In just two decades, digital technology and the internet have gone from exciting the dreams of a revolutionary new era to embodying fears about a world gone deeply wrong. The digital revolution now threatens to undermine values that it was supposed to advance—personal freedom, democracy, trustworthy knowledge, even open competition. It isn’t as though the technology did this to us on its own, or that we stumbled absentmindedly into an alternative dystopian universe. Today’s technological regime grew out of critical choices to ignore lessons of the past and allow private power to go unregulated.
Three problems—monopoly, surveillance, and disinformation—sum up what’s gone wrong and what needs to be fought and fixed in order to have any hope of recovering the promise of the new technology.
The explosive growth of the online economy in the 1990s and early 2000s appeared to validate the idea that markets were best left to themselves. The internet of that era was neoliberalism’s greatest triumph. After the federal government financed key breakthroughs and then opened the internet to commercial development, digital innovation and entrepreneurship created new online means of exchange, new wealth, and new communities. But that online economy now looks altogether different with the rise of platform monopolies. Amazon, Facebook, Google, Apple, and Microsoft control whole ecosystems of the digital world, dominating key choke points for commerce and news.
Just as the early internet fostered the illusion that it was inherently supportive of competition, so it fostered the illusion that it was inherently protective of personal autonomy. After all, no one compelled you to disclose your true identity online. Yet the digital world today has made possible the most comprehensive system of surveillance ever created; networked devices track our every movement and communication. A new form of enterprise has emerged that Shoshana Zuboff calls “surveillance capitalism,” as Google, Facebook, and other firms sweep up data about our lives, preferences, personalities, and emotions “for hidden commercial practices of extraction, prediction, and sales.”
The digital revolution began under the state, moved to the market, and now illustrates what goes wrong when the dominant players are unrestrained by law.
The reality has turned out to be less benign. The online economy has destroyed the traditional business model of journalism, resulting in a dramatic decline in professional reporting. And because Google and Facebook dominate digital advertising, no alternative online model has emerged capable of financing the same reporting capacities, particularly at the regional and local level. Meanwhile, social media platforms have replaced the old mass media gatekeepers, shaping the public’s exposure to news and debate through their algorithms. Those algorithms—for example, in Facebook’s news feed, Google’s search and YouTube recommendation engine, and Twitter’s trending topics—now influence which content and viewpoints gain visibility among users. Instead of promoting better-informed public debate, however, social media have become powerful vectors of disinformation, polarization, and hatred.
These problems afflict other countries as well as the United States, but the digital economy has been a distinctly American development. How we arrived at the present crisis and what we can do about it have become urgent political questions.
The Neoliberal Turn
The growth of the internet and digital economy has been a paradigmatic story of American technology since the Cold War. The digital revolution began under the aegis of the state, moved to the market, and has now become an illustration of all that has gone wrong when the dominant players in markets are unrestrained by law.
From the 1940s to the early 1970s, the federal government financed and guided most of the development of computers and electronic communication, largely via the Defense Department. It was a Defense office, the Advanced Research Projects Agency, that funded and supervised the creation of ARPANET, the forerunner to the internet.
During the mid-20th century, the United States also extensively regulated telecommunications. While AT&T had an effective monopoly, regulatory policies constrained telephone rates, promoted universal service, and barred discrimination, requiring telecom companies to act as common carriers. Because of its treatment of capital investment, the regulatory system gave AT&T an incentive to devote ample funds to research, and its research arm, Bell Labs, produced an extraordinary array of advances, including data networking, the transistor, the laser, and cellular telephony. Bell’s advances were subject to compulsory licensing, which meant they were available for others to build on.
Thanks to this mixed economy, the computing and telecommunications industries in the United States developed an enormous lead over their counterparts in other countries, the source of America’s head start and comparative advantage in digital innovation. Later on, the myth would develop that individual geniuses working in garages gave us computers and the internet. But their work would not have been possible without the investments and technical advances that the federal government and regulated telecom industry had already made.
