Back in the 2000s, influential “thought leaders,” as they had come to be known, peddled the idea that the world no longer needed news organizations. In TED Talks and best-selling books like Here Comes Everybody and Cognitive Surplus, Clay Shirky proclaimed that a critical mass of the world’s citizens would rise up from watching TV on their sofas and become hyper-collaborating, hyper-informed news producers on something called “social media.” Shirky and many other techno-prophets in those days predicted—coining a number of new terms in the process—that these “web denizens” would replace the established press by “crowdsourcing” in “frictionless,” “peer-to-peer” networks that would overthrow ignorance and tyranny across the globe. As proof of concept, they pointed to the putative role that social media platforms like Facebook were playing in the Arab Spring and democracy movements around the world.
This vision might seem laughable in hindsight, but it was not entirely wrong. During the flowering of Web 2.0, interactive digital technology enabled what became the blogosphere, a place where ordinary citizens, if they had the talent and energy, could find and collaborate with diverse audiences. These bloggers made important contributions to the nation’s well-being, like documenting the many failures of the George W. Bush administration that elite media outlets were underplaying or getting wrong. Today, the elevation of racial justice issues largely results from ordinary citizens, not professional journalists, documenting instances of police brutality with their cell phones and social media accounts.
But these are the exceptions. On the whole, the idea that the raging hive minds of social media networks could replace or even reduce the need for an institutionalized free press seems ludicrous. Maybe Facebook could have evolved in ways that had the net effect of checking atavism and advancing the values of the Enlightenment through the wisdom of crowds. But it didn’t. Facebook instead became a corporate giant that used hate speech, conspiracy theories, and fake news to algorithmically generate engagement by users and then micro-target them with ads. Rather than a flourishing of democracy or an age of reason, this has led to pathologies ranging from the subversion of elections by actual and aspiring authoritarians, both here and abroad, to genocidal flash mobs in Myanmar and Ethiopia.
Shirky, however, was right that the established press would suffer serious damage, even if he missed the major reason why. As Google and Facebook emerged as giant monopolies, they used their domination over advertising markets to steal away the resources required to sustain the kind of diverse and independent press we have never needed as badly as we do today.
Historically, Americans have relied on advertising to pay for the high cost of producing quality reporting and editing, particularly in realms like local and public service journalism, on which a functioning democracy depends. Advertisers seeking to reach a particular market segment would place ads in a publication—or in particular sections of a publication, such as the business or style pages—that were read by their target audience. The revenues publishers earned from these ads in turn paid for the production of all kinds of journalism that benefited civic discourse and knowledge, but for which the average American was not prepared to pay a lot of money out of pocket. In this sense, advertising has long been the primary way our democracy has financed the free speech it needs.
But now this model has been largely displaced by behavioral advertising. If you have ever googled “hotels near Disney World” and found yourself chased around the internet for weeks afterward by ads for central Florida attractions, you know what behavioral advertising is. It’s the practice of targeting individual consumers with individualized messages based on what surveillance of their online behavior reveals about their individual needs and preferences. Before, these hotels would have to pay publications to advertise their services. Now, they pay Google instead.
Believing this behavioral advertising model to be inherently more efficient and effective, many people who are concerned about journalism have resigned themselves to a world in which the press must rely on other means of support, from greater use of paywalls to government subsidies to charitable contributions from philanthropies and high-net-worth individuals. These alternative models, unfortunately, have big drawbacks. Paywalls can make quality journalism too expensive. Subsidies and donations can make journalists too dependent on handouts from the powerful people and institutions they need to cover without fear or favor.
But fortunately, this resignation rests on a dubious premise. Today’s behavioral advertising markets are not intrinsically more efficient even from an advertiser’s point of view. Google and Facebook subject advertisers to monopolistic, needlessly high pricing and predatory business practices, and a growing body of evidence shows that it doesn’t even work very well in generating sales. Nor does behavioral advertising reflect the inevitable march of social progress. It’s based ultimately on corporate spying into your personal life and on pervasive new and old forms of discrimination that are deep threats to civil liberties, self-governance, and democratic institutions.
None of this is inevitable. Behavioral advertising has come about only because the government failed to apply policies that Americans used successfully in the past to manage competition in communications networks, like telegraphs and telephones as well as broadcast radio and television. These policies included long-standing laws and regulations that prohibited owners of essential communications infrastructure from engaging in lines of business that would put them in competition with their own suppliers and customers; that forbade them from engaging in price and other forms of discrimination; and that strictly limited their ability to monetize other people’s personal data.
