How much of a median American family’s earnings get skimmed off by corporate monopolies and oligopolies? How else does market concentration corrode basic economic and democratic functioning? When I want to ask such questions, I pose them to Thomas Philippon. This present conversation focuses on Philippon’s book The Great Reversal: How America Gave Up on Free Markets. Philippon is the Max L. Heine Professor of Finance in New York University’s Stern School of Business. He has been named one of the “top 25 economists under 45” by the IMF, and has won the Bernácer Prize for Best European Economist under 40. Philippon currently serves as an academic advisor to the Financial Stability Board, and to the Hong Kong Institute for Monetary and Financial Research. He was previously an advisor to the New York Federal Reserve Bank, and senior economic advisor to the French finance minister.
¤
ANDY FITCH: During the late 1990s, as you settled in the US for grad school, you found most goods and services surprisingly cheap, especially given the high wages available. That all contributed to solid purchasing power for workers. And none of that, you tell us, happened by chance — although sometimes it might have felt like it did. So which key choices from which decision-makers had most effectively harnessed this particular convergence of economic forces?
THOMAS PHILIPPON: So first, for a perfect example of how dysfunctional this country has become: I’m like an hour from New York City, on the highway, and I get this poor cell-phone connection, which is what’s making it so hard for you to hear me right now [Laughter].
But anyway, I arrived at MIT in 1999. As a student, you care about the price of pasta, and the price of computers. I think pasta cost about the same as in Europe, but laptops, cell phones, and Internet connections were much cheaper than in Europe. So were plane tickets. Students could fly to conferences, rather than taking the train or bus.
All of that came about because US policymakers had deregulated industries in the 80s and 90s (depending on the industry). Industries had become very competitive. Firms had to compete by offering lower prices.
Since that late-1990s moment, US workers’ purchasing power has diminished significantly, and certain firms have established pronounced market dominance. Again, you say, none of this has happened by chance. So where should we think of market concentration as a symptom, and where as a cause, of increased disparities in both economic and political power?
This gets tricky, because higher concentration in a particular industry could come from two basic causes. In the book, to keep it simple, I typically refer to “good” or “bad” concentration. But we have a rich and complex history to consider.
For an example of “good” concentration, consider what happened with Walmart and the broader retail sector in the 1990s. Walmart came to dominate mostly just because of its greater efficiencies. It could undercut competitors’ prices and increase market share. You can call that a harsh business practice, but it basically just meant Walmart was better. And in that case, we don’t have to worry as much about the increasing concentration. We might eventually worry about broader market functioning, but efficiency gains and competition still should make the whole industry better off.
Airline concentration happened quite differently. The US airline industry consolidated through mergers. Over 12 years, we went from eight major airlines to four. Now many routes only get covered by one or two airlines, who can offer lousy service and high prices.
Telecoms operate similarly these days, with zero entry into the market. Nobody in their right mind would call AT&T and Verizon efficient and innovative companies at the moment — or especially Comcast. Dominance here comes from incumbents protecting themselves, making it so hard to enter the market that they don’t need to worry about competition. The healthcare sector works somewhat similarly.
For retail today, you do see concentration, but you also see real competition — whereas in the telecom and healthcare industries, significant size and market share make firms less efficient. Most other sectors would find themselves somewhere between these two extremes. And COVID-19’s aftermath of course will mostly make concentration worse.
Here The Great Reversal poses the clarion rhetorical question: “Where’s the outrage?” Yet you also acknowledge that a cluster of interpretive snags (including the difficulty of comparing prices across countries, or of recognizing the slow but steady consequences of domestic price rises and hidden fees) have in fact made it quite challenging to get a clear picture on present-day US market concentration. Which data points stood out as most salient as you finally began to recognize how much concentration you yourself had been living through for some time? And why should we always consider multiple forms of measurement as we track concentration and assess its consequences?
