Category: Monopolization

I, too, have been enamored of sports metaphors in antitrust. How can the level playing field not convince?

Two wrestlers meet on the floor. If it is uphill for one and downhill for the other, neither will excel. One will find it too easy to win, and so train little. The other will find it too hard to win, and so train little. So, too, in business. If Amazon stands at the top of the hill, because Amazon owns the floor and has chosen to put itself there, then it will do little to improve itself, for it can too easily win against the third-party sellers that it has placed at the bottom of the hill.

But the level playing field is but the pretense of fairness. A way, only, of highlighting a much greater unfairness that we in fact revere. For when the athletes meet, one wins, and not, we like to think, by chance, but because one is better. And why is that one better? Forsooth, because that one does not compete on a level playing field at all. His muscles are better developed. He has better stamina. He is a quicker thinker. He has the focus of mind required to train more. His intuition is better. He has a better spatial sense. And so on. That is, he has an advantage that he does not share with his opponent.

Let us say it is his muscles. In muscle space the field is not level; he stands at the top. And he self-preferences, for he does not, say, starve himself for a week before the bout in order to waste his muscles a bit and thereby level the playing field in muscle space. No! He seizes his advantage. He uses it to win—inevitably to win—and despite this inevitability he feels that he deserves this win, that it is an expression of who he is and not of the tilt of a field.

Why should he feel that his victory is about him given that it was not earned on a truly level playing field? Because it is not a complete leveling that we really seek in any contest. What we seek is to reveal the character of a field that we value. Once we have isolated that field, we glory in whatever tilt we find to it.

If what interests us is who is the strongest, then we want to level the irrelevant fields, and then watch which way the parties slide on the field of strength. If it is the strength of wrestlers, then we level the floor, so that we can better perceive the tilt in their relative strengths.

We can therefore only really object to self-preferencing if the particular instance of self-preferencing at issue relates to a field that we do not think important. We cannot oppose self-preferencing itself, for to do that is to oppose all tilts of field, which is to say, to oppose excellence. It is to insist that no one win the match, or equivalently, that it only ever be determined by a flip of the coin.

(You ask me why the strong massacre the weak in war and want to celebrate it. What challenge is there in that, you say? I say: what challenge was there in your successes, dear reader, any of them, apart from the anxiety you may have felt over whether you would succeed, an anxiety born of your ignorance regarding where your strengths lay? Do you not massacre your opponents too, and call it achievement?)

We can oppose Amazon’s self-preferencing only if it lies in a space that we think irrelevant. If, instead, it reflects a superiority that we desire—if, in the commercial context, it is a superiority in product space, meaning that the self-preferencing delivers better products to consumers—then we must celebrate Amazon’s blood-letting.

We might legitimately say that in giving priority to ads for its own products, Amazon is tilting a field about which we do not care much—the field of marketing—and that prevents the tilt of a different field, that of product quality, from determining the outcome of the game, as we would want it to do.

But we might also conclude that on an ecommerce platform rife with unregulated and unsafe products, the field of information about products should be tilted in favor of Amazon, because then, at least, it is easier for consumers to find the products that Amazon actually stands behind: its own, for which Amazon can be sued if the products turn out to be defective.

So I do not see how the sports metaphor ultimately adds anything to antitrust analysis. It certainly does not teach that the heart of antitrust is fairness, the rules of the game. It merely takes us back to the question that is the heart of all antitrust analysis: does the slaying of competitors improve the product entire, including our ability to find it?

When the New York Journal cabled Mark Twain in London on June 2, 1897 to inquire whether he was gravely ill, Twain famously replied that the reports of his death were greatly exaggerated.

William Randolph Hearst, the Journal’s publisher, could have saved his scoop by having Twain shot on the spot. Fortunately, he didn’t, and we got another 13 years and “Captain Stormfield’s Visit to Heaven” out of the great humorist.

Publishers of university textbooks wouldn’t have been so patient.

Reports of the demise of the printed page, popular since the dawn of the Internet, have turned out to be greatly exaggerated: sales of print books are surging.