Nonetheless, the telecom regulatory regime had a serious downside. It gave AT&T the power to control every aspect of the telephone network, including what devices could connect to it. Like any monopoly, AT&T sought to protect its privileged position. After the Defense Department received a proposal in 1964 for a communication network similar to the internet, an AT&T executive said, “Damned if we are going to allow the creation of a competitor to ourselves.” The law at that point was on AT&T’s side.
The federal government’s involvement in computers and telecommunications began to decline, however, in the late 1970s and 1980s, coinciding with the general neoliberal shift in national policy. Here I am using the term “neoliberal” specifically to refer to ideas and policies that seek to create markets and rely on market forces. The neoliberal arsenal includes such measures as privatization, free-trade agreements, deregulation, tax cuts, and reductions in social spending. What distinguishes neoliberalism from 19th-century laissez-faire is that it has arisen after a period of liberal and social democratic state-building. In the standard history, neoliberalism owes its philosophical origins to Friedrich Hayek and his circle in the 1940s and emerged as a political force with Margaret Thatcher and Ronald Reagan in the late 1970s and 1980s.
Politically, however, deregulation differed from other market-oriented measures. Some forms of deregulation attracted support from prominent liberals and progressives—including Senator Ted Kennedy and Ralph Nader—on the grounds that the regulatory agencies had been captured by the industries they were supposed to regulate and were no longer serving the public. Limiting AT&T’s power enjoyed support across the ideological spectrum.
Although the big step would come in 1984, when a court broke up AT&T, the federal government had already begun loosening the telephone monopoly by that time. In cases in 1956 and 1968, federal authorities reduced AT&T’s control over devices attached to the telephone network. These early deregulatory steps, together with the subsequent development of microcomputers, opened the way for consumers and businesses in the late 1970s and ’80s to obtain modems (enabling computers to talk to each other) and dial-up access to early online bulletin boards and proprietary networks.
Those proprietary networks—the “big three” were CompuServe, the original leader in the 1980s; America Online (AOL), which grew rapidly in the early ’90s; and Prodigy—looked for a time as though they would dominate “computer information services.” Each company had its own distinct news sources, discussion groups, email systems, and rules (for example, AOL restricted its forums to no more than 23 people, effectively limiting the reach of any individual user). At the time, it was by no means obvious that the internet—originally limited to government use, then expanded to universities and research institutes—would emerge as the framework for electronic communication.
Greg Gibson/AP Images
For Al Gore and Bill Clinton, a new “information superhighway” symbolized generational change. They steered clear of any regulation of the emerging online economy.
But the internet had a more open architecture, including a design principle known as “end-to-end” that distinguished it from other networks. As Lawrence Lessig explained it in these pages in 2001, “This model regulates where ‘intelligence’ in a network is placed. It counsels that intelligence be placed in the applications”—that is, among the users at either end of the network, rather than in the center with the network manager. As a result, unlike the proprietary networks, the internet was permission-free: It invited and decentralized innovation.
Opening up the internet to wider access, including commercial development, was therefore simultaneously a move toward the market and away from proprietary control of the network itself. That was what happened in the first half of the 1990s, when rules against commercial use of the internet were dropped, the internet “backbone” was privatized, and a host of new applications were created, including the World Wide Web. Much of this new software was also developed on a nonproprietary, open-source basis, though that did not reflect any legal or technical requirement (indeed, a new architecture of control could be built on top of the internet, which is what the online platforms would eventually do).
The politics of internet policy in the 1990s were less ideological than generational. In 1992, when Bill Clinton and Al Gore championed the development of a new “information superhighway,” they were signaling a generational change in national leadership. As a senator, Gore had done more than anyone else in national politics to open up and expand the internet.
National policy during the Clinton administration steered clear of any regulation of the emerging online economy. Internet service providers were even subsidized by being exempted from network access charges, and internet intermediaries received broad immunity from liability for user-generated content under Section 230 of the Communications Decency Act, adopted as part of general telecommunications legislation in 1996. The best approach to policy, according to William Kennard, a chairman of the Federal Communications Commission (FCC) appointed by Clinton, was to allow the “marketplace to find business solutions … as an alternative to intervention by government.”