Applying the same principles to today’s digital media markets will greatly restore the ability of advertising to support quality journalism while benefiting the public interest in many other important ways. It won’t solve all the financial challenges. But by using smart competition policy to break the dominance of platform monopolies, particularly Google and Facebook, over media markets, we will do far more to save a free press, in America and around the world, than any other single course we might take.
Many journalists have issues with the role that advertising has long played in their profession. That’s not without reason. Plenty of us have worked for publications where reporters and editors must fight fierce battles with the “business side” over whether and how to bend editorial direction in favor of advertisers.
In the 1970s, at rich and powerful news organizations like, say, The New York Times, CBS News, or Time magazine, reporters and editors generally had the upper hand in such struggles because there were so many businesses lined up to buy ads. If you wrote for the elite print media in those days, at worst you might have to produce some copy about luxury products for the new Style section, or compete for space with other journalists who did.
But if you have ever worked for financially weaker publications, you know that the pressure to bend story selection and story lines to please advertisers can be deeply demoralizing. When I began my career as a staff writer for the struggling New Jersey Monthly magazine in the late 1970s, we were able to do award-winning investigative journalism and deep dives into politics and public policy. But to pay the bills we increasingly had to do service journalism that we could sell upscale ads against, like reviews of the Garden State’s Top 10 Golf Courses and directories of its celebrity dentists. And, of course, we would never do a harsh review or exposé of any major advertiser.
It’s no longer just small outlets that have come under advertising stress. In recent years, such tensions have gotten worse for most publications, including elite ones. The pressure is especially evident on websites, where publishers, desperate to replace falling ad revenue, sometimes turn to pop-ups and pseudo-journalistic “native” ads that readers hate and that can destroy a publication’s brand.
Notice, however, that what is driving these tensions is not too much ad revenue but too little. If there were more potential advertisers, or if advertisers were willing to pay more, compromising on content wouldn’t be necessary. For all its inherent problems, journalism financed by a healthy, diverse advertising market has a lot going for it.
That’s especially true when you compare it to the alternatives. Consider, for example, the early days of American journalism, where advertising played a very limited role. Until the mid-19th century, newspapers depended instead primarily on political parties and government printing contracts controlled by politicians to pay their bills. This business model resulted in a rabidly partisan press that may have stirred up political engagement with salacious stories about the sex lives of Thomas Jefferson and Andrew Jackson. But it did little to develop the independent reporting and public education needed by a self-governing people.
That began to change only as publishers raised more revenues from advertising. This new business model not only allowed for greater political independence; it also had the benefit of democratizing access to nonpartisan journalism by dramatically lowering paywalls. Typical newspapers went from costing six cents to just a penny. This new “penny press” produced lots of sensationalist journalism. But it also for the first time allowed some publications to come to market with balanced in-depth reporting at a price readers could afford. One of the most successful penny papers was the New-York Daily Times, which would change its name to The New York Times in 1857.
Advertising also allowed for a flourishing of journalism dedicated to advancing social justice. Frederick Douglass financed his crusading newspaper, The North Star, by publishing ads for products ranging from fashionable boots to “hydropathic water cures.” William Lloyd Garrison kept the lights on at The Liberator by running ads for ladies’ “Champooing and Hair Dyeing” and for the “Boston Trecothic Calisthenic Academy and Gymnasium.” A few decades later, advertising for women’s designer clothes as well as Washington, D.C., gift shops paid for the advocacy journalism of , an essential part of the Congressional Union for Woman Suffrage.
Advertising also paid for the kind of muckraking journalism that revealed the abuses of corporate monopolies like Standard Oil and corrupt political machines like Tammany Hall. To finance the exposés of pioneering investigative journalists like Ida Tarbell and Lincoln Steffens, the publisher of McClure’s Magazine sold ad space to other corporations, hawking, for example, the Gillespie-Eden Corporation’s “Supreme Achievement in Electric Washing Machines” and the La Creole brand of dandruff shampoo.
Later, advertising for products like stereos and sporty cars in publications like Esquire and Rolling Stone financed the new journalism of the 1960s by such dissident writers as Hunter Thompson, Truman Capote, Joan Didion, and Gay Talese. In those revolutionary times, classified personals financed the growth of counterculture publications like the Village Voice, the Berkeley Barb, and many other free underground city papers. Ads by book publishers sustained niche political magazines ranging from the National Review to Dissent, Ramparts, and the Washington Monthly.