So textbook publishers have decided to kill the printed page themselves.

According to a recent antitrust class action brought by university students, all the big names in textbook publishing have been working together to funnel cash to universities in exchange for commitments to assign online-only textbooks to students instead of print books.

It’s working: more than 1,000 universities have agreed to assign publishers’ online-only editions, millions of students have already been forced to purchase them, and publishers are preparing to phase out print textbooks entirely.

Studies show that students, like most readers, prefer the printed page, and textbook companies have seemingly had no problem jacking prices up to astronomical levels in recent years, with the average price of textbook in a core undergraduate course like statistics retailing for more than $300 dollars. So what do publishers have to gain from their assault on the printed page?

A lot, it turns out.

The Rise and Fall of the Internet Used Textbook Market

Eliminating print allows publishers to wipe out competitors that have depressed sales for years.

Before the Internet, textbook publishers had little to fear from the used book market, apart from an occasional copy with a yellow “Used” sticker on the spine that would make its way onto the shelf of a university bookstore.

The Internet changed that, by creating a national–indeed, international–market for used textbooks. Sales volumes of new textbooks plummeted, as students could now pass books along to each other from semester to semester through the medium of online booksellers.

For years, publishers more than offset their losses by jacking up new book prices, but it turns out that there is a limit even to what students with no-questions-asked access to loans are willing to spend for a new textbook.

Indeed, just as excessive tax increases can reduce tax revenues, excessive textbook price increases reduce profits as students start locating bootleg copies on the Internet or shaming their professors into distributing textbook pdfs in violation of copyright rules.

Publishers tried to stem the tide by accelerating the rate at which they put out new textbook editions, even–and rather humorously–in such timeless subjects as basic physics, in order to drive used books to obsolescence and force students to come back to the market for new books.

It didn’t work, which is not to say that it put the major textbook publishers in jeopardy of closing up shop. Textbooks remain the most profitable books in publishing. But publishers preferred to go back to minting money at the old rate. And that’s where online-only books come in.

The Supreme Court has held as recently as 2013 that publishers cannot prohibit students from reselling their textbooks. But it is a staple of Internet law that online publishers can prohibit users from reselling access codes for online material. By killing the printed page, publishers kill the used book market.

The Antitrust Case against the Publishers

There was just one wrinkle that publishers couldn’t iron out on their own: getting universities to assign online-only books. To achieve that, publishers had to buy off the universities, and violate the antitrust laws.

Paying someone to deny your competitors an essential input is called “exclusive dealing” in antitrust lingo, and it’s illegal if the perpetrators have market power and the denial does not help them improve their own products.

But that’s just what publishers do when they pay universities to assign online-only books.

A university’s textbook choices are an essential input into the used book business. If schools don’t assign print books, used book sellers have no textbooks to resell in competition with publishers’ new books.

With used book sellers frozen out of the market, publishers end up with 100% of the textbook market, far in excess of the market shares generally required by the courts to establish market power.

And students end up blinking into the glare of an inferior product.

So this should be an easy antitrust case. But before an increasingly pro-business judiciary, it is anyone’s guess whether the courts will actually get this one right.

From Bad to Worse in the Information Age

The rise of the Internet used book market twenty years ago was itself a disaster for those who love books.

Back in the 1990s, biblioagnostics–those who were indifferent to studying off a new book or a used book–subsidized the bibliophiles who much preferred new books, because the ‘agnostics had to buy new books they didn’t want for lack of a robust used book market. Sales to ‘agnostics kept prices down, enabling bibliophiles to buy new books they could not otherwise afford.

In freeing ‘agnostics to save their money and buy used books instead, the Internet put an end to that subsidy, forcing bibliophiles who could not afford $300 for a new edition to put up with tattered, highlighter-marred tomes.

But if publishers now succeed at killing the printed page, everyone will suffer, not just bibliophiles. For the other thing publishers have to gain from the move to online is the demise of the university itself.

