What Will It Take to Create Competitive Digital Markets?‌

Competition in digital markets is a topic that seems omnipresent right now. What are the elements that have emerged that make this into an issue that needs a lot of discussion?‌

The platforms are extremely important for growth and innovation in GDP, and for how we live our lives as a practical matter. Keeping up with friends and family, shopping, searching for information, maps, etc. I carry around an iPhone and it’s not very frequently I can put it down. At the same time, as the services are innovative and offer paths to growth for many businesses, those businesses are finding they can get to their consumers through only one or two gatekeeper platforms. These markets are extremely concentrated and those platforms extract much of the surplus business that end users would get if there were more competition. So the gatekeepers are important for competition, for innovation in the services we all use, and for the distribution of surplus between themselves and all other firms, among others.‌

In addition, the firms are not just powerful economically; they’re powerful politically. Part of the reason they’re powerful politically is because they’re so young that the founder is still in control, even if they’re publicly traded. Microsoft is now a mature company where Bill Gates no longer runs the company, but even though Meta and Amazon are publicly traded, they have founders who are in control. And that means that when they support a president or don’t, it has consequences for our democracy and our society, as well as our commerce.‌

Americans have the view that there’s supposed to be competition. Nonetheless, we have a bunch of monopolies in digital markets, and we haven’t tried to stop them from maintaining those monopolies until very recently. The Europeans began their first case against Google Search in 2010; the United States’ Department of Justice started its Google Search case in 2020, so we’re a full decade behind. ‌

The real origin of this book is 2019, when three reports were published on the problem of digital platform markets and competition. Harvard economics professor Jason Furman chaired one in the UK. In Europe, Margrethe Vestager asked three scholars, including the economist Jacques Crémer to chair another. In the U.S., the first Trump administration wasn’t interested in studying this problem, but the University of Chicago Stigler Center organized a report that I chaired and was the equivalent for the U.S. ‌

The process of writing that report forced me to figure out what was wrong with these markets, articulate it, and then understand how to translate that into the competition enforcement function—the antitrust approach. But it was clear to me by the end of the process that you weren’t going to get it all done with only antitrust enforcement. At the end of an antitrust case is a remedy, and the idea that one judge was going to find a small set of remedies for a giant digital platform and effectively enforce them for 10 years seemed ambitious. Such a task would be far easier with a specialized regulator and clever regulation. Almost at the same time, the first draft of the EU’s Digital Markets Act was leaked. So I started down the path of studying the regulation, and that turned out to be farsighted for someone based in the U.S. ‌

You mention in the book that the economic principles at work here are timeless. What are the economic principles that people should keep in mind as they think about this stuff?‌

I will mention my top four here. First, is network effects, of course. This term means that the more people that use one of these platforms, the more valuable it becomes to all of the other users as well as to other “sides” of the platform, such as merchants or app developers. Second would be economies of scale, the idea that larger companies benefit from lower unit costs. There is a particular form of economies of scale related to data, which is the third. An increase in user data often causes an improvement in the quality of the service, which in turn increases the number of people using the product, again increasing the amount and quality of data, which increases the number of people using it, so you have this virtuous circle of data. Those are probably the three biggest features. Then there’s the idea of an ecosystem business model, which is newer. The idea here is that the platform offers a package of complements, and those complements are more valuable together than any one of them apart. So if I have a mobile OS, I can stimulate demand for it by providing ( or getting others to provide) a handset, a map, a browser, and an app store and so forth.‌

It sounds like you think governments are behind on the regulation of digital markets. What are your takeaways for them from this book? ‌

Well, the Europeans have the DMA [Digital Markets Act]. And the DMA is, I think, an excellent law in principle. EU enforcement of the 22 rules of the DMA began in March of 2024. The Commission is, by necessity, going quite slowly because when you do something for the first time, every single step gets appealed to the courts. You don’t want to make any mistakes and you don’t want to do something too creative or without foundation, because you want the very first case the court sees to be correct. You want the court to say, “Yes, go ahead and enforce this law as you are doing” So enforcement in the EU is slower than I was hoping.‌

On top of that, Apple in particular is more or less refusing to comply with critical elements of the DMA. They’re not allowing rival app stores, for example. In some cases, Apple is doing almost the opposite of the law, which is remarkable because the EU is a large democracy, and just implicitly saying “my corporation won’t do this,” is not a sustainable position. Eventually the court cases catch up with you. So that’s a big issue. ‌

