Marshall Steinbaum on the path forward in antitrust

Steinbaum is a consulting expert on the antitrust case against Uber and Lyft

It’s been a big week for antitrust in Oversharing! We’ve talked about the antitrust case against Uber and Lyft and the antitrust case against food delivery platforms. To round it all off, I’m delighted to have another conversation with Marshall Steinbaum, assistant professor of economics at the University of Utah and a consulting expert on the ride-hail antitrust litigation. We talked about the two lawsuits, whether anticompetitive practices like those alleged in the complaints might also be contributing to inflation, and why we may be in a “recoupment phase” of predatory pricing.

Oversharing interviews with experts like Steinbaum are available exclusively to paid subscribers. As a reminder, I offer a discounted group rate at this link for companies and organizations interested in purchasing four or more subscriptions.

This interview has been condensed and edited for clarity.

Oversharing: Hey Marshall, good to have you back! Last time we spoke in April, we discussed your great piece on the antitrust case against the gig economy. Since then, the antitrust case against Uber and Lyft has actually been filed, on which you are a consulting expert. I wrote about that case for subscribers a few weeks ago, but can you briefly take us through it?

Steinbaum: So this case is brought by rideshare drivers in California on behalf of a class of rideshare drivers, who have been classified as independent contractors, at least putatively under Prop 22. Their allegation is basically that the control the rideshare platforms exercise over the way that rideshare drivers do their jobs violates California State antitrust and California State consumer protection laws.

There are four broad headings in which the conduct is alleged to violate either or both of those laws. One is the price-fixing that the platforms do. It's the rideshare platforms and not the drivers themselves who decide what price consumers pay for each ride. The competitive effect of that is to prevent drivers from discounting rides on lower-take rate platforms. If a platform paid drivers more of a given gross fare, the drivers would want more customers to patronize that platform and not the platforms that pay drivers less. And the way the drivers would get customers to patronize the lower-take rate platform is by discounting on that platform.

So the price-fixing argument is that when Uber and Lyft set the prices, they prevent drivers from being able to engage in competitive pricing behavior that would also be good for consumers?

That's right. And it eliminates competition between the two rideshare platforms over their take rates. If this price-fixing weren't in effect, the platforms would compete with one another by reducing their take rate. The effect of the price-fixing is basically to effectuate a cartel in the rideshare industry. It's like you've got two players: Do those two players compete against each other very hard, or do they have a very comfortable life agreeing to their high take rates where they both earn a lot of money at the expense of both sets of counterparties? And we're in the latter world, thanks to the price-fixing.

So that’s price-fixing. What are the other three broad headings in the case?

Non-price vertical restraints. So that's like the bonus systems that require drivers to de facto single-home in order to achieve a certain bonus, and then you can't make a living as a rideshare driver without getting the bonus. So even though the platforms might say the drivers are multi-homing, and even though in principle the drivers might be signed up on both apps, in practice, in any one shift, many of them have to be de facto single-homing to make the shift worthwhile. So that's item number two.

Items three and four kind of go together, which is the fact that they don't share complete ride data with drivers in advance of accepting or rejecting a ride. So there the drivers are basically guessing about how much money they'll make off of the ride. If they accept and then discover information that indicates the drive will not be worthwhile to undertake, if they cancel then, then they face severe penalties from the platforms. So they're basically locked in on the basis of incomplete information when the platforms themselves have complete information. And then the other extension of that is another non-price vertical restraint, which is the tiering of data access conditional on minimum acceptance rate. So it's like, you can see some data that's relevant to making the decision, but you have to accept a minimum share of rides and then that defeats the whole purpose.

Do you see antitrust as a new path forward for gig rights, after a decade of failed litigation on the question of employment classification?

I definitely think there's a path forward in antitrust. You said in one of your publications that this takes as given that the drivers are independent contractors, and I don't agree with that statement. My view is that the economy is regulated to some degree that allocates the power to make economic decisions to different agents in the economy. And if the drivers were employees, then the conduct that is being alleged as anticompetitive would be legal. Because employers can tell their employees which specific work they have to do; employers can decide how much to charge individual customers on behalf of their employees. So all of those things are baked into the employment relationship.

My view of where the policy landscape might go as a result of this opening up of the box of antitrust, if that's indeed what happens, is it sort of shifts the calculus on the part of the gig companies because they're going to want to be employers. If their business depends on being able to control drivers in all of these ways, and they face antitrust liability for doing so if they treat their drivers as independent contractors, then they're going to want to treat their drivers as employees. So that shifts the ground pretty significantly from where we've been for the last decade.

There was also an antitrust suit filed against Grubhub, Uber, and Postmates (the last of which is now owned by Uber) in July 2020 in U.S. District Court. The complaint alleges that high commissions and restrictive terms imposed on restaurants by food delivery platforms ultimately lead to higher prices for all diners, even if you don’t use the apps. It survived a motion to dismiss back in March. How does that case compare to the ride-hail one?

The food delivery case is being brought by plaintiffs saying that they've been harmed because thanks to the food delivery platforms’ conduct, prices for restaurant delivery meals and in-person dining as well—an interesting argument that they make—are higher than they would be if the platforms didn't engage in their anticompetitive conduct. So that is a traditional antitrust case in that the harmed party is very clearly consumers. In the rideshare case, the plaintiffs are drivers, but the theory of harm is very much a theory that also has consumers being harmed—that the price-fixing prevents discounting by which both consumers and drivers would benefit if they were able to have autonomy over prices on rideshare platforms.

The key term in the food delivery case is “Most-Favored-Nations” clause, or MFN. Can you define that for folks?