That general attitude continued to dominate policy toward the internet under George W. Bush and Barack Obama. While Republicans opposed a broader role for government in general, Democrats opposed it for the internet and the tech industry in particular. The internet had flourished seemingly without government: Why risk endangering it? Those early beliefs about the internet—in line with general neoliberal premises at the time—left both policymakers and the public unprepared for the rise of platform monopolies and surveillance capitalism, the devastation to professional journalism, and the use of digital communications for disinformation.
If the internet had emerged in a different period, it might have developed differently. But the online economy has developed in an era when the three chief means of keeping corporate power in check—antitrust, economic regulation, and public ownership—have all been in retreat. The federal government did bring one important antitrust suit when it sought to break up Microsoft during the 1990s, a case that finally ended with a consent decree in 2002 and probably prevented Microsoft from squashing Google in its infancy.
Since then, however, the government has raised no obstacles as online platform companies have expanded, bought out potential rivals, and gained monopoly power. Nor has the government raised any obstacles to the platforms’ accumulation of personal data; unlike the European Union, Congress has enacted no general legislation protecting consumer privacy online. And despite the collapse of journalism, America has been unwilling to consider the level of support for public media widely accepted in many other liberal democracies. Only in the last few years has serious attention focused on changes needed to deal with the new concentrations of unaccountable power.
Taking On Monopoly
This is not the first time that a communications revolution has seen a rapid turn from wide-open competition to concentrated control. The same thing happened with the telegraph between the 1840s and 1860s, when Western Union gained a monopoly. It happened again with the telephone between the mid-1890s and 1910s, when AT&T took over the industry. And it happened a third time with radio from the early to the late 1920s, when NBC and CBS became the dominant national networks. That monopolies would arise yet again should have surprised nobody. Although the internet changed many things, it did not change the tendency toward monopoly in network communications.
The internet’s effect on economic concentration, however, may be even greater than the effects of the earlier communications media. Today, Amazon alone has nearly half of online sales, Google and Facebook are taking virtually all the growth in digital advertising, and venture capitalists hesitate to fund some new startups because the big tech companies can so easily drive them under. Instead of diffusing wealth, the digital revolution has been concentrating it in a few giant tech firms and their shareholders.
Antitrust and regulatory policies might have limited the growth of monopolies and abuses of market power. But since the 1980s, the federal government has greatly relaxed antitrust enforcement against big corporations, thanks to the influence of theories holding that corporate dominance of a market is no problem if it improves “consumer welfare,” interpreted largely to mean lower consumer prices. That criterion has made it difficult to prosecute antitrust cases against companies like Facebook and Google, which rely on advertising and user-generated content and charge consumers nothing, or against Amazon, which has sacrificed profits for market domination. Federal authorities have waved through mergers such as Facebook’s acquisition of Instagram and WhatsApp, even though those mergers reduce consumers’ leverage in the marketplace (for example, on privacy policies), and reinforce the monopoly power of the platform giants.
During the struggles against monopoly power in the railroads and other industries in the late 19th and early 20th centuries, Congress and the courts took steps not only against horizontal mergers but also against predatory pricing (cutting prices below costs to drive out competitors), price discrimination (varying prices to individual buyers or sellers according to their characteristics or circumstances), and vertical integration (combining stages of production or distribution normally operated as separate businesses).
In recent decades, however, the courts have tended to dismiss these concerns, failing to anticipate the new potential for monopoly power in digital platforms, which benefit from network effects. The larger a network grows, the more valuable it becomes to each participant it connects and, conversely, the greater the cost of being excluded. As that cost of exclusion rises, so does the market power of a platform company. In the digital world, scale also brings the capacity to extract data from users to train systems of machine intelligence; only the largest companies can compete effectively.
Platform monopolies pose a new threat. Because of the high cost of exclusion from its marketplace, Amazon has other sellers at its mercy.