The Assault on the Printed Page Is an Assault on the University

Publishers sell more than just online textbooks. They sell everything a school needs for online learning, offering tests, quizzes, lecture notes, and PowerPoint slides to go along with the textbooks they peddle. And they hire university faculty members to teach instructors how to teach their materials.

When universities accept cash from the publishers in exchange for moving books online, and allow their faculties to indulge in the teaching aids that publishers offer as a perk to make the switch, schools effectively outsource instruction to publishers.

It is not hard to imagine publishers one day cutting out the middleman by offering courses and degrees directly to students. That would wipe out virtually all academic scholarship, save for the sponsored research common in the hard sciences.

Tuition covers about half the cost of instruction at universities, with the difference coming from subsidies. But faculty spend up to half of their putative instructional time producing scholarship, which means that student tuition dollars mostly pay for research, not teaching.

The publishers, however, won’t get any subsidies if they take over the instructional function, so the fees they will charge students will go entirely to instruction. But unless universities are able to make a strong case for conducting scholarship without teaching, which seems unlikely, the subsidies will dry up, and so there will be no money, either tuition or subsidies, left for scholarship.

Universities have been able to force students to pay for scholarship because university education is an oligopoly: brand loyalty–universities call it reputation–makes entry into the market by startups almost impossible, allowing schools to choose their prices without fear of competition.

But it is a virtuous oligopoly that subsidizes a public service, much the way the local advertising monopolies enjoyed by newspapers for most of the 20th century subsidized investigative reporting that was not strictly necessary to attract readers (tabloid headlines suffice for that).

The Internet has already come for newspapers, which lack the extreme brand loyalty enjoyed by universities, but one day it will come for universities too.

If their complicity in the assault on the printed page is any indication, they won’t know what hit them.

(I thank Chris Bradley for comments on a draft of this post.)

One strand of the new antitrust is the notion that bigness is fragile: concentrate wealth and power in a single firm and you put all your eggs in one basket. If the firm fails, the economy is done. Better to break up your behemoths to give the economy resilience in the face of crises.

It turns out, however, that the lesson of Japan’s millenarian firms is quite the opposite: the best way to last a thousand years is to cultivate a monopoly position.

At least according to the Times, which reports that “the Japanese companies that have endured the longest have often been defined by . . . an accumulation of large cash reserves,” which economics teaches is only possible for firms that have market power, and consequently the ability to raise prices above costs and stash the difference.

Indeed, according to the Times, many long-lived firms “started during the 200-year period, beginning in the 17th century, when Japan largely sealed itself off from the outside world, providing a stable business environment.” Read: when the government limited competition.

But wait. Aren’t Japan’s old firms small businesses? Today’s antitrust movement is all for giving small businesses their own mini-monopolies. Possibly because when a small business fails the economy won’t come down with it. (But more likely because this isn’t really about antitrust, but about wealth redistribution.)

So doesn’t the longevity of Japan’s businesses actually support the view that small is resilient? It turns out no. Japan counts big companies like Nintendo among its long-lived firms.

But America doesn’t need medieval Japanese business wisdom to understand that it’s competition, and not monopoly, that’s fragile. We have Schumpeter, who made the resilience of the big the centerpiece of his theory of creative destruction.

He argued that in a world that is more like a stormy sea than the water cooler at the Chicago Board of Trade, the apparent excesses of monopoly are in fact mostly examples of redundancy, the mainmast’s apparently unnecessary girth useful when the big storm hits.

Schumpeter writes:

If for instance a war risk is insurable, nobody objects to a firm’s collecting the cost of this insurance from the buyers of its products. But that risk is no less an element in long-run costs, if there are no facilities for insuring against it, in which case a price strategy aiming at the same end will seem to involve unnecessary restriction and to be productive of excess profits. . . . In analyzing such business strategy ex visu for a given point of time, the investigating economist or government agent sees price policies that seem to him predatory and restrictions of output that seemed to him synonymous with loss of opportunities to produce. He does not see that restrictions of this type are, in the conditions of the perennial gale [of creative destruction], incidents, often unavoidable incidents, of a long-run process of expansion which they protect rather than impede.