Enforcement is also a little bit slow because the platforms are required to comply and then write a report to the Commission about how they are complying, and then the Commission assesses the report and asks for changes, and that process repeats over time as both sides learn. But while it is happening, unfortunately, not very much of the discussion is public, so we don’t really get to see the interesting part of what’s going on. ‌

So the DMA is coming, but it’s slow. In the UK we have the DMCC [Digital Markets, Competition and Consumers Act]. This legislation was forecast to be in place before the DMA, but then the Conservative government was disorganized (all those prime ministers!) and didn’t get it passed. The DMCC only became legislation about a year ago, with designations in January of 2025 and release of the first rules slated to arrive around October. However, the Labour Party today seems to be worried about being nice to big tech and is making noises like they will discourage the CMA [Competition and Markets Authority] from enforcing the DMCC the way it was intended. So we won’t know until October what is happening in the UK. ‌

That brings us to the United States, where we are doing nothing by way of digital regulation. We have three Google cases where judges are choosing and imposing remedies. But recall that this is a huge job for one person, even if they create a “technical committee” to help them administer the remedy. A court is not a regulator with specialist skills and the primary purpose of overseeing corporations in one industry. They cannot keep tabs on technological change and consumer demand and update their regulations accordingly; nor do they have a mandate from Congress to achieve broad goals. Rather, they are constrained to whatever was decided in the antitrust case and a narrow solution to that.‌

‌What do you think would be an effective remedy in the case where Google was found to have maintained an illegal monopoly in search?‌

‌Last fall some colleagues and I wrote a paper for Yale’s Tobin Center proposing a set of remedies, including prohibiting Google from paying to be the default search for a handset or browser, forcing it to spin off the Chrome browser and the Android operating system, and mandating that it license its search index to other search engines. There are other remedies that could be helpful also, but the ones we put forward in the paper would prevent Google from blocking rival access to default status on a handset, browser, or other distribution channel.‌

‌In the book you mention the idea of a moat, which is something that a company loves to have, but that can eventually become harmful for the rest of us. What’s the point where a moat is a problem?‌

When the moat surrounds a firm that has monopoly power in an industry, that lessens competition in that industry and harms consumers.‌

Think of cars. We have lots of car companies. We have a lot of choice in cars. Those cars may be differentiated and therefore able to charge a small premium to those who like their features best, but the “moat” a car company can create around its brand is small. Think of how much choice we have in restaurants or shoes. Again, there are firms with strong market positions and markups that come from superior design or service, but there are limits to those moats because it is so easy to enter into the market for shoes or restaurants. By contrast, we do not have choice in operating systems for mobile handsets. There are only two. ‌

That leads to a lack of competition in mobile operating systems. A handset maker can only license Google Android if it wants to sell in the U.S. because iOS is vertically integrated with Apple’s hardware. So Google can impose conditions on the handset maker that lead to Google controlling the search revenue. Apple can impose conditions on cloud gamers and app stores that limit the functionality of the services on its handset. If those platforms faced more competition in mobile OSs, perhaps entrants would allow cloud gaming or be cheaper and consumers would choose them sometimes.‌

That market power spills out of the operating system market into adjacent markets. Let me give a concrete example. Suppose you were an entrant in luxury electric cars. You said, “Right, I’ve got this innovative expensive vehicle. Now I’m going to make an app so that owners can control the vehicle with their handset.” And Apple said to you, “We are going to charge $1,000 for each automobile that wants to connect to the app through the App Store.” What would you do? I think in the end you would choose to pay that $1,000 because you can’t sell a luxury electric car to wealthy consumers in the United States if it doesn’t have an app in the Apple App Store. Wealthy people disproportionately carry iPhones and they don’t want a car that doesn’t have an app. The result would be that you are stuck paying Apple whatever fee it demands.‌

This is an example of the economic power of Apple over much of GDP, because we access a giant share of GDP through our handsets. If the controller of the handset can control what you see, what you buy, where you buy it, how you buy it, and how much it costs, that makes all the other businesses in the economy quite anxious. ‌

And these forces underly the political economy that may give us laws that regulate digital platforms. Apple, of course, would lobby against being regulated, but there are a lot of other business users that are very keen on Apple being regulated so that they can get to their own customers without interference or undue expense. ‌