A platform market like Amazon, Lina Khan argues in the Yale Law Journal, has clear incentives to pursue growth over short-term profits, a strategy rewarded by investors that makes predatory pricing “highly rational—even as existing doctrine treats it as irrational and therefore implausible.” Amazon’s sky-high market capitalization testifies to that logic. (Enjoy that free delivery for Amazon Prime while you can!) The size of Amazon Marketplace makes it essential for other sellers, even though by participating in it, they provide Amazon with critical data, which it sometimes uses to swoop in and undersell them with its own branded versions of their most lucrative products. As a result of the high cost of exclusion from its marketplace, Amazon has other sellers at its mercy and can impose onerous terms on them.
So how to limit the power of market-dominant platforms? Senator Elizabeth Warren proposes to do two things. As president, she would appoint “regulators committed to reversing illegal and anti-competitive tech mergers” like Facebook’s acquisitions of WhatsApp and Instagram, Google’s acquisitions of Waze and DoubleClick, and Amazon’s acquisitions of Whole Foods and Zappos. And she would seek legislation requiring the biggest tech platforms—those with global revenues of $25 billion or more—to be designated as “platform utilities” and kept “structurally separate” from businesses participating in the market. Amazon Marketplace, for example, would have to operate separately from Amazon’s own sales. All platforms, above or below the $25 billion level, would have to meet a standard of “fair, reasonable, and nondiscriminatory dealing with users.”
Antitrust investigations of big tech are already under way. The Justice Department Antitrust Division is investigating Apple and Google, while the Federal Trade Commission (FTC) is handling Facebook and Amazon. State attorneys general are investigating Google and Facebook. But even under a President Warren, breaking up big tech might be difficult. Historically, such cases have typically dragged on for years, and the odds of success today are probably not high in view of prevailing judicial doctrine. Federal authorities may also be reluctant to break up American tech companies at a time when weakening them may indirectly strengthen their Chinese rivals.
That’s not to say the kind of structural change Warren calls for should be off the table. Reducing the market power of platform monopolies may get support from substantial segments of business, not just consumer groups. A new administration hoping to curb predatory pricing and other abuses of market power may have more success, however, going directly after those abuses through new legislation or existing regulatory authority. In fact, new federal legislation may now be likely in one area—privacy rights—though possibly not for good reasons.
The Challenge of Surveillance Capitalism
The United States has long been a laggard on data privacy. American law treats privacy mainly as an aspect of individual liberty threatened by the state, not by the market. Pressure from corporate interests has limited legal protections of personal data collected privately and exchanged in the marketplace. While federal law regulates data in some areas like credit reporting and health care, even those statutes mostly predate the rise of the internet. As a result, consumer privacy online has largely depended on corporate self-regulation, and the FTC has been able to act on privacy abuses chiefly when companies have violated their own privacy policies or otherwise deceived consumers.
In contrast, European law has conceived of privacy as an aspect of human dignity threatened by intrusions of all kinds, private as well as governmental. Consistent with that tradition, the European Union has had stronger consumer data protections, a pattern that culminated in 2016 in the EU’s adoption of the General Data Protection Regulation (GDPR). At least on its face, the GDPR requires companies to give consumers the right to control data gathered about them, potentially upending the business model that has turned Google and Facebook into behemoths.
But the politics of data privacy may now have changed in the United States. The Cambridge Analytica scandal and a seemingly endless stream of disclosures about Facebook have helped raise awareness about the limits of corporate self-regulation. Last year, California adopted major new consumer privacy legislation, and several other states are considering bills. After long resisting a federal privacy law, corporate interests are now pushing for it as a way of preempting stronger state measures.
While scandals have provoked public outrage, the financial interests at stake have never been greater. Personal data is now the lifeblood of giant tech firms. Google and Facebook aggregate data from their own sites and other sources on individuals regardless of whether they have accounts with the companies. With that constant and comprehensive stream of data, they are able to target advertising more precisely and efficiently than the mass media.