Right, here again, concentration by itself does not confirm something bad has happened. You need to bring together multiple indicators, along with some theory, to figure out what’s going on. Prices by themselves won’t give us enough information. Our measurements might determine that some price is too high. But “too high” compared to what? Too high relative to the marginal cost? But how much do variations within this particular industry, or this particular firm, impact that marginal cost? Gathering this kind of data is also extremely hard. Firms won’t give it to you. You have to estimate their profit margins.
So instead, you might try comparing one price to another price. And after looking at tons of data, the best comparisons I could find put certain prices in the US alongside certain prices in Europe. That doesn’t work for everything, but it does for a few well-defined products. How much does it cost to get broadband Internet at home? How much does it cost for a cell-phone plan with a specified amount of data? When you can make the points of comparison this precise, you really see how Americans get ripped off right now.
Pivoting back then to causes of concentration, could we first bring in how comparisons to European markets confirm that 21st-century economic concentration has emerged more acutely in the US than in comparable societies — presumably as a consequence of policy choices, rather than of technological or business innovations (such as winner-take-most digital-platform dynamics)?
One irony here comes from the fact that the EU adopted US-style policies to make European industries more competitive, and to ensure that consumers get a good deal. The policies making European markets competitive right now yielded similar results for the US 20 years ago. So today’s European telecom market looks a lot like the US telecom market two decades back. Consumers can find good prices.
And like you said, this European model can help us to separate “good” concentration from “bad” concentration — and to see whether technology has driven concentration, or policy has. We still might struggle to offer something like empirical proof on these questions. We often can’t find clear real-world comparisons between two societies with similar structural features, yet which make quite different policy choices. But with the US and the EU, the similarities in technological development make this much easier.
Here we can take transport as our example. Car, bus, train, airplane technology does not differ much between the US and the EU. Similarly, with telecoms, no technological difference would make one market have significantly higher costs or prices than the other. So this gives us a clearer sense of how policy-making has produced certain outcomes (especially in terms of the prices consumers pay, and the quality of products they buy). Technology may still play some role, but certainly not a crucial one.
Europe and the US have had telecom industries shaped by both technological and market forces ever since the 19th century. We have a whole regulatory apparatus to deal with them. What’s new is the reluctance in the US to use these tools to regulate the platforms. Internet platforms are admittedly difficult to analyze, and we don’t have them (not yet, at least) to the same extent in Europe. I mean, we have Spotify, but not at the same scale as Facebook in the US. Still I don’t think that network effects by themselves explain US platform firms’ concentrated market share. These platforms have managed to convince everybody that because they’re so amazing and great, we just don’t need to worry about regulating them. And the fact that they can even propagate this myth suggests how much things have changed in recent decades. But from a broader historical perspective, the challenges presented here are actually nothing new.
Could you outline then a few additional factors combining to make EU consumer markets more free: perhaps starting, as you said, from a studious importation of mid-20th-century US economic principles, but then further accelerated by a distinct set of multi-nation negotiations — in which mutual suspicion has superseded tendencies towards both state dominance and state capture? And as a result, what can this EU model teach the whole world today about the broader societal benefits of robust regulatory agencies and related institutional bodies (such as the European Central Bank) pursuing their proper functions?
That gets at the second irony of present-day European markets. Alongside American-style competition happening more in the EU than in the US today, this outcome in Europe should surprise us given Europe’s own history. EU markets still do have significant problems. But if you think back to the mid-90s, when many of these competitive practices were first put in place, the EU had about 15 to 18 countries — none of them with a strong history of free markets. All had long-standing traditions of politicians taking a hands-on approach, controlling everything, mostly concerned about maintaining their own power (really the opposite of what we tend to think of as healthy market conditions). So you might assume that, if you put 18 countries like this around the table, you’ll get a similar system at the EU level.
But instead we got the opposite. At the EU level, these very same people decided to create the most robust and independent market regulators in the world. And this actually shouldn’t puzzle us much. At home, German and French political leaders might have wanted to keep their hands on the industrial policies. But as they started thinking about EU regulations, they mostly worried about some other country gaining power over the regulators. That made for a very different game. And then all of the smaller countries (still important countries, though people tend to leave them out of this history) had to anticipate these two very large economies dominating the policy discussion. The smaller countries needed to ensure that Germany and France and other big countries couldn’t just team up and impose their will on everyone else.
