligopolies” — a mashup of monopoly and oligopoly — because you cannot understand these firms just by looking at their share of outputs in narrow markets, like search for Google or e-commerce for Amazon. And so all of these firms to some extent compete as oligopolies in a wide area of market segments and at the same time enjoy a core position in an origins markets, but it would be, I think, foolish to think about the competitive pressure bearing on these firms by just looking at the segment in which they have a large market share.

If you think about Google, it of course is very dominant in search but at the same time competes with Apple for attention on mobile devices by placing its operating system in Androids, with Microsoft in software and productivity applications like docs and spreadsheets, and so on. And you cannot make sense of what Google is doing in search without making sense of what it’s doing in the other markets and how these dynamics work together. So I call them moligopolies. Some people call them ecosystems, and I think once you take this perspective of Big Tech firms’ size as ecosystems, you can see that there’s a lot of complexity that the monopoly moniker tends to ignore, and this is pretty bad.

 

Now, let me just give you one last point about that. The behavior of these companies is inconsistent with monopoly behavior. A monopolist is a lazy company, doesn’t invest in R&D, doesn’t invest in marketing, doesn’t invest in employment. What you see here for some of these firms is investments into demand expansion that are orders of magnitude higher than what you would meet in a standard monopoly situation. So there’s clearly no way just by looking at the evidence — but just also by thinking analytically — to categorize these companies as monopolies.

What do you mean by “investing in demand expansion”?

So, you know, they are constantly investing into complimentary products that increase value for users connected to the ecosystem. For instance, when Google adds emails, calendar functionality, navigation functionality to the search engine, every minute I spend on Google’s search provides increased utility to me being connected to the network because these complements work together to generate even more value. So it’s like Apple putting you into its ecosystem where the AirPods work seamlessly with the iPhone, which works seamlessly with the iMac, and so on. The difference between Apple and Google is that Apple does that in a way which is very closed, whereas Google does that in a way which is more modular and open and working maybe with less internal resources and relying more on external providers.

But I think what matters is that these ecosystems of complimentary products raise the utility to consumers. What I’m saying also is that the set of ecosystems that we will be using as consumers is not finite, and it might change. I mean, today we have one ecosystem for search, one ecosystem for handset devices, and one for e-commerce, but there might be new ecosystems of applications which will rise, as we saw with Zoom during the pandemic. So it’s a very dynamic environment.

That’s an interesting point, because the picture Big Tech critics give is that this is not a dynamic environment. You have these big companies. Google is dominant in search. Amazon is dominant in e-commerce. So there are these companies, they have this business that they dominate, and unless government acts, we are done. These are forever companies who now are so powerful that they cannot be challenged. No one would fund you if you tried to go create a company that would challenge Google in search (that’s an example that’s often given), and they’re all in their little silos that they dominate. And “ecosystem” is the wrong word, because that suggests it’s a living, breathing, organic, dynamic market, but really these are dead — or unchanging, stagnant — markets dominated by these big companies. That’s where we’re at, and there’s no sense that there’s a dynamic, evolving, churning aspect.

But you’re saying all of that’s not true?

Yeah, so I’m saying that it’s a matter of not looking into the right place. Anyone trained in a business school — and probably the people working in these companies and in competing companies — are trained to think that they should not go try to compete in a red ocean, and they should search for a blue ocean — meaning that you need to look for a product line where there are large margins to make, and that you’re not going to make your profits and find economic opportunity by trying to commoditize what exists. And so it might be true that there’s six or five platforms that we know they have already conquered a red ocean. But if that’s true, then you should watch elsewhere and see whether firms are trying to find a blue ocean upstream or downstream or in adjacent markets.

It’s like the drunk person looking for the keys under the lamppost: It’s where the light is, but it’s probably not where the keys are. And so if you turn your eyes away from the red oceans and you try to watch the blue oceans, you’re going to see a lot of activity. You’re going to see intense, competitive entry and attempts to penetrate this market. So let me just give you —

Yeah, so what are the blue oceans where these companies are attempting to sail?

