JANUARY 28, 2019
THERE ARE TWO TYPES of books in the Silicon Valley canon: books on programming and blockbuster memoirs by entrepreneurs. When I moved to San Francisco after graduating college, I had read just enough of the first kind to land a software job. After receiving my company-issued hoodie and learning the hard way when not to wear it (at parties), I decided it was time to catch up on the second kind of book.
The first thing I noticed is that the West Coast techie strain of elitism is far more flagrant than the East Coast strain I was used to. For instance, in his memoir, Twitter co-founder Biz Stone announces that he has come up with a new “social good”–oriented definition of capitalism (fact check: he is actually just repackaging Andrew Carnegie’s 1889 “The Gospel of Wealth,” a treatise justifying the concentration of wealth as a means of promoting public welfare). Among his catalog of good deeds: switching his company from bottled to filtered tap water (bottles of Starbucks’s Ethos Water available for visitors) and reserving the @red Twitter handle for Bono’s humanitarian brand. Truly, he is virtuous.
However, it’s not just the feigned benevolence nor the arrogance that struck me in these memoirs but how these self-made millionaires talk about consuming, rather than overcoming, competition. Venture capitalist Marc Andreessen celebrates software companies’ “eating the world,” ingesting entrenched industries in their path. Peter Thiel, co-founder of PayPal and Palantir, declares that “competition is for losers,” and claims that companies can only reach their highest creative potential if they don’t waste energy fending off pesky competition. Capitalism, according to all these men, is about frictionless, seamless engorgement.
Their argument that monopolies can be good for the economy and good for consumers is not new — it does not originate in Silicon Valley, as their memoirs would have us believe. It emerged in the late 19th century during the United States’s Gilded Age, advanced by men of unprecedented wealth, like J. P. Morgan, the steel, shipping, rail, banking, and electricity magnate; Andrew Carnegie, the union crusher unafraid of violence; and John D. Rockefeller, the oil man who at his death was worth roughly three times Jeff Bezos in today’s dollars.
What is different, then, is what might be called the stylistics. For one thing, the image of those old top hat–clad robber barons with their cultivated urbaneness is sartorially different from that of today’s hoodie-clad tech titans. And for another, their business tactics are different. The major tech companies don’t extort the public and shake down competitors like the monopolies of old. They give us what we want, or think we want, offering most of their services for free (email, messaging, maps, cloud storage, et cetera). On the whole, consolidation in tech seems actually to help consumers. Or does it?
Tim Wu’s The Curse of Bigness: Antitrust in the New Gilded Age, a history of antitrust in the United States, makes a compelling case that it doesn’t. More generally, he argues that we are, in fact, in an all-out return of the Gilded Age — and perhaps only the clothing has changed. A professor at Columbia Law School known for coining the term “net neutrality,” Wu argues that the very bigness of present-day companies — especially those in the tech sector — does not just harm consumers, but that it also threatens innovation and undermines the power of government.
In a nutshell: His prescription to end the new Gilded Age is the same bitter pill that ended the old one — trust-busting.
It’s worth revisiting the history of trust-busting, which he does in some detail in his book. During the first Gilded Age, powerful trusts were celebrated as an embodiment of social Darwinism — in other words, as survival of the fittest. As Rockefeller himself said, “The American Beauty rose can be produced in the splendor and fragrance which bring cheer to its beholder only by sacrificing the early buds which grow up around it.” Between 1895 and 1904, the number of manufacturing companies operating in the United States declined from 2,274 to a mere 157. Those who survived, including General Electric, US Steel, Standard Oil, and AT&T, used predatory-pricing exclusionary cartels and virulent lobbyists to keep rivals at bay.
Few businesses were safe from the trusts, especially from the United States’s most aggressive bud clipper, Rockefeller himself. In 1904, investigative journalist Ida Tarbell said that, for him, “nothing was too small: the corner grocery store in Browntown, the humble refining still on Oil Creek, the shortest private pipe line. Nothing, for little things grow.” Tarbell’s 19-part exposé on Rockefeller’s Standard Oil would mark a turning point in setting public sentiment against the trusts. And it set the stage for one of the earliest antitrust cases, led by President Theodore Roosevelt.
Though he venerated size and power, Roosevelt forcefully brought his big stick down on the likes of Rockefeller because he understood that the Gilded Age trusts posed existential threats to the power of government. Fewer firms meant more powerful firms that could easily coordinate on prices and unify their lobbying efforts to change laws in their favor. He cleverly gave teeth to the vaguely worded Sherman Anti-Trust Act of 1890, which he then strategically deployed to win landmark cases. In the process, he became known as the trust-buster par excellence.