So here again the obvious answer was to make the commission completely independent. Everybody had to give up their chance to exert dominance, but really as a self-protective measure. At that time, people didn’t really see all the consequences this choice would have. They understood the game. They understood why they had made certain decisions — but they still couldn’t predict certain outcomes.
Take lobbying. I don’t think the EU’s designers worried much at that time about firms lobbying the various individual states. They worried more about these states seeking political influence over the economy. But by making, for example, the antitrust regulators so independent from politicians, the EU also ensured its own independence from lobbyists. You didn’t see US-style campaign contributions, or that kind of pressure put on regulators. And still today, corporate lobbyists have less chance of succeeding in Brussels than they do in D.C.
So what can this all tell us about the US itself? Well, starting from the obvious, we need to push for campaign-finance reform. You can’t create a system that resists corruption here unless you rein in campaign spending. Every democracy in the world limits corporate spending during elections. And today you need absolute transparency.
Lobbying itself is fine. Lobbying means exercising your right to express your opinion. You need limits on campaign spending, and you need strong regulators. But if you have that, then lobbying doesn’t need to interfere with market competition.
Here most antitrust theorists I talk to would appreciate The Great Reversal’s lucid demonstration that competition has declined across the whole US economy (not in patchy fashion, as seems likely if prompted by disruptive technologies, or superstar innovators), that this lack of competition has come about largely through policy choices made by corporate-funded political leaders, and that such concentration has brought a host of corrosive consequences including: lower wages for workers, lower rates of investment, lower productivity gains for firms, and slower macroeconomic growth. Many of these theorists would then go on to claim that antitrust approaches need not only reinvigoration, but significant reinvention, especially by rejecting a narrow Borkian focus on consumer pricing. Yet your own analysis suggests to me a more ambivalent perspective. So did relatively robust antitrust enforcement operate sufficiently to bring about those catalyzing conditions you found in the 90s? Could resuming such antitrust enforcement by itself dramatically improve today’s market functioning? Or to what extent did we then or do we now need to move beyond Chicago School approaches prioritizing consumer interests?
I might sound hopelessly middle-of-the-road here. I consider myself an extreme centrist. For Bork specifically: everybody cites him, but nobody has read him. The Antitrust Paradox does offer a couple good ideas. Some of it is half-baked nonsense, and he also has no data. Still I do believe that we need to watch out for excessive bureaucratic enforcement. I do believe that we should focus on consumer welfare. Everything else in the book I’d describe as mostly wrong. This idea that you can equate consumer welfare and short-term prices…everybody knows that’s just ridiculous.
This University of Chicago approach assumes that if markets remain contestable, everything else will work fine. They didn’t worry about monopoly power, because they said: “Look, if this firm makes excess profit margins, that will attract competition. Other firms will enter, and the margins will diminish.” For me, and my approach to economics, that could offer a good hypothesis for a paper to explore through the data. But it definitely wouldn’t fit as your paper’s conclusion!
And if you do look at the data, you find Bork’s account roughly right for his time. When an industry started making loads of money, people did notice, and enter that market and compete. Five years later, margins would come down significantly.
But precisely that doesn’t happen anymore. This model of free entry just doesn’t fit the data today. Still at least we can frame this as an empirical question. My own generation of economists has been much less ideological, because we build on theory to look at data. If you want to know what actually happened, you look at the data and you can hope to get an answer — if you’ve framed your question right (which is why you need theory). Other economists most likely will accept this answer, as long as you can offer convincing data. But Bork didn’t even bother with the data.
On the other hand, we have people described as neo-Brandeisian (after Louis Brandeis) who want antitrust to solve every problem in democratic society. Yes, I do think we need to worry about corporate power. But I wouldn’t rank antitrust as our number-one response to corporate power. I would put campaign-finance reform first. Often on conference panels, I can’t help asking: “Wouldn’t campaign-finance reform have a much bigger impact than antitrust?”