Right, so driverless cars is a low-hanging fruit. They are all going there, right? And the equilibrium in this market is not yet there. I mean, we don’t know where the value will be in the software, in the navigation system, in the AI brain that will cover all that in the car itself. So there’s a lot of concerns here, and this drives hundreds of millions — or even tens of billions — of dollars of investment every year. Let’s think about B2B middleware markets — companies like Stripe, Salesforce, Zoom, Slack, Nextdoor, Twilio, Shopify — I mean, these companies did not exist 20 years ago. So actually, the point is: We are probably not looking at the right place. We should look elsewhere, and there, we would see a lot of activity, but also our time horizon is surprisingly short. I mean, we seem to expect more entry from digital than from any other industries and less stability. That’s very weird.

In your group of companies, some are obvious, like Apple and Amazon. But why is Netflix in that group and not, say, Tesla? Do you consider Tesla a technology company, or is it a software company trying to be a car company? Could you speak a little bit about why you picked certain companies and not others, or how you classify them?

Right, so let me be clear: There was no cherry-picking. I actually followed the press, and the press decided to use the acronym “FANG” (Facebook, Amazon, Netflix, and Google). In Europe, we have another acronym: We call them “GAFAM,” so that’s Google, Apple, Facebook, Amazon, and Microsoft. And so I decided to put the two acronyms together and use the names used by the press because I didn’t want to be accused of cherry-picking.

Now, you could probably reproduce some of the analysis in my book for other tech firms like IBM or Salesforce. As for Tesla, I don’t know. Maybe, or maybe not. You could say it’s an energy company to some extent. It’s also competing with a bunch of players in energy markets, so, I mean, I’m not sure, but there is a lot of sense in the idea that Tesla is also a tech company.

How concerned are you about these companies purchasing all of these smaller companies in order to prevent competition? The classic example I would give is Facebook buying Instagram, but more broadly, do you think that’s a problem — that they’re buying up too many potential competitors whose full potential we’ll never see?

That’s a great question, and it’s a subject of considerable importance, because if policy decides to put an end to that practice of intense M&A with small firms by Big Tech, then this will have a lot of counter-incentive effects on entrepreneurship and innovation. You have to know that most venture capitalists which fund these small startups really expect this M&A to take place. They don’t expect IPO as a dominant way for a startup to exit a market. A startup is there to die, to exit, and in two thirds or three fourths of the cases, VC funders really expect exits by acquisition. So you certainly do not want to kill the acquisition routes for the startups to exit. I think we have to be serious about the counterfactual worlds that we think about when we think about the ideal world and the real world, and the real world is one in which you see a lot of M&A and this is the normal way for startups to enter and scale and then exit.

Now, one thing I want to add to this is: People talk about Facebook and Instagram all the time as the canonical example of something that went wrong with policy, with the failure being that we didn’t prohibit Facebook from acquiring Instagram. Now, people tend to forget, when Facebook acquired Instagram, Facebook was a very young, publicly listed company. There was a lot of uncertainty around it, and what happened after? Facebook didn’t kill Instagram. It scaled Instagram to be really, really big. So regarding this idea of killer acquisitions where large Big Tech firms acquire small startups to kill them, Facebook-Instagram is clearly not the best example to use. The capabilities of Instagram have been grown and increased to a huge extent under Facebook ownership. So you might be unhappy today, but back in the day, this was not obvious. And clearly when we look at the history, we can see that we’ve not seen a killing. We’ve seen a rise.

It’s unclear to me if we had separate Facebook and Instagram companies — and we had Google North, Google East, Google West, Google South — it’s not clear to me how that would deal with the issues which are commonly brought up as the biggest problems with these Big Tech companies: issues about privacy, about what we might call “fake news,” and about hate speech. I don’t understand how antitrust is the solution to these problems. Antitrust seems to be just the most obvious thing to talk about because we have this history of antitrust in this country, and these are big companies, and we remember other big companies in the past like AT&T and Standard Oil. But I don’t really understand the relationship between the most common critiques of these companies and antitrust as the solution.

I agree. I think there’s two lines to that argument for breakups. One is, I think, emotional to some extent. So some people are angry with these big companies and they want to do something about them and a breakup is a vindictive, punitive way to show muscle and show who’s the boss here. It’s a sort of emotional way to act on a perceived frustration about policy and do something.

