What Antitrust Is Really Against, Part II: Why Did Rockefeller and Gates Frighten People More Than Apple?

Controlling a whole market on merit can be tolerated, and taking most of an industry's profit isn't much of a problem either. The trouble starts when a company uses its monopoly to decide who may use a piece of infrastructure.

On March 16, the Christchurch shooting in New Zealand killed 50 people. Beyond the attacker, blame landed on two things: gun control, and how Facebook — which has a monopoly on the spread of digital information — helped spread hatred.

It had been a week and a half since Elizabeth Warren, a candidate for the 2020 U.S. presidency, called for breaking up the big tech companies, and the argument over antitrust action against them had grown louder. Regulators in the EU, the UK and India joined in.

One question kept coming back: how big does a big company have to get before antitrust regulators need to step in?

As the New York Times reporter David Streitfeld put it, “antitrust is the nuclear weapon of regulatory policy.” In the United States, the country that invented antitrust law, this weapon once dismembered Rockefeller's oil empire, and came close to dismembering Gates's Microsoft.

Rockefeller's Standard Oil and Gates's Microsoft were each the most successful business story of its era. Each held 90 percent of its market, made a great deal of money, and was at one point the most valuable company in the world.

None of that, by itself, is grounds for antitrust action. Apple is the example. Apple never took 90 percent of the market, but the share of the phone industry's profit it extracts is far above anything the oil company or the PC company ever reached, and for four years running it has been the world's most profitable and most valuable company.

Apple's real antitrust troubles didn't start until this year — after it stopped being the most valuable company.

In more than a century of antitrust history, being too good at making money has never been the problem. Neither has owning a market outright.

Standard Oil Won the Whole Oil Business on Commerce and Technology — But That Wasn't the Point

John Rockefeller and Standard Oil did not discover oil, just as Henry Ford did not invent the automobile.

But as Henry Demarest Lloyd wrote in an 1881 issue of The Atlantic, “Very few of the forty millions of people in the United States who burn kerosene know that its production, manufacture, and export, its price at home and abroad, have been controlled for years by a single corporation — the Standard Oil Company.”

In 1859, Rockefeller was 20. He had just quit his job as a bookkeeper at a trading firm and started his own produce business. That same year, the first oil well was drilled in Pennsylvania, a state rich in oil. The “black gold” rush began, and people started digging for oil the way they had once dug for gold.

Unlike most oilmen, Rockefeller was not obsessed with drilling. What he wanted was refining. Drilling carried risk; refining produced steadier cash. By 1863, at 24, he had a refinery of his own.

Soon a new rail line linked Cleveland, where he lived, to Pennsylvania, and his market grew fast.

This is a story about technology and invention. In the early days, Rockefeller noticed that men at his refinery used 40 drops of solder to seal a five-gallon oil can; he tried it himself and found 39 would do. He made his own cans for 96 cents each rather than buy them for $2.50. He and his partners ran experiments and found better, cheaper ways to refine. These small, dull savings added up.

An oil empire took shape out of that.

By 1869 Rockefeller's plant was the largest refinery in the Cleveland area, with its own trains, barges and warehouses, refining 1,500 barrels of crude a day. Many small outfits could refine only a barrel or two a day, and even the bigger rivals often needed a week to do 1,500. A year later he founded the Standard Oil Company of Ohio.

By January 1871 it held 10 percent of the refining market, making it the largest player in the business. By 1873, through mergers, it held a third of the market and refined 10,000 barrels a day. By the end of the 19th century, Standard Oil controlled 90 percent of America's refining.

Consumers were not hurt along the way. One reason Standard Oil bought up other refineries was Rockefeller's worry that oil from shoddy plants would explode and damage the business he had built on high-grade product. So he was glad to buy substandard refineries and upgrade them to turn out better kerosene. And as the technology improved and production got more efficient, Standard Oil's prices fell rather than rose.

“That's why Rockefeller called his company Standard Oil,” said the economist Jim Johnson. Beating rivals and taking a market with better technology is, in fact, exactly what the natural-monopoly theory of market economics endorses.

