- The defining feature of markets over the past decade is the meteoric rise of the digital platforms: Software has eaten the world and a majority of Americans now fear the sector has gotten too big, too powerful, too quickly.
- One consequence is that antitrust enforcement is back, after a quiet two decades: However, the tech behemoths have amassed huge war chests for legal defenses and such complex cases typically take 4+ years to play out. Google’s trial begins Sept. 12, 2023.
- It is certainly not a coincidence that Biden’s nominee for Attorney General, Merrick Garland, has extensive experience in antitrust: However, his votes on a variety of antitrust issues suggest a balanced jurist and not a judicial activist.
- The cases against Google and Facebook are focused on two activities, exclusionary deals and acquisitions whose intention is to stifle competition: Be careful of confident predictions regarding the outcomes though, as there are few antitrust precedents, especially for the digital age.
- The “Policeman at the elbow” theory of antitrust enforcement: Previous cases demonstrate even unsuccessful suits (e.g., IBM, Microsoft), have a dramatic effect on behavior. All tech companies will now be much more careful about activity that could be construed as anti-competitive.
- For investors, antitrust has historically represented more of an opportunity than a threat: Major cases, including Standard Oil, AT&T, IBM, and Microsoft, had a positive impact on competition and unleashed a wave of innovation. Typically, even shareholders of the targeted firm benefited.
- Historical precedents suggest the recent antitrust cases will unleash a surge of new platform companies challenging the middle-age incumbents: This will likely kindle greater breadth in tech market leadership and the emergence of entirely new sub-sectors.
- Beyond antitrust, there are bipartisan calls for reform of Section 230 (indemnifies tech platforms for items their users post): Democrats argue § 230 allows content that is violent and misleading, while Republicans allege it empowers platforms to censor conservative voices. Regardless, there are three approaches to reining in § 230 that we believe are likely to make progress.
- Additionally, digital platforms face the inevitability of their own regulatory agency: Over the last 150 years every major emerging technology gets regulated, e.g., railways, radio, electricity, autos, airplanes, nuclear power, TV, telecoms, pharma. Today’s dominant platforms will not prove an exception.
- Tech policy will change under Biden and the Democrat trifecta, but don’t expect revolution: There will be additional antitrust cases filed, but that would have happened regardless. Moreover, while there exists bi-partisan agreement on the need to rein in tech, there remains profound disagreement regarding which policies to prioritize.
- There are, however, five initiatives that possess bipartisan support: (1) More funding for the FTC and DoJ, so they can take on wealthy, heavily lawyered tech companies; (2) Changing the burden of proof, so the onus is on big tech to demonstrate acquisitions will not harm competition; (3) Ensuring a level playing field, so platforms can’t favor their own listings above those of competitors; (4) Reform of § 230; and (5) Data privacy and portability rules that give consumers more control.
- Regardless, we believe tech will continue to be the most dynamic sector of the economy: Digital platforms will constitute a sizeable majority of market cap by the end of the decade, and we expect greater breadth in tech market leadership and the emergence of entirely new sub-sectors.
“Shall the industrial policy of America be that of competition or that of monopoly?”
—Louis Brandeis (U.S. Supreme Court, 1916-1939)
The recent cascade of antitrust suits is indicative of the growing backlash against big tech, a movement that appears destined to usher in several major developments. After flying under the radar for decades, tech has now attracted the attention of all three branches of government and for one simple reason. Following their meteoric rise over the last decade, the digital platforms are perceived as having gotten too big, too powerful, too quickly. In a rare display of bipartisanship, Democrats and Republicans now agree big tech needs to be confronted on antitrust, § 230 immunity, data privacy, and more.
Have They Not Noticed the Product is Free?
Big tech counters that they’ve been highly innovative and provide enormous value to consumers. They also argue that antitrust and regulation is unnecessary because, as Google’s Larry Page likes to quip, “Competition is only a click away.” Moreover, many of the dominant tech companies of the past seemed equally unassailable but then faced unexpected competition due to technological disruptions that created new markets and new companies. The prime example is IBM, whose dominance in the 1960s and early 1970s was obliterated by Microsoft and Intel, reflecting the emergence of the PC. However, this convenient narrative ignores the crucial role played by the antitrust investigations and trials against IBM from 1967 to 1982 which helped create the opening for the next wave of innovators. Like it or not, for big tech, 2020 marked the year in which government intervention became an inescapable fact of life.
The remainder of this note kicks off with a brief discussion of the key antitrust precedents (Standard Oil, AT&T, IBM), clarifying the lessons they hold for today. Next, we demonstrate why antitrust is currently in the spotlight and then dive into the main takeaways from the Google and Facebook cases. After that we explain why Section 230 is such a lightning rod, why reform is so difficult and what progress is most likely to occur. The following section demonstrates why tech is about to become a regulated sector and what that means for markets. Finally, we explain how tech policy is likely to change under Biden, before concluding with the key takeaways for investors.
John D. Rockefeller and the Granddaddy of Antitrust Cases
The Sherman Act of 1890 was passed to curtail the growing dominance of the nineteenth century trusts, which dominated sectors such as oil and steel. In 1911 the U.S. Supreme Court ordered the break-up of the Standard Oil Trust, a company that produced more than 90% of America’s refined oil. Rockefeller’s creation was convicted because of its pattern of acquiring oil refineries throughout the country, together with various exclusive practices toward remaining independent rivals. It is often said “data is the new oil,” and it is hard not to be struck by the parallels between the allegations against today’s digital platforms and the anticompetitive behavior Standard Oil was convicted of.