And then there’s, of course, the more analytical idea: When you break up companies in multiple segments or subsidiaries or business units, then you are going to create competition by rivalry, right? And so rivalry will lead to better outcomes, because competition is generally a good thing. This idea ignores two problems.

First, if you create rivalry, you’re going to raise outputs. So now, if you complain about the existence of fake news or not enough privacy or too much personalized advertisements, you don’t want to raise outputs, right? I mean, you have to be consistent. You don’t want to raise the output, the number of news, you don’t want to raise the number of ads, you don’t want to raise competition for personalization. So that’s one.

And the second thing is that economic theory shows very clearly that there are large network effects on the supply and the demand side. So you’re going to lose a lot of efficiency if you break these companies into multiple business units.

These Big Tech companies are American companies. Why aren’t half of the biggest tech companies European companies?

That’s a very hard question. I mean, if I knew that I would probably be a billionaire.

I think Europe has a bunch of structural, cultural, and economic problems that explain that. The cultural problem is that Europeans are generally more risk averse than American entrepreneurs. The structural problem is that we don’t really have a single market, contrary to what people often say. So very trivial things, like opening a bank account from one country to the other in Europe, are very difficult if you’re not a resident of the country, so think about a startup willing to expand the number of geographic markets in which it’s operating in Europe. This is already a hurdle.

So you have these problems and, of course, there are things like language — Europe has many, many languages, and this puts a break on labor mobility, which explains a lot of things, I believe.

Do you think any of it has to do with specific public policies, whether it’s taxes or regulation or anything like that? Here’s one reason I ask: I remember reading a Financial Times piece saying that Europe has decided that regulation of technology was going to be its comparative advantage. That doesn’t sound very entrepreneurial to me.

Yeah, that’s right. I mean, exactly. I think there’s often the belief in policy circles that you can create and promote innovation by legislation. And some people have referred to that in the academia as the “Brussels effect” regarding Brussels dictating norms and standards for the worldwide community, but there’s also this idea in the policy community that the Brussels effect is also trying to create innovation by legislation. I prefer to call that the “Brussels defect,” because I don’t think this is the way to create innovation. One deep, ingrained belief in Brussels is that competition creates innovation. And I think this has a lot of truth to it, but I think that it’s probably more true that innovation creates competition, much as Joseph Schumpeter said before. And I think Brussels is too much into the “competition creates innovation” and not enough into the “innovation creates competition” paradigm.

In the past, there have been technology companies that seemed to be dominant, like Nokia and MySpace. People thought they might be around forever, but then they were replaced. Do you think that the current tech titans are just so powerful that this process of churn has ended? Will today’s tech giants still be dominant 25 years from now?

Yeah, that’s a great point. It’s true that it’s been a while since we’ve seen a Nokia or a MySpace in the tech community, and it’s true that it has been a while as well that Google and Facebook and Amazon have been around. Now, what I said before is: Maybe we’re just not looking in the right place and there’s a lot of indirect entry.

The other thing is: The history of innovation suggests that new platforms or innovation often arrive all at once. So there are windows in history in which most of the innovative deliveries come around the same time space and we have to wait until the next window. That’s pretty well-documented, and it’s not a surprise.

Now, I would nonetheless submit that there is a lot of indirect entry. These firms are also extremely flexible. I mean, a company like Google could have experienced a near-death predicament when Apple created the smartphone industry with the iPhone. And they were very clever to have the capabilities to scale Android; they had acquired Android a little while ago, and they invested a lot into that to stay on top of the business. And so you cannot discount the dynamism that goes behind all these apparent monopoly positions. And so I hear people say, “Oh, Facebook has been a monopoly for 10 years.” I’m like, “That’s an interesting question, but the more interesting question is: Is Facebook today the same firm as it was 10 years ago?”

And there’s reason to think that while these companies may remain big and powerful, they will not be exactly the same companies as they were 10 years ago. It’s not a story of stagnation.

Exactly.

My guest today has been Nicolas Petit. Nicolas, thanks for coming on the podcast.

Great to be with you, Jim."