By the time Standard Oil had the whole refining market, oil was cheaper and safer than before, and had become a major engine of the American economy.

A company started from nothing made its founder the richest man in the world while he was still in his thirties and forties. Consumers gained too: life got easier.

Standard Oil's story is not very different from the inspiring business stories that came after it.

The Trouble Began When Standard Oil Teamed Up With the Railroads to Rig the Market

In 1867 a Lake Shore Railroad line connected Cleveland, home to Standard Oil, with the oil region.

Rockefeller's partner went straight to the railroad's vice president and offered to ship crude from the oil region to Cleveland at 35 cents a barrel, and kerosene to New York — usually for export — at $1.30 a barrel, at least 15 cents below what competitors paid.

In exchange for the discount, Standard Oil promised the Lake Shore volume too large to turn down. Here is how James Devereux, then the railroad's vice president, explained it:

“At that time a freight car took an average of 30 days to make the round trip between Cleveland and New York. Sixty cars a day meant 1,800 cars for one full cycle, and because orders were unpredictable the average cost per car ran about $500, so the business demanded a $900,000 investment. But the study showed that if sixty carloads a day could be absolutely guaranteed on a regular schedule, the Cleveland–New York round trip could be cut to ten days, the business would need only 600 cars, and the total investment would fall to $300,000.”

This plainly made it harder for other refiners to compete with Standard Oil. But Rockefeller did not stop there. He drew other refineries into an alliance with him, and membership brought the discounted freight rates. The alliance later bought a company that held a special transportation charter, whose value lay in the vague but sweeping powers it granted. Under its cover, the alliance set up two freight rates: members settled at a cheap “contract rate,” outsiders were forced to pay more.

Before the railroads, only cities on a river or a coast could hope for commercial prosperity. After them, the United States became one huge single market. The railroad was the most important infrastructure of the age. And Standard Oil decided who could use it, so its competitors were no longer competing on technology or scale, and nearly every form of resistance was futile.

In the end the government stepped in.

In 1890 Congress passed the Sherman Antitrust Act, which gave the federal government a legal basis for acting against corporate monopoly.

Theodore Roosevelt won reelection in 1904 and promptly directed the Justice Department to open an antitrust investigation into Standard Oil. His successor, President Taft, went further. Roosevelt wanted monopolies restrained; Taft wanted them broken up.

In 1911 the Supreme Court found Standard Oil guilty of monopolizing the oil business through a series of anticompetitive acts, and broke it into 34 separate, competing companies.

Antitrust Law Kept Evolving, But Its Core Never Changed: No Company Should Control Infrastructure to Restrict Competition

American antitrust law is not one statute but a body of them.

The Sherman Act came first; the Clayton Act and the Federal Trade Commission Act, both passed in 1914, are also counted among the most important antitrust statutes. The 1930s brought the Robinson-Patman Act, the Wheeler-Lea Act and others.

The legal definition of price restrictions also starts with the Clayton Act. Section 2 let merchants charge different buyers different prices for the same good, but made such price discrimination illegal if it substantially lessened competition or was meant to create a monopoly. The Robinson-Patman Act went further: charging different prices to different people or regions, and cutting prices selectively to kill off a competitor, were both illegal.

In 1950 came the Celler-Kefauver Act, a broad restriction on corporate mergers also known as the Merger Act. Two decades later, three or four more acts tightened merger rules further. These laws shaped American judicial practice on mergers enormously — a deal in which a company with 1 percent of a market bought another company with 1 percent of it could be ruled a monopoly.

A run of federal regulators was set up in the same period. The Federal Reserve System, which acts as the central bank, was founded in 1913, and the Federal Trade Commission a year later. Then the Federal Power Commission (1920), the Federal Home Loan Bank Board, which oversaw savings and loan institutions (1932), the Federal Communications Commission and the Securities and Exchange Commission (1934), and the Civil Aeronautics Board (1940, forerunner of the Federal Aviation Administration).

But when those agencies eventually worked with the Justice Department or the FTC to examine whether a company was behaving as a monopoly, the questions came down to two: did it restrict competition, and did it hurt consumers?