After World War II, trust-busting became a bipartisan issue undergirding American nationalism. This was in part because people blamed German monopolists for enabling Hitler’s rise to power. Support for antitrust continued to grow until it reached fever pitch in the 1960s. The Johnson administration was so aggressively litigious that it prevented the merger of two Los Angeles–based grocery chains, Von’s Grocery and Shopping Bag Food, which together would have controlled a mere 7.5 percent of market share.
But it was in this very same period, 600 miles west of Washington, that the ideological wheels favoring the return of bigness were already in motion, according to Wu. Known as the Chicago School, a group of libertarian, classical economists at the University of Chicago, including the conservative intellectual heavyweight Robert Bork, were arguing that the sole purpose of the Sherman Act was to promote “consumer welfare.” Monopolies, they said, should only be broken up if they raised prices for consumers, and not simply because they were too big or anti-competitive. When a young Bork first presented his thesis on the consumer welfare standard in 1964, he was considered to be on the “lunatic fringe,” but within 20 years his thesis would shape the majority opinion of the Supreme Court.
As Wu and others have rightly pointed out, the straightforward Chicago School standard overlooks, among other things, the stifling effect monopolies can have on innovation. As a case in point: when AT&T was broken up in 1984, a torrent of new products came on the market, everything from the first answering machines to early ISPs. Until then, “Ma Bell” had prevented other businesses from selling attachments to its phone jacks. Antitrust again shaped the internet in 1998 with the defunct case against Microsoft. Microsoft could easily have crushed early internet companies like Google and Amazon in their infancies if not for regulatory scrutiny, which effectively stopped it from forcing users onto Internet Explorer.
Today, despite what Thiel says about the creative potential of unbridled monopolies, it feels like we are again in one of those innovation bottlenecks. In the decade between 1998 and 2008, the tech sector went from stuffing mailboxes with AOL free-trial CDs to releasing the iPhone 3G, Google Docs, Amazon Prime, and the Facebook News Feed. Another 10 years on, our technology actually looks much the same, and the promises of virtual reality, IoT, chatbots, and blockchain continue to elude us. And yet the value (and influence) of the big tech companies has only grown, now making up five of the top six most valuable companies in the world by market cap. In 2008, none were in the top six.
As we enter the third presidential administration in a row without a major antitrust case, and with none on the horizon, tech companies are consolidating with impunity.  The software that is “eating the world” is eating itself too — Facebook has made 67 straight unchallenged acquisitions, including competitors Instagram and WhatsApp; Amazon has made 91 acquisitions, including Whole Foods; and Google has made 214, including Waze and YouTube. Budding companies that resist being acquired are pruned through flagrant copying. Snapchat experienced this firsthand when its stock price and user base crumbled as soon as its signature “story” feature appeared almost simultaneously on Facebook Inc.’s four largest platforms, Instagram, Messenger, Facebook, and WhatsApp.
The New Gilded Age even holds the risk of something unprecedented in the United States: an undemocratic alliance between big tech companies and the government. In a recent interview, Wu said that, referring to while he was an advisor to the FCC,
Facebook is always trying to be helpful because they want to be our friends, in government … so they came over and they say something like … “We know what everybody’s faces in the world look like, so if you’re looking for someone, let us know.” Now they said this in the context of missing children … [but] you can see that, pretty quickly, if you have a company which has spying powers that actually frankly transcend those of the CIA or the FBI … you have the possibility of totalitarianism that we’ve never really fully witnessed …
But The Curse of Bigness leaves room for optimism in two ways. First, Wu offers a number of specific recommendations (some of them perhaps too specific to be readily understood by a general audience) on how to strengthen antitrust law, including reviewing more mergers, increasing opportunities for public comment, having the government once again take on big cases, and generally moving away from the “consumer welfare” definition of antitrust toward one of protecting competition. Second, he argues that the economies of scale that incentivize bigness are only beneficial up to the point when they become “diseconomies of scale.” In other words, layers of management and internal control systems eventually make larger companies less efficient. So long as they cannot change the rules of the game, they become less competitive as well.
Another book, The Internet Trap: How the Digital Economy Builds Monopolies and Undermines Democracy, casts doubt, however, even on these two glimmers of optimism. Its author, Matthew Hindman, a media studies professor at George Washington University, argues that the rules governing tech are different — bigness inherently translates into the kind of dominance that prunes the sector down to a few winners and a whole lot of losers. He adamantly disputes the claims of yesteryear that the internet is a Wild West meritocracy, where any programmer with grit, gumption, and a good idea could succeed. The internet does not create a level playing field — and nowadays only PR professionals still believe that “competition is just one click away.”