Joseph A. Schumpeter, Capitalism, Socialism and Democracy 88 (Harper & Row 1975).

When the Times writes of Japan’s long-lived firms that “[l]arge enterprises in particular keep substantial reserves to ensure that they can continue issuing paychecks and meet their other financial obligations in the event of an economic downturn or a crisis,” it’s hard not to see these firms’ business philosophy as fundamentally Schumpeterian.

Of course, there’s a limit to this kind of thinking. Business longevity and economic growth are two different things. I really am glad we’re not still using fax machines. But although there are plenty of problems with monopoly, fragility is not one of them.

Observations:

• The decline in cases brought by the Department of Justice since the 1970s is consistent with the story of Chicago School influence over antitrust. What is perhaps less well known, but clearly reflected in the data, is that the Chicago Revolution took place in the Ford, and especially the Carter, Administrations, not, as is sometimes supposed, in the Reagan Administration, although Reagan supplied the coup de grace.
  
  Indeed, we have only five monopolization cases filed by DOJ over the course of the entire Carter Administration, as compared with 58 filed during the part of the Nixon Administration and the Ford Administration covered by this data series. This is consistent with the broader influence of the Chicago School over regulation of business. It was also under Ford and Carter, not Reagan, that deregulation got underway, with partial deregulation of railroads (1976), near-complete deregulation of airlines (1978), and partial deregulation of trucking (1980) (more here).
  
  The timing suggests that the Chicago School’s victories were intellectual, rather than merely partisan. As Przemyslaw Palka has pointed out to me, Milton Friedman consciously pursued a strategy of intellectual, rather than political warfare, because he understood that victory on the intellectual plane is more complete and enduring (a nice discussion of this may be found here on pages 218-221). As these numbers suggest, Chicago prevailed by converting its adversaries, so that even when its adversaries were nominally in political power under Carter, they implemented Chicago’s own agenda.

• To the extent that the early part of the FTC data series is reliable (more on that below), the story in the FTC case numbers is that of the six monopolization cases brought over the past five years, following a twenty-year period during which the FTC brought only three cases. With the exception of Google, which has just been filed, there has been no corresponding uptick in monopolization cases filed by the Department of Justice.

• The private litigation data show that in some years (1998 and 2013), private litigation across the entire United States has produced fewer monopolization cases (against unique defendants) than did a single federal enforcer–the DOJ–in 1971. The private litigation numbers for 1997 to 2020 also show that, on average, about twenty defendants face new monopolization actions each year when federal enforcers are filing near-zero complaints. To the extent that the numbers for 1974 to 1983 are reliable (of which more below), they suggest that private cases have also declined markedly since the 1970s, although there was a lag of several years between the two effects, perhaps due to the tendency of private plaintiffs to file follow-on cases to government cases.

• Altogether, one is left with the impression that corporate America has been awfully well-behaved since about 1975.

Notes on the Data:

• The cases brought by the Department of Justice (DOJ) come from the Antitrust Division’s own workload statistics, so I assume the numbers are accurate. For DOJ cases investigated, as well as filed, see here.

• The cases filed by private plaintiffs come from two sources. The first, for the years 1997 to 2020, is a search for Section 2 complaints in federal court dockets via Lexis CourtLink. I must thank Beau Steenken, Instructional Services Librarian & Associate Professor of Legal Research at University of Kentucky Rosenberg College of Law, for figuring out how to search CourtLink for Section 2 cases (no easy task, it turns out).
  
  These are only cases for which the plaintiff, in filing the complaint, indicated the cause of action as Section 2 of the Sherman Act in the court’s cover sheet. Apart from deleting a few cases in which DOJ was the plaintiff, and a few cases in which the case was filed by mistake (e.g., the case name reads: “error”), I did not examine these cases at all, other than to note that many of the defendants look plausible (e.g., Microsoft comes up a lot in the late 1990s or early 2000s).
  