Some of this “behavioral advertising” is harmless, and indeed many people may even like getting ads that are more “personalized.” But personalization has rich possibilities for exploitation. In a guide for Australian and New Zealand advertisers, Facebook noted, “By monitoring posts, pictures, interactions, and Internet activity, Facebook can work out when young people feel ‘stressed,’ ‘defeated,’ ‘overwhelmed,’ ‘anxious,’ ‘nervous,’ ‘stupid,’ ‘silly,’ ‘useless,’ and a ‘failure’”—in short, the moments when they are “most vulnerable to a specific configuration of advertising cues and nudges.” A Wall Street Journal investigation of apps that send Facebook highly personal information found that one of them—with 25 million downloads—provided Facebook with estimates of when women were ovulating. Tracking users and stockpiling sensitive data about them inevitably risks abuses. (Facebook disowned responsibility for both the advertiser guide and the app once they were exposed.)
One of the things that companies want most, as Phillip Longman explains in the Washington Monthly, is to know the maximum price you’ll pay for whatever they’re selling, whether it’s a ride on Uber or a product on Amazon—and “it’s getting harder and harder to avoid dealing with marketers who can estimate with increasing accuracy just how much you, personally, are willing to pay, and who charge you accordingly.”
Jenny Kane/AP Images
Mark Zuckerberg say he is very sorry about divulging data to Cambridge Analytica. Notwithstanding such apologies, Facebook’s many scandals have helped clarify the limits of corporate self-regulation.
In her book The Age of Surveillance Capitalism, Zuboff points out that the surveillance business is fundamentally about creating “prediction products,” which are valuable not just to advertisers who want better predictions of what you’ll buy and how much you’ll pay. Insurance companies, landlords, and employers are also interested in more personalized predictions about potential policyholders, tenants, and workers.
Moreover, surveillance capitalism has now moved from the virtual world into the physical one as our phones, apps, and networked devices in the “internet of things” report back to the data companies where we are and what we are doing. The companies, Zuboff writes, want to “nudge, tune, herd, manipulate, and modify behavior in specific directions by executing actions as subtle as inserting a specific phrase into your Facebook news feed, timing the appearance of a BUY button on your phone, or shutting down your car engine when an insurance payment is late.” If unregulated, the new technology has an awesome potential for a new social regime, operated in the interests of the dominant companies.
Both the European GDPR and California’s new privacy legislation aim to give consumers more control of their information, but they do it in different ways. The GDPR requires an opt-in, whereas the California law requires companies to give consumers an opt-out. Under the GDPR, companies must get individuals’ consent for their data practices in a “freely given, specific, plainly worded, and unambiguous” agreement. Among the rights given to consumers are a right of access to information collected about them; a right of “erasure” of data under certain circumstances (more qualified than the earlier “right to be forgotten”); and a “right of explanation” to algorithmic decisions about them. What these rights mean in practice will depend on regulatory and judicial decisions.
The early results since the GDPR went into effect in 2018 have not been encouraging. Complying with the GDPR requires a significant investment. Those costs are not a problem for big tech companies like Google, Facebook, and Amazon, which have been able to get consumers to click on consent forms, but the costs are a problem for small firms and startups. The GDPR could therefore end up strengthening the platform monopolies.
In addition, the GDPR delegates regulatory enforcement to the country where a company has its principal facilities and “data controller.” For the major American companies, that country is Ireland, which has long catered to the tech industry it is now supposed to regulate. Irish officials are now reaching closure on some major enforcement cases, and critics are skeptical about how rigorously they will interpret the GDPR. One key question is whether Facebook can require its users to consent to its data practices as a condition for using Facebook.
Behavioral advertising may seem harmless, and some people may even like getting “personalized” ads. But it has rich possibilities for exploitation.
The California Consumer Privacy Act (CCPA), which goes into effect in January 2020, is not as broad as the GDPR and does not require consumer consent for companies to collect personal data. Instead, the CCPA requires large firms to post in a “clear and conspicuous” place a simple and specific opt-out: “Do Not Sell My Personal Information.” Since the law defines large firms as those that have annual revenues over $25 million or personal data on 50,000 or more individuals, it is unlikely to cause the perverse effects of the GDPR in strengthening big tech. The CCPA also explicitly gives consumers a right to receive equal service and pricing from a business regardless of whether they exercise that opt-out or other rights. And it defines “selling” personal data broadly to include “disclosing, disseminating, making available, [or] transferring” that data. The internet industry doggedly tried to amend the CCPA to get an exception for digital advertising, but they failed in the legislative session that ended in September.