The Sherman Act as passed banned monopoly outright, and that kind of blanket ban created new problems. Any agreement or act involving trade is by nature a restraint of some kind; ban them all without exception and free trade itself becomes impossible. That, in turn, limited what courts could actually do.

Justice Edward White, who took part in the Standard Oil breakup, applied what was later called the “rule of reason”: reasonable restraints on trade were not a crime, only unreasonable ones were. That became the reading used to find Standard Oil guilty of monopoly and of restricting competition.

What counts as “unreasonable” has shifted with the times — from competition within one industry, out to the suppliers and buyers along the chain, and then to influence across industries — with the effect on consumers weighed throughout.

Zhao Wei, a professor at Zhejiang University's School of Economics, put it this way in a paper on how U.S. antitrust policy developed: “Even after all these federal agencies and antitrust laws, American law's definition of anticompetitive conduct stayed fairly vague — and the vagueness was deliberate, meant to give the courts and the administration plenty of room to decide.”

The Standard Oil breakup is one example. The Sherman Act declared “restraint of trade” and “monopolizing commerce” illegal, but never said precisely what those acts were.

The same goes for the rule of reason. It was meant to separate good monopolies from bad ones; the good ones were not illegal.

That line kept moving. In its 1945 ruling on the Alcoa case, the court held that “monopoly itself is illegal, reasonable conduct or not,” and that “controlling 90 percent of an industry's business certainly violates antitrust law. Even 60 percent could be held a violation.” Against that backdrop General Motors watched its market share carefully, working hard to keep it under 45 percent. But when the court threw out a monopoly ruling against Kodak in 1979, its reasoning was that a large company that gains a competitive advantage “merely because of its efficient size” is no monopoly.

This back-and-forth runs through the whole of American antitrust history, but the central test has never changed: how far a large company uses infrastructure to do harm.

Microsoft Nearly Got Broken Up for the Same Reason — It Used the Internet's Infrastructure to Crush Competitors

In May 1998 the Justice Department under Clinton, joined by the attorneys general of 20 states, sued Microsoft on antitrust grounds. Gates's company then controlled more than 90 percent of the PC operating-system market. The suit accused it of using the power of Windows to monopolize the browser market: Microsoft bundled Internet Explorer with every computer that shipped with Windows, signed exclusive deals with internet service providers, and put the IE icon on the desktop by default.

After more than two years of tug-of-war, the court ruled that Microsoft had illegally used monopoly power to squeeze out rivals, and at one point ordered it split into two companies, one for the operating system and one for applications.

An appeals judge overturned that order, chiefly on the ground that Microsoft's browser was free. The court kept the finding that Microsoft was a monopoly, but held that although its conduct hurt weaker competitors, consumers still gained, and that maximizing consumer benefit was good for the economy as a whole.

Microsoft finally settled with the Justice Department — paying fines and opening part of the Windows source code — and escaped a breakup.

The Justice Department's investigation findings, released later, show that officials reasoned this way: after the PC arrived, Microsoft designed the operating system that ran on it and used that technical lead to take the largest market share. Once the internet went commercial, servers linked millions of computers together, and those computers used browsers to reach the pages stored on the servers.

The personal computer was, at that moment, infrastructure you could not get online without. And Microsoft, having invented the operating system, decided what software a computer had to ship with — which meant that at the birth of the internet it decided whose browser consumers used to get there. Microsoft was accused of withholding support from browser developers, of denying them a development environment as fair as the one its own IE team enjoyed, and of making third-party products harder to build.

Thomas Jackson, the federal judge who favored breaking Microsoft up, compared it to Standard Oil and compared Gates to the richest man of a century earlier, John Rockefeller. In the end the Justice Department's monopoly case against Microsoft rested on the same ground as its case against Standard Oil: how each used infrastructure to restrict competition.

The browser turned out not to matter so much — the search engine, not the browser, became the internet's front door. But to people at the end of the 1990s, deciding which browser consumers used was a frightening amount of power.

Gates stepped down as Microsoft's CEO in 2000, and industry figures interviewed by The Seattle Times and Forbes still thought it might have had something to do with the antitrust case, guessing that he had simply grown tired of the long fight with the Justice Department.