After the Supreme Court broke Rockefeller’s behemoth into 34 components, many immediately flourished (e.g., Exxon, Mobil, and Chevron) and soon ranked amongst the most powerful companies in the world. Moreover, shareholders did fantastically well following the break-up, with Rockefeller quintupling his wealth. The break-up of AT&T in the 1980s was similar in several respects and we could expect a comparable outcome if courts force divestitures by Facebook or Google.
“Standard Oil was broken into its constituent parts, among them seven “majors,” …. within a year, the value of what had been Standard Oil had doubled, and in several years, had increased five-fold.”
— Tim Wu, “The Curse of Bigness: Antitrust in the New Gilded Age”
Rockefeller tried hard to convince President Theodore Roosevelt and Congress that his company was good for the nation and was largely responsible for the dramatic decline in oil prices paid by consumers. He also issued dire warnings of the damage a break-up would inflict on the economy. However, the courts were not swayed by such arguments and the inefficiency of his leviathan became clear post-break-up. As typically occurs following antitrust cases, the dissolution of Standard Oil ushered in an enormous boon to innovation and the industry’s further expansion.
Why were the new oil companies so much more innovative than their parent? One example serves to illustrate this point. In 1909, a Standard Oil chemist, William Burton, invented “thermal cracking,” which vastly improved the process of turning oil into gasoline. The Indiana branch applied to headquarters to put $1 million into developing the process, but HQ said no, despite the burgeoning auto industry (the first model T was produced in 1908). Standard Oil primarily sold kerosene, and viewed the investment requested by Burton as too risky to what had become a smug, slothful monopoly.
Luckily for Burton and early auto enthusiasts, immediately after the break-up Standard Oil of Indiana authorized $709,000 to build 120 stills. The thermal cracking method was patented in early 1913 and the “Burton process” was widely used until 1937 when it was superseded by catalytic cracking. In the interim, the entire industry was licensing patents from the company. While all shareholders of the parent Trust did fantastically well, Indiana shareholders did even better. Moreover, this innovation was a key driver of the era of cheap gas, which owed some gratitude to the dissolution of Standard Oil.
As a final point, antitrust always moves like molasses. The case against Standard Oil was filed decades after its greatest offenses. The same is true of the suits against AT&T and, to a lesser extent, IBM. By this standard, the cases against Facebook and Google (which are focused on acquisitions of Instagram in 2012 and WhatsApp in 2014, or YouTube in 2006 and DoubleClick in 2007), are coming remarkably quickly. With that, the next section moves on to briefly discuss the breakup of AT&T and why it is relevant to today’s antitrust cases.
AT&T: Breaking Up is (Not) Hard to Do
A common response to the idea of breaking up dominant firms is to argue that the costs and difficulties are overwhelming, as it is the equivalent of unscrambling eggs. However, the historical record, based on large complex dissolutions such as Standard Oil and AT&T, suggests this is not the case. Breakups are just not that hard to do.
To illustrate, an insider’s account of AT&T’s forced divestiture1 describes the process of constructing eight new companies out of the titan. It required reallocating 70 million customer accounts, 200 million customer records, 24,000 buildings, 177,000 motor vehicles, and one million employees, all in a period of two years. The mandate was to separate long distance from local service operations, and to create seven geography-based regional Bells. Despite protests that the task was unrealistic and would lead to economic ruin, divestiture was accomplished, on time, and with considerable benefit to competition, consumers, and shareholders.
A recent article2 examines the AT&T experience to fathom how a forced divestiture of Instagram and Whats-App could proceed. He concludes that each of the three firms should receive non-exclusive licenses to one another’s technology, along with shared access to the customer database, although with customer rights to opt out. For customers this would mean each could communicate and share data seamlessly with users from the other two. These principles guided the highly successful AT&T break-up forty years ago and could prove equally effective today.
Does this mean ATT shareholders benefited from the divestiture? The short answer is yes. The market value of AT&T, just before the settlement was announced in January 1982, was $47.5 bn. A decade later, the combined market value of the eight successor companies (the stripped-down AT&T, plus Ameritech, Bell Atlantic, BellSouth, Nynex, Pacific Telesis, Southwestern Bell, and US West) came to $180 bn, significantly outperforming the S&P 500.
“The constant curse of scale is that it leads to big, dumb bureaucracy.”
Why do forced divestitures, such as those experienced by Standard Oil and AT&T, tend to result in higher market cap? One reason might be that, post-break-up, not having antitrust investigators constantly at your elbow must help. A second is that for such behemoths, eventually diseconomies of scale kick in, including of the type emphasized above by Charlie Munger. Or, as Harvard economist Michael Jensen rather dryly puts it: “Managers have incentives to cause their firms to grow beyond the optimal size. Growth increases managers’ power by increasing the resources under their control.” So maybe everyone benefits except for the top dogs who lose a bit of their managerial power.
IBM: Profound Impact, Despite the DoJ Dropping the Antitrust Case
Turning to IBM, business historian Alfred Chandler observed that the American software industry itself was arguably a consequence of Big Blue choosing to unbundle software from hardware in 1967 as a result of antitrust scrutiny. Columbia University’s Tim Wu extended Chandler’s observations, asserting that antitrust investigations and trials against IBM from 1967 to 1982 generated massive innovation in storage, processing, printing, modems, and personal computing. Moreover, antitrust commentators such as Matt Stoller3 argue that Microsoft and Intel might not have been able to become significant players without that suit. That is, even unsuccessful suits can unleash innovation and deliver enormous benefits. We now turn to a second case that ended in settlement, that against Microsoft.