The Great Reversal here makes the persuasive case that with incumbent stability and market share and profit margins (presumably at consumers’ expense) all systematically linked, the free entry of new firms into markets proves crucial. And your book also stresses that we need to consider how firms enter or exit a market: whether, for instance, they stake out new technological ground (or seek to escape some dominant incumbent’s shadow), whether they depart through liquidation (or through consolidation). So again, what does this focus on free entry as an essential rebalancing mechanism tell us about the health of US markets today?
I actually consider this the most useful indicator. It’s pretty intuitive, and it fits with what many market participants perceive. When you see an industry with firms making a lot of money, and with nobody else entering, you can assume barriers to entry exist. Sometimes you will find misguided regulations. Sometimes you find abuse of market dominance. Often you find both.
I also consider this focus on market entry so handy because it should generate bipartisan support. US approaches to competition and regulation used to be very bipartisan. There’s no reason it shouldn’t be that way now. And to restore this bipartisan approach to regulation and antitrust, we can look at entry. Both sides should want free entry. One side resists regulation. One side opposes dominant market players preventing entry in various ways. Again, both of those perspectives offer a lot of compelling points.
Still in present-day political terms, we no doubt have had some highly profitable firms emerge in recent decades. But your data suggest that, from a historical perspective, GAFAM (Google, Amazon, Facebook, Apple, Microsoft) firms’ rates of investment, their share of workers, their impact on broader economy-wide productivity gains, and certainly their tax contributions don’t justify attempts to cast these corporations as exemplary superstars deserving special status or treatment. Many of us might not look back at Walmart’s late-20th-century market dominance as exemplary in those ways either. But how does such a 1990s superstar at least live up to this billing better than today’s GAFAMs?
Of course some people consider these marquee firms of the new economy just so amazing, so fundamentally different from past superstar firms, that the old rules shouldn’t apply. But anybody with a historical sense will tell you that this class of argument (“Today is different”) is often bogus. That argument itself always has existed. And it always has come with these same exaggerated claims about certain firms breaking from the past.
The data show that applying this argument to these present GAFAM firms just doesn’t make sense. These firms certainly have amazing qualities. But if you look at top US firms decade by decade, they all had certain amazing features for their time. If you look at rates of productivity, and profit margins, and market value, you see companies that look a lot like Microsoft and Google today. So we do have impressive firms, yes. But the US economy always has had impressive firms. We shouldn’t consider these current firms any different, and we shouldn’t hold them to different standards. And like you said, they certainly don’t spread their amazingness to the rest of our economy.
Here if we adopt an approach less of punishing “bad” corporations than of promoting broader market flourishing, what potential might you still see for these GAFAMs, operating under updated conditions and constraints, to truly power our economy in productive ways? And in any case, given newfound prospects for data-hoarding firms to entrench themselves all the further, where do we undoubtedly need more effective regulation around data privacy, property rights, and portability?
The US clearly has to do a better job protecting people’s data. California already has something in place. California basically extended what the EU has done. But for the big platforms, we’ve reached the point where data protection won’t suffice. If you want competitive markets, you need to intervene. Whether you break them up, or put constraints on any future acquisitions, these platforms need competition.
And in terms of data, we’ve also reached a point where we just don’t trust these companies. I mean, think of which firms would be natural candidates to improve our healthcare system: the most inefficient market in the US. Google and Apple stand out as obvious choices. You can imagine health apps efficiently collecting data on everybody, and your doctors easily accessing it, and all of that. But we don’t trust these firms enough for this. We sense they already have too much power. Even if they do have good intentions today, we can assume at some point they’ll abuse this power. That just means they’ve become too big, period.
Certain tech firms’ labor-market monopsony dominance likewise stands out for weakening worker power through reduced wages, restrictive contracts, non-poaching agreements. Certain professional trade associations further constrict labor markets through burdensome occupational-licensing requirements and corresponding forms of regulatory capture. So here again, which measurements might you cite to counter any single-minded focus on job-displacing automation as the principle cause of workers’ diminishing labor-market power today?