Yeah, so Most-Favored-Nations clause, that kind of odd phrase comes from actual international trade agreements. When two countries have a trade agreement, a provision of that agreement can say you have to charge no higher tariffs on us than you charge on your most-favored other trading partners. So in other words, we can't be disadvantaged relative to some other country in trading with you. That has been sort of ported over to the relations between companies, as in this food delivery case. So not between nations but rather between platforms and their suppliers. And what it means there is that the suppliers—restaurants in this case—must charge the same prices for restaurant meals for individual items on their menu to consumers who purchase through the food delivery app as to consumers who either order for delivery directly from the restaurant or, in some cases, dine-in as well.

Another interesting aspect of the case is how it identifies three different markets: restaurant platforms (takeout and delivery orders that happen through a food delivery platform); direct takeout and delivery (orders placed directly with a restaurant); and dine-in (orders placed and consumed in a dining establishment).

The whole question of how to define markets and establish market power in antitrust platform cases is very much a matter of ongoing academic and legal debate. I have a paper called Establishing Market and Monopoly Power in Tech Platform Antitrust Cases that's exactly about this. So basically, everyone is trying different approaches to this and there's no one accepted way of doing it so far. So I was interested to see this case do it in a fairly unique way that I thought was compelling and seemingly successful, at least as far as they’ve gotten past motion to dismiss.

The central allegation is that when food delivery platforms restrict restaurants from charging prices lower elsewhere than what's being offered through the apps, the result is that your meal in the actual restaurant might also be more expensive. What do you make of that?

I want to highlight that point because I think it has broader policy significance beyond this antitrust case or really antitrust as a whole. I'm getting a little ahead of my skis here as an economist, but I suspect that the existence of platform MFNs is significantly responsible for the current inflation that is going on, and it's through the mechanism of increased fees and take rates on all sorts of platforms, all of which impose some sort of MFN, which then means that prices go up across the board, because they can't go up specifically for the people who patronize that platform. So yes, as you described it, the allegation in this case is that even if you have nothing to do with Grubhub, you could go to a restaurant, sit down, and pay a higher price as a result of Grubhub’s Most-Favored-Nations clause. Because when Grubhub increases the fees that restaurants have to pay, restaurants can't just increase the price of ordering through Grubhub, they have to increase all the prices if they want to stay in business.

That's fascinating. So when you're theorizing about broader implications around inflation, we’re obviously talking about more than the gig economy here. What would some other examples of that be?

Well, I mean, we use platforms in every aspect of our lives. So credit cards count there, increased advertising loads on Google search. I don’t know that this is true, but suppose Google is increasing its advertising rates and then advertisers have to charge more for all of their goods, not just on Google.

Super interesting. The last thing I want to touch on is the role of the consumer welfare standard, which basically says that monopolies aren’t inherently bad, they’re only bad if they lead to higher costs for consumers. How are we seeing that play out in the ride-hail and food delivery cases?

Antitrust laws are vague in the United States. They prohibit harm to competition. The whole legal wrangling since they were passed about what exactly they prohibit is about, well, how do we know that competition has been harmed? We've had debates and different paradigms have prevailed at different times and even simultaneously. Cases can be decided that have implicitly different definitions of harm to competition as being the deciding factor in the case. The consumer welfare standard is a contention about what should count as harm to competition within the meaning of the antitrust laws.

The big case that people who like the consumer welfare standard will point to is: Suppose you have one dominant company that comes to have a monopoly in a given market because it has the most efficient cost of production and can lower its price because its costs are so much lower, and that puts all of its putative competitors out of business and then it becomes a monopolist. That company becoming a monopolist is increasing consumer welfare. So it hasn't harmed competition, even if the number of competitors went from 10 to one as a result of that company coming into the market.

Unless it were to then raise prices, right?

Yes. This is where we get into predatory pricing and whether that's what's going on. I think the other the other piece of the inflation story is basically recoupment from past predatory pricing. So with predatory pricing, it's suppose that all the competitors have the same cost of production, but one charges a lower price in order to drive everyone else out of business and claim the entire market. Then it becomes a monopoly. Then it can charge a higher price and make back all the losses it made when it was competing vociferously against the competition.

It's very hard to prove predatory pricing because low price is supposed to be the sign of competition. So you have to show not just low prices—lower prices than the cost of production—you also have to show eventually higher prices or a strong probability of higher prices after the predatory scheme has been effectuated. And that's very hard to predict. For example, there were predatory pricing cases against the rideshare platforms early in their life cycles, like in the early 2010s. They all died because the court said there is no definitive probability that after this low price goes out of existence, there will eventually be a high price that makes the predatory pricing scheme economical.

Now we do see high prices. So arguably, we have the fact pattern that would justify accusation of predatory pricing against the rideshare platforms. And I think to get back to the inflation thing, it's that Most-Favored-Nations clause that makes that doable, because if they didn't have the Most-Favored-Nations clauses on all of these platforms, that's the case where you could see that recoupment wouldn't be possible. Because then they raise the price again, after the predatory period, and that causes everyone to leave the high-price competitor and go to some new would-be entrant.

This is the reason why predatory pricing became de facto legal under the antitrust laws. Because the prediction was, oh, if you did raise prices after you put all your competitors out of business, then you would get new competitors. Your high prices would bring in more people to the market. So you wouldn't actually be able to get away with the recoupment phase. We're seeing people get away with the recoupment phase. And the way they're getting away with the recoupment phase is the MFN.

Thanks so much Marshall, always great talking to you.

Glad we could do this.