Microsoft: The Template for the Case Against Google
Microsoft was accused of illegally maintaining its monopoly position in the PC market, through restrictions put on manufacturers and users who wished to uninstall Internet Explorer and use other browsers such as Netscape and Java. Microsoft countered that bundling its flagship IE web browser software with its Windows operating system was the result of innovation and the two were now the same product and inextricably linked. Moreover, consumers were receiving the benefits of IE for free.
The suit argued that Microsoft’s actions constituted unlawful monopolization under §2 of the Sherman Act. No one questioned whether it dominated the market for operating systems (it did). The legal question was whether Microsoft was using its contracts with PC manufacturers to require them to install IE in a way that excluded potential competitors.
While the suit accused Microsoft of exclusionary, anticompetitive behavior, it was largely an innovation-harm case. However, such harm requires a counter-factual or alternative history, so it is impossible to provide direct empirical evidence. Inevitably, expert economists provide testimony and clever models for both sides, but a black-and-white outcome is inherently improbable. This helps explain why the Microsoft case took three years to try and ended up in settlement, and why the consensus expects a similar denouement in the two recent cases.
“If we had not brought the case against Microsoft there might not have been a Google.”
—Tom Miller, Attorney General of Iowa (the longest-serving state AG in U.S. history, helped bring cases against both Microsoft and Google)
Even though the DoJ did not win the suit, and Microsoft’s share price outperformed during the antitrust case (Figure 1), the proverbial “policeman at the elbow” helped create the necessary space for start-ups including Google to grow. With Microsoft management focused on fighting the antitrust suit, attention was diverted from product development, especially updating its browser and developing its search engine. This allowed Google to steal a march and ultimately come to dominate both products. It seems possible that a similar dynamic will play out today as well, with management at Google and Facebook distracted for at least the next three years.
Unfettered Markets and the “Chicago School” of Antitrust
“For every complex problem there is an answer that is clear, simple, and wrong.”
—H. L. Mencken
The Microsoft suit was filed in 1998 and the U.S. waited more than 20 years before filing another major antitrust case. In fact, the number of antitrust cases dropped dramatically from the 1970s (Figure 2), when America’s regulators became so welcoming that critics have painted them as doormats. This decline reflected the influence of the “Chicago school” of antitrust law, which since the 1970s has successfully argued for a much more hospitable approach to antitrust. The Chicago school reached its zenith in the writing of the legal scholars Robert Bork and Richard Posner, who combined a narrow “consumer welfare” focus with the ideological perspective that government interference in the operation of free markets does more harm than good.
However, the last two decades have witnessed rising concentration in a majority of sectors4, with the most successful technology companies growing into veritable titans. Most economists studying the subject now worry that a lack of competition is an economic drag, and legal scholars are moving away from the Chicago school (see, “The Great Reversal: How America Gave up on Free Markets,” by T. Philippon and “The Curse of Bigness: Antitrust in the New Gilded Age,” by T. Wu). From this perspective the cases against Google and Facebook are long overdue and constitute an important test of the Chicago school. But why has the interest in antitrust surged during the last year or two?
Techlash: The Vast Majority of Americans Believe Digital Platforms Have Too Much Power
“Google and Facebook may be the most powerful political agents of our time. … Google and Facebook have the power of ExxonMobil, the New York Times, JPMorgan Chase, the NRA, and Boeing combined. Furthermore, all this combined power rests in the hands of just three people.”
—“Stigler Committee on Digital Platforms: Policy Brief,” U Chicago
Americans across the political and demographic spectrum staunchly believe political leaders are not paying enough attention to tech issues. The platforms are perceived as having grown too powerful (Figure 3) and Americans now have largely negative views of big tech companies’ impact on society. Moreover, there is a growing perception that tech acquisitions are unfair (Figure 4) and often stifle competition, reduce consumer choice, and hamper innovation. Behind this apprehension is a mind-boggling statistic: Since 1998, the big four platforms collectively purchased more than 550 companies, with the antitrust agencies deigning to not block a single acquisition.
In 2020, For the First Time, More Than Half of Advertising Revenues Accrued to Digital Platforms
“Senator, we run ads.”
—Mark Zuckerberg responds to Senator Orrin Hatch who questioned Facebook’s viability given it’s free to users
For Facebook and Google, their ascent to power has been fueled by advertising revenues. The meteoric rise of digital platforms has come at the expense of traditional players such as TV, newspapers, and radio (Figure 5). The reason for the industry’s vivid transformation is simple: Digital ads allow marketers to better target and measure the performance of their ads, an advantage that keeps growing as consumers spend more of their time and dollars online.
The Tech Industry is Rife with Natural Monopolies
Moreover, the top two companies’ ad revenues have soared, with no signs of slowing down (Figure 6). They accounted for 54% of U.S. digital advertising in 2019, with Google’s share at 31% and Facebook’s at 23%. If Amazon’s rising claim is taken into account, the three firms accounted for more than 70% of online advertising globally. At the firm level, 98% of Facebook’s global revenue was generated from advertising, while the corresponding figure for Google was 71%. At core, the two firms run quite simple businesses. They are attention merchants, developing social media, search, and other services so they can better attract eyeballs, with the intent of monetizing them by selling highly lucrative ads. However, it is now being alleged that anti-competitive behavior was a critical component of building and maintaining their dominant positions.
With the Google Lawsuits, the Long Antitrust Winter is Over
“If we let Google continue its anticompetitive ways, we will lose the next wave of innovators and Americans may never get to benefit from the ‘next Google’.”
—William Barr, U.S. attorney-general
The first suit was filed in October and gives the appearance the DoJ just copied the Microsoft case, added a bit of whiteout, and scribbled in Google’s name. The case alleges a violation of §2 of the Sherman Act, which prohibits maintaining a monopoly through “exclusionary” or anticompetitive conduct. Investors have not panicked though and since the suit was filed, Google has marginally outperformed the SPX, even during a period when many tech names were struggling (Figure 7).