The symbiotic relationship between advertising and journalism evolved after the rise of television. By the 1960s, advertisers of broad national brands like Coca-Cola found it increasingly more efficient and effective to advertise on national television networks that could reach tens of millions of consumers at a time. The trend broke the backs of once-thriving mass-audience magazines like The Saturday Evening Post and Life by the early 1970s and made it increasingly hard for newspapers to compete for national accounts as well.
But advertisers still needed a way to target ads for products, like Adidas jogging shoes and Cuisinart food processors, that most people watching Gunsmoke or the CBS Evening News with Walter Cronkite would never buy. This reality created a vast opportunity for new forms of advertising-supported journalism that appealed to narrower market segments, particularly as the American public became more and more diverse in its tastes and unequal in its income after the 1970s.
In this era, advertising financed a golden age of specialty publications, ranging from high-quality trade publications like American Banker and Regardie’s to award-winning state and city magazines like the Texas Monthly and Washingtonian. The need for advertisers to more efficiently reach niche audiences also breathed new life into journals specializing in politics and public policy like National Journal and Governing. And it led to myriad publications focused on the emerging feminist, environmental, and gay rights moments like Ms., Mother Earth News, and The Advocate. Finally, it led to new editorial products ranging from Runner’s World to Creative Computing.
Beginning in the 1980s, the endurance of this model was tested by economic and regulatory shifts. Lax enforcement of antitrust and other competition policies allowed Walmart, for example, to use predatory business practices to close down local retailers. This meant that local newspapers, as well as local television and radio stations, found fewer and fewer businesses they could sell ads to. Meanwhile, large corporate mergers eventually destroyed the advertising base for many business publications. Newspapers in this era also faced declining revenues from classified ads, including to rival print products like Autotrader.
But despite these ominous trends, the strong symbiotic relationship between advertising and journalism continued. Surviving city papers became prime real estate for surviving local advertisers. Companies still paid, and reporting thrived.
At first, the internet was a boon for news outlets. Online advertising services, collectively known as “ad tech,” helped increase the flow of ad income for many existing and new editorial products. For example, in 1995 Kevin O’Connor and Dwight Merriman rolled out a new company called DoubleClick. Based on software they had developed in O’Connor’s basement, DoubleClick enabled both large and small internet publishers to host internet display ads in standardized formats while also allowing ad buyers to measure instantly how well their pitches were working. DoubleClick did collect data on customers’ online behavior, but it did not use it for targeting ads. As such, this and similar ad tech did not erode the financing of journalism, because it still required that businesses seek out the specific websites they thought prospective customers would visit. There was no micro-targeting, and no behavioral advertising. Editorial content remained the chief means of reaching targeted audiences. It simply became easier for publications to sell online ad space.
Similarly, a few years later Google developed a product called AdSense that marketers could use to automatically place ads on websites that matched the advertisement’s content. So, for example, AdSense would find a website containing lots of content about sports and automatically place pop-up ads for products like sports memorabilia and sports tickets, giving the website’s owners revenue they otherwise would not have had. This was helpful for small or new publications and for websites in particular.
The unfolding of ad tech remained a net positive for the financing of journalism throughout the mid-2000s. A competitive ad tech marketplace for display advertising allowed many bloggers, such as Joshua Marshall at Talking Points Memo, as well as innumerable “mommy bloggers,” to earn at least a modest living from ad revenues. We saw a wave of new digital publications in this era, including the Daily Dish in 2000, Gawker in 2002, Pajamas Media in 2004, and the Huffington Post in 2005. Finally, many legacy publications, including the Washington Monthly, got a second wind by creating digital editions and their own blogs, selling display ads against this content on the ad tech exchanges. But soon, the growing monopoly power of Google and Facebook over advertising market changes destroyed all that.
Let’s start with Google. In 2007, Google bought DoubleClick for $3.1 billion. At the time, Google told Congress and the Federal Trade Commission that it would not combine the increasing amounts of personal data it was collecting through Google Search, Google Maps, and Google Mail with information gained from DoubleClick about which consumers visited which digital publications and viewed what ads. But Google went back on this promise and meanwhile purchased more ad tech software companies, acquiring an even greater share of the advertising market, including ads on mobile devices. Moreover, it collected even more personal data from newer products like Google-owned “smart home” devices and the Google-owned Android operating systems found in most of the world’s cell phones.