The California law effectively applies to large firms throughout the country, which are unlikely to deny everyone else rights they have to give Californians. Nonetheless, other states have been considering legislation as well. A proposed New York law, applicable to companies of any size, would go beyond California’s in creating a private right of action, enabling consumers individually to sue companies for privacy infringements. It would also treat companies as “data fiduciaries,” requiring them to act like other kinds of fiduciaries such as physicians and lawyers (who have to keep information confidential) and put the consumers’ interests first.
But a proliferation of state laws could boomerang. By exposing companies to conflicting requirements and the risk of being inundated with litigation, laws beyond California’s would strengthen the case for federal legislation to override separate state rules, perhaps with much weaker national ones.
Still, there is at least positive movement on privacy rights. Like antitrust, privacy law now provides a framework for reform that has considerable public legitimacy. The third challenge posed by the internet—disinformation—is much tougher, conceptually and politically.
Disinformation and the Degradation of the News
Two effects of the digital revolution have degraded the quality of news and information. American journalism has been in free fall as a result of the collapse of its traditional business model. About a fifth of local newspapers have closed, and many of the survivors are ghostly shadows of what they once were. Whole areas of the country have become news deserts, without any local reporting, leaving many people reliant on social media.
But the social media platforms, freed from any accountability for user-generated content, have until recently had no incentive to invest resources to distinguish truth from falsehood, much less to block outright disinformation. They have even taken ads—for example, from the Russians in 2016—in the form of “dark posts” in the news feeds of targeted groups, invisible to the public at large. The platforms’ algorithms have been optimized to build scale and make the sites as “sticky” and ultimately as profitable as possible. If conspiracy stories and other sensationalist and extreme content had that effect, it was of no concern to the platform companies. Google’s YouTube—which in late 2016 passed one billion viewing hours a day, close to total viewing time for all television and growing more quickly—has been a prime example of this pattern. A 2018 investigation by The Wall Street Journal found that after detecting the political biases of users, YouTube typically fed them videos echoing “those biases, often with more-extreme viewpoints.”
For all the limitations of the predigital media, journalistic organizations at least traditionally accepted responsibility for the reliability of the news they transmitted. The digital revolution shifted power to social media platforms that have disclaimed any such responsibility. If reforms are to rectify that loss of responsible editorial judgment, they will have to do two things: first, help reconstitute the economic basis of professional journalism and, second, hold social media platforms accountable for their role in shaping exposure to different contents, sources, and viewpoints. Democratic government can and should play a role in the first; it is not clear how, if at all, it can play a role in the second, except by reducing platform monopolies and requiring some degree of transparency.
Government support for journalism is widely regarded as anathema, but it shouldn’t be. Many liberal democracies provide substantial financial support to journalism through mechanisms that protect the integrity of the press. The BBC has served as a model of tax-financed, public-service broadcasting, while the Scandinavian countries have subsidized newspapers without impairing press freedom. Early American democracy would not have had its vitality without the subsidies the government provided to the press through below-cost postal rates as well as printing contracts and other means.
Since the late 1800s, however, American journalism has existed almost entirely on a commercial basis. Indeed, during the 20th century, newspapers and broadcast networks made so much money that the best of them amply cross-subsidized areas of coverage that could not have been justified as contributing to their financial bottom line. Now the profits are disappearing, but a majority of Americans have not registered the significance of the change. According to a survey by the Pew Research Center in 2018, 71 percent think their local news media are doing well financially.
Newspaper publishers would like to get some of the ad revenue now going to the platforms. They’re supporting a bill that would give them a four-year exemption from antitrust laws to enable them to negotiate collectively with Google and Facebook. But even then, the news media would be in a weak bargaining position; they need the traffic from Google and Facebook more than the platforms need them. The antitrust exemption wouldn’t resolve journalism’s crisis.