A successful suit will require the DoJ to prove Google (1) has monopoly power and (2) used that power to exclude competitors. The first requirement couldn’t be easier (Figure 8), so the focus will be on Google’s exclusionary behavior. In particular, the complaint alleges Google has struck exclusive deals with Apple, android handset makers and mobile networks to ensure its flagship search engine is the almost universal default option, effectively shutting out rivals like Bing and DuckDuckGo and preventing any emerging competitors from gaining the scale required for success.
“There’s a difference between dominance and excellence.”
—Eric Schmidt, former Google CEO (commenting on the DoJ’s suit)
One focus of the complaint is that the search titan entered into a long-term agreement with Apple that required Google to be the default general search engine on iPhones, and that Apple was paid $8 bn -$12 bn per year for this exclusive arrangement. According to the DoJ, Google’s exclusionary agreements cover more than 80% of all U.S. search queries on mobile devices. This type of “exclusive deal” is a classic antitrust offense and is reminiscent of the conduct that got Microsoft in so much trouble.
What favors Google in the case is the almost universal acknowledgement that their search engine is the best (other than by the privacy-inclined). However, the government may respond that Google is only “the best” because their exclusionary behavior has prevented anyone from getting sufficient search volume to create an engine with similar performance.
“Competition is only a click away.”
—Larry Page, Google
“Monopolists lie to protect themselves … usually by exaggerating the power of their (nonexistent) competition.”
Google will also insist the arrangements aren’t in fact exclusionary because consumers could always install another browser (even if they rarely did). Further, Google will emphasize that, even though it is the world’s dominant search engine, nearly half of Americans start their product searches on Amazon.
In terms of the harm caused by Google’s conduct, the DoJ will argue that a monopoly, if immunized from competitive forces, can behave with impunity. In this case, that behavior might include raising prices (to advertisers), downgrading the quality of its product (with increased advertising), and weakening privacy protections. Regarding remedies, the DoJ’s lawyers emphasized “nothing is off the table.”
The second antitrust case against Google was filed in December and takes broad aim at its monopoly over the online advertising ecosystem. Among other things it alleges Google and Facebook entered into an unlawful agreement to reduce competition in online advertising. More specifically, the suit accuses Google of giving Facebook special privileges in ad auctions in exchange for it promising not to support a system that competes with Google’s ad platform. The agreement, the complaint charges, “fixes prices and allocates markets between Google and Facebook.” Most antitrust cases are tried under §2 of the Sherman Act. The alleged conspiracy between Google and Facebook is different. It falls under §1, which makes it illegal for two or more companies to make any contract or agreement “in restraint of trade.”
“If the free market were a baseball game, Google positioned itself as the pitcher, the batter and the umpire.”
—Ken Paxton, Texas attorney general
More broadly, the complaint accuses Google of taking advantage of its control over each step of the online advertising chain—from the tool businesses employ to place ads, to the platform publishers use to make their ad space available, to the exchange where the two sides meet—in ways that stifle competition and increase the prices paid by advertisers.5 Much of Google’s power as an ad broker stems from acquisitions of ad-technology companies, especially its 2008 purchase of DoubleClick for $3.1 bn, a deal that has since been criticized as central to Google’s dominance. Google now operates the leading tools for both buying and selling advertising, and also runs the biggest marketplace where online ad deals happen. Moreover, much of Google’s ad revenue comes from ads place on its own properties (YouTube, Google Search). It is little wonder that Google’s dominant role has led to calls for regulatory action. The digital ad market was tiny a decade ago, but now its size demands a more transparent, competitive market microstructure.
The third lawsuit was also filed in December and targets Google’s conduct in specialized search services. Google allegedly abuses its gatekeeper role to curtail traffic to specialized search providers that focus on segments like travel (e.g., TripAdvisor), home repairs (e.g., Angie’s List) and local businesses (e.g., Yelp). Those search services rely on Google for customer flow and represent a potential threat to its dominance, particularly because their niche offerings are attractive to advertisers. The suit claims Google has created a system rife with conflicts of interest, in which it abuses its unique dominance in general search to give display preference to its own competing reviews or services.
Dina Srinavasan, now an antitrust specialist at Yale but formerly an ad tech executive, has been involved in drafting one of the latest antitrust cases. She has demonstrated how Google used acquisitions as part of a strategy to dominate in all parts of the digital advertising value chain, squeezing out potential rivals and grabbing an increasing proportion of digital ads (Figure 9).
We’re at the Top of the First Inning of a Very Long Game
The multiple cases against Google will be consolidated into a single lawsuit, presided over by U.S. District Judge Amit Mehta, who suggested a trial date of Sept. 12, 2023. The next status hearing is set for Jan. 21, 2021, with the two sides expecting discovery will be completed in March 2022, with other pretrial matters not addressed until early 2023. In the meantime, both sides must gather documents, do research, file and debate procedural motions, interview executives and stakeholders, and build complex economic models for the trial. The case itself could stretch out, with a remedy phase, and then there will be appeals. When all is said and done, it could easily be five years before there’s a remedy.
Facebook: Buying its Way Out of Competition?
“It is better to buy than compete.”
Moving on from Google, last month the FTC and 46 states filed antitrust suits against Facebook alleging the tech giant engaged in unlawful, anticompetitive tactics to buy or kill off its rivals and solidify its dominance in social networking. Since the case was submitted, Facebook has underperformed the SPX, while Twitter has gained modestly (was up almost 15% before Jan 5-6).