  Finally, I counted only unique defendants in any given year. So for example, if there were ten cases filed against Microsoft in 2000, I counted that as only one case. The reason is that multiple consumers or competitors might be harmed by a single piece of anticompetitive conduct undertaken by a monopolist, and so one would expect multiple plaintiffs to sue the monopolist based on the same conduct. For those interested in using case counts to measure enforcement, all of those cases signal the same thing, that a particular anticompetitive practice has been challenged, and so all of the cases together really only represent a single instance of enforcement. I did not, however, check each complaint to make sure that the alleged conduct was the same across all complaints. I just assumed that multiple complaints filed in a given year against a single defendant relate to the same conduct. (I did not, however, count unique defendants across plaintiff types: the Justice Department case against Microsoft was counted toward DOJ cases and and any private cases filed against Microsoft in the same year count as a an additional single case in the private cases account.)
  
  According to CourtLink, some federal courts adopted online filing later than others, and CourtLink only has electronic dockets. I chose to use 1997 as the start year for this count, because by that year almost all jurisdictions were online and so presumably their dockets are part of the CourtLink database. According to CourtLink, several jurisdictions had not yet moved online by that year, however, and so the counts may be slightly skewed low in the first few years after 1997 because they miss cases filed in the jurisdictions that were still offline during that period. The jurisdictions that went online after January 1, 1997, and the year in which they went online, are District of New Mexico (1998), District of Nevada (1999), and District of Alaska (2004).
  
  The source of the data for the years 1974 to 1983 is Table 6 in this article. That table gives the yearly percentage of refusal to deal and predatory pricing cases in a sample of 2,357 cases from five courts, Southern District of New York, Northern District of Illinois, Northern District of California, Western District of Missouri, and Northern District of Georgia, as well as the total number of private antitrust cases filed per year. Because I suspect that my CourtLink data represents “pure” Section 2 cases–cases in which the Section 2 claim is the principal claim in the case–I adjusted these percentages using information from Table 1 in the paper about the share of those percentages that represent primary claims. Because the total yearly private cases given in the Article did not appear to be adjusted for multiple cases filed against the same defendant in a given year, as I adjusted the CourtLink data, I therefore further reduced the results in the same proportion as my CourtLink results were reduced when I eliminated multiple cases against the same defendant, a reduction of about 40%.

• I collected the FTC data by searching for cases labeled “Single-Firm Conduct” in the FTC’s “cases and proceedings” database. The cases and proceedings database goes back to 1996, and so I labeled years for which there were no hits as years of zero cases going back to 1996. However, the FTC website does caution that some older cases are searchable only by name and year, and presumably not by case type, so it is possible that this data fails to count cases from early in the period (e.g., late 1990s). I also paged through the “Annual Antitrust Enforcement Activities Reports” issued by the FTC between 1996 and 2008 and found a couple of cases not returned by the search of the cases and proceedings database. Finally, I included the FTC’s case against Intel, filed in 2009. I counted both administrative complaints filed in the FTC’s own internal adjudication system and complaints filed by the FTC in federal court. The FTC cases are nominally brought under Section 5 of the FTC Act, through which the FTC enforces Section 2 of the Sherman Act.

Still in those early days trade depended not upon the quality of the goods but upon the military force to control the markets. The Dutch consequently valued the island chiefly on account of its strategical position. From Formosa the Spanish commerce between Manila and China, and the Portuguese commerce between Macau and Japan could by constant attacks be made so precarious that much of it would be thrown into the hands of the Dutch, while the latter’s dealings with China and Japan would be subject to no interruptions.

James W. Davidson, The Island of Formosa, Past and Present (1903).

Here Davidson nicely contrasts monopolies based on product quality with monopolies based on force, capitalism with mercantilism. I do not think it is too much to say that democracy, or at least a genuine republicanism, even if autocratic in administration, is the principal bulwark between the two, and that antitrust, when used properly, is meant to round off any remaining mercantilist edges.

When used improperly, antitrust is of course a gunboat all of its own.