All this gave Google an unprecedented advantage in the battle to attract advertising dollars. No newspapers or magazines, nor even the largest publishing, broadcasting, telco, or cable companies, can ever know as much about its own customers as Google does. Believing that Google’s unprecedented power of surveillance would allow them to boost sales with micro-targeted ads, advertisers fled from contracting with publications to contracting with Google.
But Google wanted to push its advantage even further. In addition to acquiring DoubleClick, it rolled up ownership of other companies that control the servers most publishers use to list the ad space they want to sell, as well as the servers most advertisers use to list the ad space they want to buy. Google now controls the very exchanges where these ads are bought and sold and prices are set. Indeed, according to a recent study by the United Kingdom’s Competition and Markets Authority (CMA), Google now holds a dominant position—as high as 90 percent—in every layer of the ad tech market.
One result of this roll-up is that advertisers can never know if they have paid a real, fair market price for their display ads. No one except Google can know, because all ad data is locked in black boxes that Google controls. This means that advertisers also have to commit substantial time and money to verifying where Google has actually placed their ads and how the ads are performing. That’s no small worry, since, as one study in 2017 found, more than half of all dollars spent on display advertising is spent for ads that were placed on fraudulent websites, or for ads that wound up being unviewable because of faulty servers.
Google’s control of ad tech drives down publisher profits in other ways. Google doesn’t just literally own and control advertising markets; it also sells advertising space for its own properties on those markets. So, for example, while publishers are trying to sell ad space on Google-controlled ad tech markets, Google is competing against them on the same exchanges as it tries to unload ad space on properties it owns, like YouTube. Only Google knows whether an ad it places for a client on YouTube is more effective than an ad it places for a client on some third-party website.
Google’s only peer in its powers of surveillance—and therefore the online advertising business—is Facebook, which tracks users across the various platforms it controls (including such giants as Facebook Messenger, Instagram, and Whats-App), as well as across the wider open web. Facebook’s business model is different than Google’s in that it only sells ad space on its own platforms. But its immense knowledge of users’ personal behaviors combined with a lot of predatory business practices have further shunted ad revenue away from newspapers and magazines.
Now, Google and Facebook operate effectively as a duopoly in digital advertising. Precise market share data is hard to come by, but according to the recent CMA report, 80 percent of all expenditures on search-and-display advertising collected in the UK during 2019 went to just those two companies. As the CMA report also notes, Google and Facebook’s share of total digital advertising revenue is significantly greater than the share of time that internet users actually spend on their platforms, implying that they are charging monopoly prices for the eyeballs they actually deliver.
This duopoly not only siphons off resources needed to sustain quality journalism, it also enables new frontiers in discrimination, as Google and Facebook gobble up petabytes of personal data and use it to let advertisers offer different deals to different people based on what their purloined personal data suggests is their ability and willingness to pay. It’s unclear exactly how effective this is for advertisers. But producers of mere journalism just can’t compete with the ill-gotten advantage.
Adding to the imbalance of power is the fact that if publishers complain that Google is consuming the revenue they need to create content, they have to worry that Google will retaliate by making their content disappear in search results or erase it from other platforms Google controls. Similarly, publishers who dare to cross Mark Zuckerberg have to worry that Facebook will stop referring its users to their editorial content. Publishers depend on referrals from Google and Facebook for about 75 percent of their traffic.
While this duopoly has devastated digital media (and individual privacy), it may seem like a step up for advertisers, who can better reach their audiences. But as it turns out, this system has been bad for them, too. In a recent book entitled Subprime Attention Crisis, the former Google employee Tim Hwang makes the case that between fraud and overhype of Big Data, behavioral advertising delivers so little actual value to advertisers that it is akin to the overpriced mortgage-backed securities market that blew up in 2007 when investors at last realized that it was built on sand. If you find that you keep getting served ads for products you bought last week or last month, you might well see how this could be true.
So we have a system that benefits nobody except Google and Facebook, and that is especially bad for journalists and democracy. The duopoly’s power over a free press comes from every angle. It’s now to the point that more and more media outlets can only stay in business by depending on handouts from third parties, like governments, foundations, and high-income individuals. And the largest source of this charity? Grants handed out by Google and Facebook’s marketing departments, which the corporations use to silence critics (see “Beggars and Choosers”).