The lawsuit’s central allegation is that Facebook’s purchases of Instagram in 2012 and WhatsApp in 2014 marked a pattern of behavior to neutralize competitive threats, allowing Facebook to dominate the social media landscape (Figure 11).
“It wasn’t legal when John D. Rockefeller did it, and it’s still not legal.”
—Tim Wu, Columbia University
Tim Wu argues Facebook’s strategy was similar, though clearly not of the same scale or ruthlessness, to Rockefeller’s during the 1880s. He asserts both companies scanned the marketplace, searching for potential competitors, and then bought or buried them. That was the essence of Rockefeller’s business model, and the FTC will be arguing it was also Facebook’s MO. As Professor Wu puts it, “No one faults Facebook for gaining its social networking dominance in the first place, beating rivals like Myspace in the 2000s. The trouble is what it did to hold on to the kingdom.”
In addition to anticompetitive acquisitions, the FTC also argues Facebook limited third-party software deve-lopers’ access to its platform, including restricting their access to APIs that allow the developers’ apps to interface with Facebook. The suit alleges Facebook permitted access to key APIs only if the third-party developers refrained from providing similar features for its competitors.
One of the biggest hurdles in the lawsuit is showing that Facebook’s monopoly has been harmful. This is a question that has long frustrated the push for antitrust enforcement against platforms: How do you prove people are being harmed by a product that’s free? The complaint suggests the answer hinges on reduced innovation, fewer choices for consumers, the decline in Facebook’s privacy protections, and higher prices paid by advertisers. That is, Facebook may not charge a fee, but that doesn’t mean users haven’t been paying a price for its dominance.
In “The Antitrust Case Against Facebook,” the legal scholar Dina Srinivasan demonstrates Facebook’s takeover of the social networking market has inflicted a very specific harm on consumers: It has forced them to accept ever worse privacy settings. Her privacy theory has the virtue of being concrete and demonstratable: Facebook really did backslide on privacy commitments as it grew more dominant. Similarly, the higher prices paid by advertisers is also amenable to empirical proof. She concludes that, over the next few years, expert economists should be fully employed!
In cases tried under the Clayton Act of 1914 (specifically §7, which is the principal federal law governing mergers and acquisitions), there is a strong tendency for courts to embrace divestiture. Consistent with that, the FTC is seeking a court injunction that would require the divestiture of Instagram and WhatsApp (and possibly others among its 86 acquisitions). The FTC also wants to prohibit Facebook from imposing anticompetitive conditions on software developers.
“The Instagram you see today is the Instagram that Facebook built, not the app it acquired. When Facebook bought Instagram, it had about 2% of the users it has today, just 13 employees, no revenue and virtually no infrastructure of its own.”
—Jennifer Newstead, Facebook General Counsel
In response, Facebook has mounted a massive lobbying offensive to rebut the allegations, stressing that Facebook’s purchases of Instagram and WhatsApp helped them grow into viable services in a larger market where newcomers, such as TikTok, are still able to thrive.
“A strongly competitive landscape existed at the time of both acquisitions and exists today. Regulators thoroughly reviewed each deal and rightly did not see any reason to stop them at the time.”
—Chris Sgro, Facebook spokesperson
A second point emphasized by Facebook is the federal government did not block its acquisitions of WhatsApp and Instagram at the time. While that narrow statement is true, it is important to note the FTC did not “approve” the deals. It just opted not to try to block them at the time and the FTC explicitly reserved the right to take another, later look at the mergers.
A third argument raised by Facebook is that the three products are no longer distinct entities and breaking them up would be incredibly technically difficult. The products share back-end resources for computer infrastructure, operations, advertising systems, content moderation, and more. Further, as the FTC deliberates, Facebook engineers are working frantically to weave the products together ever more tightly.
This was also part of Microsoft’s defense and the court will certainly take the point into account. However, the FTC will counter that divestitures are always technically difficult, and it’s not obvious this break-up would be any more challenging than Standard Oil’s or AT&T’s. Moreover, Facebook’s user base perceives the three apps as distinct products, providing quite different experiences. A Pew Research survey found only 29% of Americans are aware Instagram and WhatsApp are owned by Facebook.
When Big Isn’t Beautiful: Antitrust Takeaways
At this point, before moving on to a discussion of Section 230, we wanted to highlight a few observations from the antitrust cases discussed above. First, be careful of confident pronouncements regarding how the cases against Google and Facebook will be decided. With so few antitrust precedents, especially for the digital age, we have little appreciation how this will play out.
Second, antitrust moves like molasses, and this has always been the case. Although investors should be encouraged that antitrust enforcement is back after a quiet two decades, the tech behemoths have amassed huge war chests for legal defenses and such complex cases typically take 4+ years to play out.
“Antitrust oxygenates the marketplace.”
—Scott Galloway, NYU
Third, antitrust cases act as a catalyst for innovation and often the creation of entirely new sectors. That was certainly the case with the suits against Standard Oil, A&T, and IBM. More recently, the DoJ suit against Microsoft opened a window for a new wave of innovative tech companies – including Google. Similarly, we expect the current cases to boost innovation and result in a broadening of tech leadership.
The “Policeman at the Elbow” Theory of Antitrust Enforcement
Fourth, experience suggests even unsuccessful antitrust cases have a dramatic effect on behavior, and not just of the targeted company. Legal scholar Tim Wu stresses there is a lot of precedent for what he calls “The Policeman at the Elbow theory of antitrust enforcement.” In the late-1990s, Microsoft executives became far more cautious and paid less attention to product innovation, creating an opening for Google in both browsers and search. Similarly, with IBM, which unbundled software in the 1960s partially because of an antitrust case, thus enabling the creation of the modern software industry. Regardless of the outcome of the cases against Google and Facebook, big-tech firms have been put on notice they are being closely watched for anti-competitive behavior.