The goals of the DMA in particular are fairness and contestability. What do those two words mean to you as an economist?‌

Contestability is the easier concept. It means that entry is possible and entry barriers are low so that competition is the likely result. If we happen to see not very many new entering firms, then perhaps there isn’t demand for them, or they are not sufficiently innovative. However, in a contestable market it is entirely possible for a new firm to enter, connect, compete, and not be disadvantaged. ‌

Fairness for an economist is about ensuring that productive activity gets a competitive return. If I have a good idea and have innovated, invested, and executed well, then efficiency is enhanced if I earn the market return from that activity rather than having a platform take my profits. I should be able to capture the returns from my innovation and not face expropriation from a monopolist I cannot avoid.‌

Today I think the problem is that too many of those innovators must go through a platform and the platform takes a chunk of their returns without having done anything especially useful. If you have a monopoly position, developers don’t have the ability to go anywhere else to get customers and must pay you. By contrast, if the developer has a choice about how to get to customers then the platform must compete for their business with favorable terms.‌

It might be that a lot of the rents these platforms capture are coming from network effects, which don’t really belong to the platform; they just are a coordination device. If both sides left together and went to a different platform, they would be equally happy. The value added isn’t coming from the platform; it’s coming from everyone being in the same place. So that’s the way we think about fairness, is appropriate returns for innovative activity and investment.‌

Can you give us an example of a behavior for one of these gatekeeper companies under the DMA that would be considered unfair?‌

Let’s say Apple is not allowing other developers to have rival app stores on its handset. The result of that is there’s a one-size-fits-all app store. If you’re Spotify, you can’t sell your music subscription on Apple without paying them 30% or 15%. And since the music labels want half of Spotify’s revenue already, if Spotify gives 30% to Apple, Spotify cannot run a business. So they want to be able to tell users to link out and go to the Spotify website and buy there, or they want to have a rival app store that doesn’t cost 30%, and Apple doesn’t allow either of those things. Yet Apple did not invest and create Spotify.‌

That particular example is in the news right now in the U.S. What do you make of the Apple argument that this is payment for the work that they have done to create software infrastructure?‌

The answer to that is that Apple is absolutely welcome to monetize their ecosystem and their intellectual property. The sale of the handset is the obvious way in which they do that and anyone who has bought one knows there is also insurance, accessories, storage, and planned obsolescence. But when a gatekeeper can charge 30% in the app store to all sorts of different business it did not create and choose to give them free distribution, discount distribution, or sell access to their own customers for a high fee, then you are arbitrarily helping your friends and harming your enemies. When Apple cuts or restrict their ability to talk to customers, then Apple is holding back investment and growth in other parts of the economy. If Apple feels its tech is worth more, it can charge people more for the handset, or for the AirPods, or for the insurance, or for the repairs. Apple has many services they charge us all for, and those charges are harmless because they don’t have competitive implications.‌

It’s also the case that an app store is what we call middleware. And middleware represents a future threat to current platforms because it makes entry easier and more profitable. Middleware could become, or could sponsor, the entry of the next operating system and will, more generally, create more competition between operating systems. If there is middleware, a user who switches to a different handset and logs into her middleware application has all her content immediately; that user experiences low switching costs across operating systems. When users have low switching costs that puts a lot of pressure on Apple to compete and charge lower prices. So the competitive implication of blocking rival app stores is really substantial, and there are ways for Apple to collect money that don’t have those competitive implications. ‌

What would healthier competition in these digital markets look like for consumers?‌

The first critical element of healthy competition is lower entry barriers. When a new firm with a good idea can enter the market and compete, then consumers get the benefit of that new choice and existing competitors have to work harder to retain their customers—perhaps by cutting prices or innovating. Lowering entry barriers in these giant industries will lead to entry and more competition and we would see the market shares that are close to monopoly (e.g. Google search) fall. That process delivers more profit to consumers in the form of lower prices or better quality. It also rewards the businesses that transact through a platform. When those businesses have more choice, they will distribute through platforms with more favorable terms. That means developers and merchants will earn more profit and have more incentive to innovate. Handset prices will fall, new app stores that cater to children or to interest groups will be able to get distribution, newspapers will see a rise in their revenue from advertising, and so forth.‌