Look across the whole of American antitrust history and the last fifty years of economic growth, and the 1990s stand out as a watershed. On one side, U.S. growth fell 0.1 percent year on year in 1991, taken as the start of the ninth postwar recession. On the other, Clinton, who took office in 1992, encouraged investment and mergers and tilted economic policy modestly toward high technology — yet on government action against monopoly he was not only more aggressive than his predecessor, he also put the focus back on protecting infrastructure.

Microsoft's case happened to land right on that shift. It was “lucky” enough to stay in one piece, but it did rein itself in when competing with rivals. Steve Ballmer, who took over from Gates, was known for savaging competitors, but Microsoft under him never forced users to see a particular website first thing each day, the way 3721, Tencent and the Sogou browser did. After the antitrust suit IE still held an enormous share, but Microsoft never used it as anything more than a tool for users — and it was a bad enough tool to give Chrome and Firefox their opening.

Once the Government Decided Internet Companies Were Controlling the New Infrastructure, a New Round of Antitrust Began

Much as a century ago, today's big technology companies, the ones widely discussed as suspected monopolists, are without exception the companies with the deepest technology and the most mature business models in their industries, and they give consumers and customers the best service available.

Just as unexceptionally, each of them owns the most important infrastructure in its industry. Amazon has taken over online shopping, Google decides which websites appear at the top of a search, and Facebook decides what you get to see other people say. Their rivals have picked up similar “skills” to varying degrees.

Of these companies Google has taken the hardest antitrust hit so far, with at least $7.7 billion in fines from the EU. In 2017 it was accused of using its dominant position in search to give its own shopping-comparison service prime placement in product results, holding competitors down. In 2018 it was accused of using Android's dominant position to force phone makers to preinstall the full family of Google apps.

Looking back, the control seems to have started early. But as Warren said, 25 years ago none of these companies existed. They rose so fast that regulators had no time to react. Facebook above all: in 2012 it was written off as a joke that could only lose money; by 2017 it was suspected of swaying elections and filtering speech, and both the left and the right believed it.

At the same time, the U.S. government really did become far more permissive about antitrust. The number of antitrust cases filed nationwide fell sharply, from an average of 15.7 a year between 1970 and 1999 to fewer than 3 a year between 2000 and 2014.

In 2016 The Economist ran three cover stories on the history of monopoly in America. They found that two-thirds of industries had grown more concentrated since the 1990s, and that since 2008 the value of corporate mergers had passed $10 trillion, with the pace still quickening.

There is reason to think regulators leaned on market principles: in a market driven by profit, natural monopolies appear, and a well-ordered market economy can run the cycle of monopoly and its undoing on its own — the way Sears was upended by Walmart, and Walmart in turn by Amazon. None of that needed the government.

Which is why it was no problem that Apple at one point took nearly all of the phone industry's premium users and profit, and charged every app in the App Store 30 percent.

Nobody called Apple a monopoly. Netflix and Spotify both accepted the App Store's 30 percent cut, however unwillingly. That is not hard to understand: the rate roughly matched, and sometimes exceeded, their own gross margins.

But as selling iPhones stopped paying as well as it used to, Apple started its own digital content plan — selling a music service, signing studios to make shows — and walked straight into Netflix's and Spotify's markets. Accusations of monopoly, and resistance to Apple, followed.

On the 13th of this month Spotify complained to the European Commission that Apple was breaking antitrust law, arguing that Apple takes 30 percent of what Spotify sells through the App Store while its own Apple Music pays nothing, leaving Spotify weaker in the fight. Netflix, after Apple announced its video service, likewise stopped supporting subscriptions bought through Apple's payment system.

Apple created the iPhone and the App Store. Like a railroad company controlling one region's market in the 19th century, it decides who may reach a billion iPhone users willing to pay, and it charges 30 percent of the value of what they sell. None of that was a problem. But when Apple used that 30 percent to squeeze rivals' margins while jumping in to offer the same services itself, the fear began.

Cover image/The Antitrust Attorney Blog