Next, many commentators assert that current antitrust legislation is outdated. The key laws were written more than 100 years ago (the Sherman Act of 1890 and the Clayton Act of 1914). Further, as we have argued in a series of “Tech is the new macro” papers, the economics of digital platforms is fundamentally and profoundly different than that of asset-heavy, bricks-and-mortar companies. To illustrate, special challenges arise in demonstrating anticompetitive harm when the product is “free.” This requires proving some combination of less innovation, reduced consumer choice, weaker privacy safeguard and higher advertising prices. While it is easy to argue these points conceptually, it is no easy task to provide hard data (as it requires a counter-factual that doesn’t exist). This suggests some aspects of U.S. antitrust legislation might need to be rewritten for the digital economy.
Any legislative updates could also include enhanced merger review to prevent anticompetitive acquisitions. Since 1998, Facebook, Google, Amazon, and Apple collectively have purchased more than 550 companies, and the antitrust agencies did not block a single acquisition. The House Judiciary Committee’s 2020 report, “Competition in Digital Markets”, provides a full listing of Facebook’s 86 acquisitions, as well as the 256 made by Google, 104 by Amazon and 120 by Apple.
Section 230: The Twenty-Six Words that Created the Internet
We now move on from antitrust, to explain why § 230 is such a lightning rod, why reform is so difficult and what is most likely to happen. § 230 of the Communications Decency Act of 1996 is a statute whose tame title belies the weighty protections it provides to the tech industry. The bulk of § 230’s impact comes from a single sentence: “No provider or user of an interactive computer service shall be treated as the publisher or speaker of any information provided by another information content provider.”
Under U.S. law, you can sue the publisher of certain kinds of content, but you can’t sue a distributor. Think of a book that contains horrible, defamatory lies about you. You can take the publisher that printed the book to court, but you can’t sue every bookstore that sold it. § 230 says social media companies like Facebook and Twitter fall into the distributor category.
The immediate catalyst for writing § 230, was an anonymous message posted in 1995 on Prodigy Services6, that accused Stratton Oakmont (the scammy investment firm immortalized in The Wolf of Wall Street) of fraud. A defamation suit followed, with the boiler-room broker seeking $200 million in damages. The New York Supreme Court ruled against Prodigy, arguing that because it moderated its message board, it was legally liable for whatever users posted. That is, if a website moderates user content, it is making editorial decisions, and should be treated like a publisher rather than a distributor.
The court’s logic created a perverse incentive: Legally, it was less risky to let users post any awful thing they wanted, because once you started moderating content you could get sued. In response, during the following year Congressmen Ron Wyden [D-OR] and Chris Cox [R-CA] inserted § 230 into the Communications Decency Act to ensure websites can moderate content without getting sued.
Absent § 230’s protection, platforms would face an economically crippling duty to review the prodigious volume of data to ensure its users’ posts didn’t contain defamatory speech or other unlawful content. Online platforms would be compelled to heavily censor user speech or disallow posting altogether to avoid the risk of liability. However, fast-forward to the present and a consensus has developed that the blanket immunity passed twenty-five years ago requires a comprehensive reevaluation. § 230 now presides over a vastly different internet from the one it was designed to govern.
With Great Power Comes Great Controversy: Social Media’s Struggle with Self-Censorship
“Facebook shouldn’t be the arbiter of truth of everything that people say online.”
Social media firms have become more active at moderating content, albeit kicking and screaming. Mark Zuckerberg does not want to be an “arbiter of truth,” but that is precisely what he has become. And as such, social media networks have edged toward being proactive publishers rather than passive distributors.
During the last two years, Facebook has been on the mother of all clean-ups (Figure 12). In addition to hate speech, Facebook removes posts related to bullying and harassment (averaging 3.0 mn posts per quarter), suicide and self-harm (2.3 mn), terrorism (8.1 mn), drugs (6.8 mn), child nudity (11.0 mn), violent and graphic content (2.9 mn) and porn (37.7 mn). Moreover, it disables some 17m fake accounts every single day, more than twice the number three years ago. These statistics are just for Facebook and don’t include Instagram or any of its other sites. Further, YouTube, Twitter and others face similar challenges.
“Software ate the world. So all the world’s problems get expressed in software. (We connected everyone, including the bad people, and our bad instincts.)”
While much of the offending content can be removed by AI, some categories like bullying require human judgement and almost artisan skill levels. As a result, Facebook now employs about 15,000 people to moderate content. To illustrate how rough a job this is, in May the company agreed to pay a total of $52 mn to 11,250 moderators who developed post-traumatic stress disorder from looking at the worst of the internet.
“With so many internal forces propping up the production of hateful and violent content, the task of stopping hate and violence on Facebook starts to feel even more Sisyphean.”
—Departing post from a Facebook data scientist
A data scientist who departed Facebook after a two-year stint on a team focused on “Violence and Incitement,” estimated that roughly 5 mn of the 5 bn pieces of content posted daily violates the company’s rules on hate speech. More stunningly, they estimated Facebook was “deleting less than 5% of all of the hate speech posted to Facebook.”
Section 230 and its Critics: Bipartisan Calls for Reform
—@realDonaldTrump, May 29 (in reaction to Twitter fact-checking two of his tweets)
Public anger is growing more intense and calls for § 230 reform have come from every corner. Democrats criticize online platforms’ failure to protect the public, reasoning that they have a responsibility to identify and limit the spread of hate speech, materials promoting terrorism, and other harmful material. On the other hand, President Trump and Republicans criticize media platforms for their perceived bias, alleging platforms’ content-censorship practices systematically silence conservative voices (Figure 13).