Well, automation certainly has had a big impact. But automation already has advanced much farther in other countries. Japan has incorporated much more automation into its industrial production. And you don’t find many Japanese people complaining about how automation has led to a big diminishment of labor’s share of income. So I don’t for one second consider that a sufficient explanation. It does apply to certain sectors of American manufacturing. It has little to do with telecoms, or airlines, or transportation more generally — or the energy market, or the healthcare system.
I’d also point out that, 20 years ago, the US had light regulation in its labor markets, and has since trended in the wrong direction. Whereas in France we had our protected pharmacies, and loved our licensing requirements, though then moved in a different direction — the US started with a low baseline of labor-market restriction, yet has since moved in the opposite direction. Here you can think of things like non-compete agreements and non-poaching agreements. These have their rationale in very specific circumstances. But they absolutely get overused. Or we know a society needs a patent system, but at present this system gets vastly overused. It gets abused, really.
You also have situations in which a firm like Amazon, on its consumer-facing side, might not raise prices. But they’ll squeeze their suppliers (who can’t help relying on Amazon), creating another form of monopsony.
So for regulatory agencies, tech innovation offers, you say, “both a blessing and a curse.” Such innovation forces regulators themselves to keep evolving. But it also takes us into uncharted terrain, where opaque, exploitative approaches might find new opportunities to entrench themselves. Within these challenging circumstances, how can we begin to parse whether regulators have made “good” or “bad” decisions? And in any case, what leeway should we give regulators to respond pragmatically to new conditions? How might we foster good regulation less by seeking perfect knowledge or flawless judgment, than by cultivating a transparent, adaptive, and when necessary corrective decision-making process?
I got worried about this while writing the book. In a few recent antitrust cases, judges ruled against the Department of Justice, which is fine. But then the way those DOJ officials got pilloried in the press was just insane — as if they’d literally committed a crime by bringing cases that ended up losing. We can’t have a system where regulators only prosecute one-hundred-percent sure-to-win cases. This is one-hundred-percent a system for under-enforcement. You need to allow regulators to try. Sometimes they might fuck up. That’s fine as long as everybody respects the rule of law, and judicial review. That doesn’t have to suggest any big problem.
Right now we have this double standard, where companies can get away with losing your data or using it in all kinds of ways — and not protecting it, and abusing it. But somehow regulators don’t have the right to make one single mistake.
More broadly then, what has researching this book taught you about the unexpected fragility of free markets — particularly as an economist trained in late-20th-century assumptions that free markets instead would impose discipline on lazy incumbents or inefficient bureaucracies or corrupt regimes?
I underestimated the extent to which big firms can protect themselves by blocking competitors from entering. Again, some of this has to do with lobbying. Some of this has to do with more aggressive predatorial behavior. But I just hadn’t realized how pervasive it was. I sensed it happening, but not to this extent.
I also hadn’t before recognized how extreme some profit margins are in the US, and how much money that adds up to. When I first developed ball-park estimates, I expected to find a few hundred million here or there. But even my conservative estimates found that more than one trillion dollars of labor income gets siphoned off every single year. For the median family, earning about 50 thousand dollars, that means close to five thousand dollars, and a 10-percent reduction in their standard of living. That’s just huge.
By extension, for progressive leaders seeing the need (even pre-COVID) for vastly expanded pro-equality initiatives across our society, how might fostering substantive economic competition go a significant way towards achieving such ends — without requiring politically prohibitive increases in public spending?
So if we take this first two-trillion-dollar stimulus package, and if you think about how lobbyists will misappropriate those funds, it’s just insane. These guys have no shame. They have so much money and influence. So for a lot of this funding, instead of it helping the people who actually need it, it will end up in the pockets of shareholders and big companies — as a direct outcome of our current system. So it doesn’t matter whether or not you call yourself progressive. If you believe in free markets, you should be pissed.
Photo of Thomas Philippon by Saskia Kahn of NYU Photo Bureau.