Coming up with reasonable proposals to support journalism is not the problem. One straightforward way would be an earmarked tax on digital platforms to support public-service journalism, with funds to be distributed on the model that the United States already has successfully developed for public broadcasting. Public radio is one of the few success stories in the recent history of the media; its federal structure—a decentralized system mostly of private nonprofit stations, with a mix of public and private funds and competing public-radio networks—helps minimize the risk of control by political incumbents.
But it is hard to imagine a worse time to advocate aid to journalism. When the president calls the media “the enemy of the people” and his party shares that disdain, proposals to support the press will inevitably be taken as partisan. Liberals should continue to think about such ideas—and, unfortunately, there will be plenty of time for reflection. The most that we can expect in the interim is a recognition among foundations and other nonprofit sources of the urgent need to bolster journalism, especially at the local level.
If your news feed is being altered by changes in a platform’s algorithm, you ought to have a right of explanation in ordinary language.
Efforts to change the policies of social media are also politically fraught. Resting their hopes on corporate self-regulation, several organizations are trying to promote independent fact-checking for use by platform companies. Both Facebook and YouTube have announced changes to their algorithms that they say will limit “borderline content.” In Mark Zuckerberg’s description, this is “sensationalist and provocative content” that “can undermine the quality of public discourse and lead to polarization.” Facebook is not blocking these posts, but it is limiting how often they show up in news feeds. In an explanation of how Facebook was preparing for the 2018 elections, Zuckerberg said, “Posts that are rated as false [on the basis of independent fact-checkers] are demoted and lose on average 80% of their future views.”
This kind of power, together with increases in “content moderation,” is both necessary and worrisome. It is necessary because viral falsehoods, deepfake videos, and other forms of deception are a genuine threat to democracy, and social media can serve as a checkpoint. It is worrisome, however, because social media can just as easily serve as a choke point for free speech and do so in ways that those subject to control cannot detect, much less contest.
Relying on independent fact-checking organizations is no escape from politics. According to a Pew survey, 70 percent of Republicans believe fact-checkers are biased, while only 29 percent of Democrats think so. In fact, the right has reason to be concerned about the fact-checking of its news sources. In a recent study of online news, Network Propaganda, Yochai Benkler, Robert Faris, and Hal Roberts find that journalistic norms continue to constrain news organizations running from the center-right (for example, The Wall Street Journal) through the center to the left. But on the right they find an insular media ecosystem skewed toward the extreme, where even the leading news organizations (Fox and Breitbart) do not observe norms of truth-seeking.
That is the context for conservative charges that Facebook, Google, and Twitter are biased against them. Ironically, after years of denouncing Democrats for supposedly wanting to bring back the Fairness Doctrine in broadcasting, some conservatives now want a new fairness doctrine for social media. Senator Josh Hawley, a Missouri Republican, has proposed legislation that would require internet intermediaries to demonstrate that they are politically unbiased to obtain the broad freedom from liability for user content conferred by federal law. The measure seems calculated to deter social media platforms from adopting criteria such as news source reliability in their news feeds and other algorithms.
The Hawley proposal ought to serve as a cautionary signal about any kind of government content regulation of social media. But transparency is one function that government could usefully advance. Legislation could require that social media and other digital platforms regularly disclose how their algorithms work, including how they affect different kinds of content and viewpoints. If your news feed is being altered by changes in a platform’s algorithm, you ought to have a right of explanation in ordinary language. Greater transparency would be a step toward what has come to be known as “algorithmic accountability.”
When the internet and digital economy were first emerging in the 1990s and early 2000s, it was unclear what shape they would take. The speed of change and its novelty were dazzling; a sober, historically informed analysis could not compete with dreams of a libertarian playground. But no one should continue to be under any illusions about the benign consequences of relying entirely on the market “as an alternative to intervention by government.” The neoliberal internet era is over. A new wave of intervention is coming. Yet the tech firms have now grown so large and powerful, and the Trumpian right wing so opposed to journalism and its norms, that the shape of things to come is unclear again. The question now is whose interests and values will prevail in the digital world that politics will shape. Monopoly, surveillance, and disinformation are the immediate stakes; the ultimate issue is what kind of society we want to live in.