Rebooting Internet Immunity: Taking a Scalpel or a Hammer to Section 230?
“The network platforms have been left blissfully unmolested by the FCC, even as they have moved into direct competition with the TV and radio stations that it regulated.”
—Niall Ferguson, Stanford
Politicians are openly debating what to do and in 2020 there were five separate bills proposed in Congress to curtail § 230. As Columbia’s Tim Wu emphasizes, the intuitive case for abolishing or weakening § 230 immunity is that traditional media companies such as Newspapers and TV stations are fully responsible for what they publish or broadcast. Why should Facebook and Twitter receive special treatment, especially given they actively recommend content and monetize it?
However, Professor Wu also warns that curtailing § 230 could hurt smaller platforms and startups more than larger ones, because the big platforms have the resources to develop the necessary AI and to pay for all the content moderators and compliance staff. Smaller platforms and startups don’t have access to these resources, so a few nasty lawsuits could kill them. In this way, § 230 repeal might even further insulate the big platforms from competition.
Regardless, trust in social media platforms is extremely low (Figure 14), suggesting the status quo is untenable. While many politicians have made rash calls to simply repeal § 230, there are three basic approaches to reining in § 230 that are likely to make progress. The first is something like a version of the “fairness doctrine” but adapted for social media. For example, under the proposed Ending Support for Internet Censorship Act (sponsored by Senator Josh Hawley [R-MO]), companies would only be protected from lawsuits if they could convince the FTC they’re politically neutral. This would partly satisfy the concerns of conservatives, and some social media platforms are already moving in that direction. In October Twitter’s CEO Jack Dorsey suggested modifying § 230 to make platforms’ moderation processes more transparent and developing clear ways for users to appeal their decisions.
The second approach concerns “carve-outs” in which legislators would make broad categories of content exempt from liability protections. This would mean firms could not avoid responsibility for copyright infringements, posts that break federal criminal law, or which enable sex trafficking. The blueprint is the Fight Online Sex Trafficking Act (FOSTA), which Congress passed in 2018 and states that § 230 protections don’t apply when a website gets sued for hosting content related to sex trafficking.
The third approach would only apply to social media firms, like WhatsApp, that limit themselves to solely transmitting information, with no editing or promotion. They could be treated like phone companies and be exempted from content liability. As the University of Chicago’s Committee on Digital Platforms emphasizes, such a system would go a long way in levelling the playing field between digital platforms and traditional media.
Does the Meteoric Rise of Digital Platforms Require a Specialized Regulatory Agency?
“Each wave of tech changes the world, and gets regulated. Every important industry gets industry-specific regulation.”
In addition to more aggressive antitrust enforcement and reform of § 230, digital platforms face the inevitability of regulation. Creating a new agency is how the government has responded to all major technological advances over the last 150 years (e.g., railways, radio, electricity, autos, airlines, TV, nuclear power, telecoms, pharma). Regulators focused on sectors requiring specialized expertise include the Federal Aviation Administration, National Highway Traffic Safety Administration, Food and Drug Administration, Nuclear Regulatory Commission, and Federal Communications Commission. Today’s dominant platforms will not prove an exception.
To illustrate this point, we will briefly discuss two examples, railways in the 19th century and radio in the 20th. The railway boom began in the U.S. in the 1820s, but it wasn’t until 1860 that rail networks connected nearly every major city in the North and Midwest. Anti-railroad agitation was widespread, with western farmers the dominant force behind the unrest, believing the railroads abused their economic power and systematically over-charged them. There were also longstanding concerns about safety. However, it wasn’t until 1887, after decades of deliberation, that the U.S. Interstate Commerce Commission was established. Its purpose was to ensure fair rates, eliminate rate discrimination and, from 1893, oversee railroad safety.
Turning to radio, the first program broadcast occurred on Christmas Eve 1906, with the initial federal law to license radio broadcast occurring six years later (in response to delayed emergency signals during the sinking of the Titanic). Almost a decade later, the first news program hit the air and 1920 was the year radio really took off. After much debate, the Federal Radio Commission was established in 1927 with the power to grant licenses and to assign frequencies and power levels for each licensee. However, it wasn’t until 1934 that FDR and Congress created the Federal Communications Commission to provide a single regulator for telephone, telegraph, and radio.
Digital platforms are the railroads and radio of the 21st century, occupying a similarly important position in society and raising many of the same policy challenges that telephone and radio broadcasting created in their day. Tech regulation is inevitable, but creating a new regulator often follows decades of fractious debate and deliberation.7
Regulation: Meeting the Demands of the Digital Economy
“The internet is the world’s largest ungoverned space.”
—Google CEO Eric Schmidt, 2013
“Despite earnest efforts, the tech community has not demonstrated convincingly that it can regulate itself.”
—Former Google CEO Eric Schmidt, 2020
Whether it’s railways, radio or digital platforms, regulatory policy deals with critical questions that tend to be full of complexities and trade-offs. That’s how policy always works, and tech regulation certainly won’t be an exception. Slogans, executive orders, and litigation will not suffice. Rather, what is required is thoughtful legislation and a regulator that possesses significant expertise regarding digital platforms and their business models.
According to a recent survey, 53% of American “expert economists” agree on the need for such a regulator, with only 21% disagreeing (Figure 15). On the continent, where tech regulation has been more prominent, almost 90% of their European counterparts agree.
Recent studies out of Chicago, Harvard, and Stanford8 suggest the broad objective of a digital markets regulator should be the protection of both consumers and competition. To achieve this, they recommend enforcing a set of basic rules of conduct, such as disallowing anticompetitive acquisitions and exclusive behavior, and ensuring platforms don’t favor their own services. To protect consumers and give them more control over their data, the regulator could improve privacy safeguards and require personal data portability.