What can be done? This crisis does not result from iron laws of economics or technological progress but from failures of public policy in three broad areas. All three of these can be fixed.
The first failure is weak antitrust enforcement. Though Google did pioneering work in search engines two decades ago, most of its market power today comes not from devising better products but from acquiring actual and emerging competitors across an ever-increasing swath of the economy. These include big household names like YouTube and Waze, but also all kinds of more obscure firms, like DoubleClick, that provide key technologies and control of market choke points. Similarly, Facebook’s power results largely from anti-competitive roll-ups, most notably of WhatsApp and Instagram. Together Facebook and Google bought more than 150 companies just between 2013 and 2018.
Many, if not most, of these deals violate the Clayton Act, passed in 1914, which prohibits mergers that may lead to even “incipient” monopolies. Many of these deals also violate the Clayton Act’s prohibitions on mergers that concentrate corporate power across different layers in markets, as when Google makes acquisitions to become simultaneously a digital publisher, a digital advertising agency, a data broker, a mapping company, a private post office, and the owner of the ad market and communications infrastructure. Prosecutors can use these statutes to roll back Google’s power.
They can also use the Sherman Antitrust Act to build cases wherever the platforms use exclusionary conduct to acquire or maintain monopoly power. No new legislation is needed to make these cases, though Congress may want to pass some statutes to clarify for activist conservative judges and timid regulators just what these long-standing antitrust laws actually mean. Fortunately, even key Congressional Republicans have recently agreed that we need stricter antitrust enforcement. And the Department of Justice, along with attorneys general from eleven Red States, have recently filed a major antitrust suit, based on the Sherman Act, that targets Google’s monopolization of search engine advertising.
We are not limited to antitrust actions, however. The next policy lever would entail transforming Google and Facebook into what they really should be: utilities. Web browsing and social media are, today, basic parts of how we connect and learn about the world, and there is a long tradition in American life of prohibiting owners of essential communications infrastructure from engaging in vertical integration, discrimination, and self-dealing. Imagine if we had allowed Bell Systems, the onetime telephone monopoly, to listen in on every conversation and to use the personal information it gleaned to allow corporations, politicians, and foreign governments to send “personalized” advertising messages and deals to different individuals. Imagine that by so doing it had monopolized the advertising industry and destroyed the business model for most of the free press. Imagine, finally, that Bell had also started using its wires to push out its own editorial content while other publishers had to rely on those same wires to get their products before the public.
Of course, that didn’t happen. Even though AT&T sought in the 1970s to become an electronic publisher, Americans used utility regulation to knock it back every time it tried to be anything but a neutral information platform. Similar principles of “net neutrality” have long governed the way we regulate other networked industries, including railroads and, for a time during the Obama administration, the internet service providers who control the hardware on which the internet itself runs. By classifying Facebook and Google as utilities, we can make net neutrality apply to them as well.
Among other virtues, this would remove the ongoing controversy over whether the owners of such digital platforms should be allowed to censor or discriminate against people they don’t like; they would not be allowed to play the role of censor, but nor would they be allowed to be in the business of publishing. Finally, applying traditional utility regulation to these platforms would mean that they would have to spin off any properties that put them in the position of competing against their own customers, such as data brokerage and advertising. Similar policies were recently recommended in a majority staff report by the House antitrust subcommittee after an 18-month investigation into abuses by Google, Facebook, and other digital platform monopolies.
The final big policy lever should have broad appeal even for people who don’t care much about the future of journalism: new federal privacy laws that ban companies from buying and selling personal data. Johnny Ryan, a colleague of mine at the Open Markets Institute and a privacy advocate with the Irish Council for Civil Liberties, describes privacy law as “Big Tech’s kryptonite.” A study he conducted found that Google’s ad tech infrastructure sends personal health data about individuals without their consent to 969 different companies engaged in behavioral advertising. Ryan and others are currently petitioning European data regulators to use the European Union’s privacy laws to crack down on Google’s ad tech business. If the U.S. similarly strengthened privacy laws and used them to prohibit the trading of personal data for advertising purposes, it would destroy the business model for surveillance capitalism that is enfeebling American journalism and with it our democracy.
Enacting these reforms will, of course, involve titanic political struggles, but no more than was required to bust up the monopoly power of Standard Oil and other predatory trusts back in the Progressive Era. The largest challenge will be doing the necessary public education before Google and Facebook gain control of virtually the entire information environment. We must win this war, and we can, but we don’t have a lot of time to lose.