The Internet Trap does a good job pulling back the curtain on how big tech companies make themselves ever bigger: by making major infrastructure investments to give themselves small but compounding advantages. Google learned the hard way that speed can be one such key advantage (or disadvantage) when, in 2000, the search engine experimented with showing 20, 25, and 30 results at a time instead of 10. The change added only half a second to the page’s load time, but that slight delay led to a 20 percent drop in traffic.
Since then, Google’s lodestar has been the “gospel of speed.” That gospel has reshaped the geography of the internet, creating barriers to entry for competitors along the way. Google has laid undersea fiber-optic cables, built data centers inside of local ISPs, and spent a good chunk of its $13.9 billion annual research and development budget, all in the name of a few milliseconds. As Hindman puts it, “Most sites cannot build a new web browser, and then make it the most popular in the world, in order to speed up their site.” 
Hindman’s underlying claim is that if one website has a wide array of content that frequently updates itself and is also tailored to users’ interests, then that website will dominate their attention. But, again, the advantage of bigness holds: the expenditure behind a feature like personalization is only feasible for the largest companies. Recommendation systems, the engines of personalization and the secret sauces behind Google’s search algorithm and Facebook’s News Feed, constitute a massive fixed cost for tech businesses but can bring about exponential growth. Hindman chronicles just how expensive they can be in his account of the Netflix Prize, a one-million-dollar reward offered by the then-DVD-rental service to the researchers who could improve the accuracy of their rating prediction algorithm by one-10th of a star.
As Hindman points out, the prize allows for a rare public glimpse into the challenges of building recommendation systems. With the contest spanning 2,000 active teams from 152 countries working over three years, many teams had to join forces after the first year to remain competitive and eke out small improvements. Eventually two supergroups emerged — BellKor’s Pragmatic Chaos and a 23-team Hydra called “The Ensemble.” In the end, BellKor’s winning solution took more than 2,000 hours of work and included 107 algorithms. By the time the competition was over, Netflix had already pivoted to streaming and the algorithm was never used in its entirety. 
Hindman covers other topics, but the book loses cohesion after the first half when it becomes mired in the statistics and academic jargon of Hindman’s own research. His economic models of the web, of internet traffic, and of local news consumption feel like tangents about methodology and fail to paint a clear picture of internet centralization. By the end of the book, Hindman shifts to advising small local newspapers on how to escape the maw of tech behemoths, and his advice — faster load times, better design, more frequent site updates, increased focus on headlines — feels like too little, too late.
Despite the specter of the Chicago School, the advantages of entrenched tech companies, and the monopolistic evangelism touted by Stone, Andreessen, and Thiel, the pendulum may be finally swinging against bigness. Besides Wu and Hindman, a group of “Neo-Brandeisian” legal scholars have started to question the legality of the tech hegemons under the Sherman Act. Most prominent among them is Lina Khan, a 29-year-old legal wunderkind whose antitrust paper on Amazon in The Yale Law Journal sent reverberations across legal and political spheres. In example after example, her paper highlights how the consumer welfare view of antitrust fails to curb the predatory pricing and vertical integration strategies Amazon uses in its relentless pursuit of growth.
Still, looking across the political landscape, it is unclear who the David is to take on the tech Goliaths. Republicans are cozy with big business,  Democrats are cozy with tech interests, and constituents may not rally behind the issue on their own if the flow of information is controlled by the companies under threat. Still, the ideas are out there, and thanks to recent media coverage, they have penetrated the popular imagination. Perhaps this generation of Silicon Valley migrants will not read the memoirs of self-important entrepreneurs that I slogged through as a young programmer, but rather the work of anti-establishment academics like Wu, Hindman, and Khan. And then, the phrase “trust-buster” may well return to American lips.
 When there have been antitrust cases, it is often to defend tech companies, such as in 2012 when the DOJ sued publishers for colluding to raise the price on popular ebooks. Amazon could set predatory prices because they could recoup their losses over the long-term by locking users to the Kindle platform.
 All of this infrastructure also means that big tech companies can move more quickly to address competitors. For example, in November when Facebook released Lasso, its clone of the popular short-form music video app TikTok, the company already had uploading, video serving, and recommendation technologies it could use from its other apps.
 The Netflix Prize also brought a boom in original research, particularly around restricted Boltzmann machines and singular value decomposition. Netflix used a blend of these two methods until at least 2012. The competition was enough of a success that Netflix attempted to run it again, but they were stopped in 2010 by an FTC lawsuit over privacy concerns.
 Republicans were not always anti-antitrust. It was Richard Nixon who in 1974 started the case that would eventually break up AT&T, then the largest company in the world.