This raises the question of whether we have reached the point where a majority of Americans support the establishment of a digital markets regulator. Most Democrats support a larger government role to address a host of tech issues while Republicans are less enthusiastic about an increased bureaucratic presence (Figure 16).
“The final scope is unclear but, tech is becoming a regulated industry … Expect Google and Facebook to have armies of compliance people in 5 years – just like banks or telcos.”
There are two main risks associated with the creation of a digital markets agency. One is regulatory overkill, including the possibility that bureaucratic meddling in a fast-moving industry could hobble innovation. However, this seems unlikely given the army of highly paid lobbyists tech has amassed in DC. The bigger concern is regulatory capture, which would entrench the leadership of tech incumbents, becoming a new barrier to entry rather than a promoter of competition, shackle innovation and impede the next generation of upstarts. Both are real concerns that all stakeholders need to be vigilant of.
Regardless, under a Biden presidency we could well see a new agency established to monitor and regulate big tech. Congress would need to draft and pass legislation, preferably with strong bipartisan support. There are many open questions (e.g., should the agency be housed in an existing administrative body like the FTC or FCC), so this will not happen quickly, but it will, in all likelihood, happen.
The Democrat Trifecta: Tech Policy Will Change Under Biden, But Don’t Expect Revolution
“This will be a “put up or shut up” moment for Democrats, no longer blocked by a Republican majority, to decide what tech policies they want to push to the finish line.”
—Ina Fried, Axios Login, Jan 7, 2021
On antitrust matters, the incoming Biden administration isn’t going to take its foot off the gas. There will likely be additional antitrust cases filed, but that would have happened regardless of who was sitting in the White House.
In politics, “personnel” is policy, and it is certainly not a coincidence that Biden’s nominee for Attorney General, Merrick Garland, has extensive experience in antitrust. Early in his career he taught Advanced Antitrust at his alma mater, Harvard Law School and also published articles in the Yale Law Journal and Harvard Law Review on antitrust issues. Since his appointment to the U.S. Court of Appeals in 1997, Judge Garland has been on several panels deciding antitrust cases and has proven to be a balanced jurist and not a judicial activist. According to a 2016 analysis by two partners at Skadden, “His votes on this variety of antitrust issues bolster the view that he is a measured jurist, whose decisions are uninfluenced by ideology.”
To rein in big tech, Biden will need to reach across the aisle, as there remains profound disagreement regarding which policies to prioritize. For example, establishing a digital markets regulator is likely to invite fervent GOP opposition. There are, however, at least five initiatives that possess significant support from both parties: (1) More funding for the FTC and DoJ, so they can take on wealthy, heavily lawyered tech companies; (2) Changing the burden of proof, so the onus is on big tech to demonstrate acquisitions will not harm competition; (3) Ensuring a level playing field, so platforms can’t favor their own listings above those of competitors; (4) Reform of Section 230 (three approaches are outlined above); and (5) Data privacy and portability rules that give consumers more control.
Data Privacy: A “Miranda rights” for Digital Consumers
Regarding this last point, lawmakers from both parties have introduced bills to give consumers more control over the personal data that technology companies collect, analyze, and monetize. One promising approach is provided by Senator Maria Cantwell’s [D-WA] Consumer Online Privacy Rights Act. COPRA would give users the right to see and delete any personal information that companies have amassed about them and require tech companies to clearly explain what they are doing with users’ data. COPRA is already co-sponsored by senators Klobuchar [D-MN] and Ed Markey [D-MA] and an updated COPRA could have legs—especially if Biden throws his weight behind it.
Investment Implications: Tech Will Continue to be the Most Dynamic Sector
“I think the idea that innovation is slowing down is one of the stupidest things anybody ever said.”
For investors, antitrust has historically represented more of an opportunity than a threat. Major antitrust cases, including Standard Oil, AT&T, IBM, and Microsoft, had a positive impact on competition and unleashed a wave of innovation. Typically, even shareholders of the targeted firm have benefited.
Moreover, historical precedents suggest the recent antitrust cases will unleash a surge of new platform companies challenging the middle-age incumbents. Until recently, the trend has been toward the consolidation into a few mega-platforms that bundle together a wide variety of services. We expect this trend to slow and possibly reverse due to regulatory backlash. While this may limit the growth of mega-platforms, it will likely foster the emergence of numerous, highly innovative single-product platforms.
This suggests greater breadth in tech market leadership and the emergence of entirely new sub-sectors. Finally, we expect tech will continue to be the most dynamic sector of the economy, with digital platforms constituting a sizeable majority of market cap by the end of the decade.
1. Tunstall, B., “Disconnecting Parties,” 1985.
2. Hovenkamp, H. “Antitrust and Platform Monopoly,” Yale Law Journal, 2021.
3. Author of “Goliath: The 100-Year War Between Monopoly Power and Democracy”
4. Reflecting the impact of globalization and technology, as well as lax antitrust enforcement
5. For an excellent description of the digital ad market, see “How Google Edged Out Rivals and Built the World’s Dominant Ad Machine: A Visual Guide,” WSJ, Nov 9, 2019.
6. At that time Prodigy was the third largest online service and peaked with 3.1 mn subscribers
7. See “The Case for the Digital Platform Act: Breakups, Starfish Problems, & Tech Regulation,” by Harold Feld, 2019.
8. “Stigler Committee on Digital Platforms: Policy Brief,” “The Case for a Digital Platform Agency and a New Approach to Regulatory Oversight,” and “Report of the Working Group on Platform Scale,” respectively.