Why Don’t We Just Ban Targeted Advertising?

YOU PROBABLY REMEMBER this moment. It was April 2018, the peak of the Cambridge Analytica scandal, and Mark Zuckerberg was testifying before an angry Congress. Republican senator Orrin Hatch, then 84 years old, asked how Facebook could make any money by offering a free service. “Senator, we run ads,” Zuckerberg replied, breaking into a smirk. The exchange went viral as a testament to congressional ignorance—can you believe this old guy doesn’t know how Facebook works?

In fact, Hatch did know. Senators, like lawyers, often ask questions to which they already have the answer. But the moment was deeply revealing for another reason. In the nearly two years since that hearing, policy makers have been trying to figure out what to do about Facebook and other social media giants. They have argued over whether to revoke platforms’ Section 230 immunity, launched a barrage of antitrust investigations, and introduced a number of competing privacy bills into the Senate. Above all, they’ve tried to browbeat the companies into adopting better policies around things like fact-checking, content moderation, and political ads.

What they haven’t done is question social media’s underlying business model.

Jump ahead to another hearing, this past January. Another old Republican member of Congress, Ken Buck, was questioning another young tech executive, Basecamp cofounder David Heinemeier Hansson. “I don’t really care if they tell fifteen tee-shirt companies that I’m out looking for a tee-shirt,” Buck said. “It’s another thing when you’re trying to use that information in ways that I explicitly don’t want that information used. And so, what’s the answer there?” This was nothing new: a lawmaker in Washington who took for granted that our online behavior will be shared with advertisers, only then to wonder how one might contain the damage that ensues. But this time, his tech-world witness would reject the premise.

The solution to our privacy problems, suggested Hansson, was actually quite simple. If companies couldn’t use our data to target ads, they would have no reason to gobble it up in the first place, and no opportunity to do mischief with it later. From that fact flowed a straightforward fix: “Ban the right of companies to use personal data for advertising targeting.”

If Hansson’s proffer—that targeted advertising is at the heart of everything wrong with the internet and should be outlawed—sounds radical, that’s because it is. It cuts to the core of how some of the most profitable companies in the world make their money. The journalist David Dayen argued a similar case in 2018, for the New Republic; and since then, the idea has quietly been gaining adherents. Now it’s taken hold in certain parts of academia, think-tank world, and Silicon Valley.

The thinking goes like this. Google and Facebook, including their subsidiaries like Instagram and YouTube, make about 83 percent and 99 percent of their respective revenue from one thing: selling ads. It’s the same story with Twitter and other free sites and apps. More to the point, these companies are in the business of what’s called behavioral advertising, which allows companies to aim their marketing based on everything from users’ sexual orientations to their moods and menstrual cycles, as revealed by everything they do on their devices and every place they take them. It follows that most of the unsavory things the platforms do—boost inflammatory content, track our whereabouts, enable election manipulation, crush the news industry—stem from the goal of boosting ad revenues. Instead of trying to clean up all these messes one by one, the logic goes, why not just remove the underlying financial incentive? Targeting ads based on individual user data didn’t even really exist until the past decade. (Indeed, Google still makes many billions of dollars from ads tied to search terms, which aren’t user-specific.) What if companies simply weren’t allowed to do it anymore?

“To me, banning targeted ads is the ultimate root-cause solution when it comes to privacy,” Hansson told me. Seemingly every week, there’s another article exposing some company’s creepy behavior. A recent European study, for example, found that the gay dating app Grindr was sharing user data, including precise location history, with 35 different third parties.

But while each revelation is disturbing, the stories shouldn’t really shock us. The behavioral advertising business model has given rise to a teeming ecosystem of adtech firms, including data brokers, that pass user information through each step of the chain between publishers and advertisers. It’s all perfectly legal and very profitable, which explains why established companies like Adobe, Comcast, and Amazon have been getting in on the action. “The only reason that Facebook and others are collecting this data, buying this data—stealing this data—is because the data is so valuable,” Hansson said. “If you reduce the value of that data to near zero, then the entire incentive disappears.”

“Privacy” is only one way of describing the issue. Other experts blame the ad-driven business model for the proliferation of hateful and false content on social media. In a 2019 essay for the Knight First Amendment Institute, Jeff Gary and Ashkan Soltani argued that “restricting or lessening” the ability to microtarget ads would be more effective, and raise fewer free speech issues, than any law policing online discourse. The market for such ads creates incredible demand for users’ attention on both the front and back ends: the more time you spend on Facebook, the more finely it can target you and the more ads you’ll see. Combine that with the fact that users gravitate toward provocative content, and you can see where things might go. In the past, at least, sensation-seeking publications had to worry that sinking too far into the gutter would alienate their advertisers. Now the gutter is a money pit.

Then there’s politics. A 2018 study by researchers at Data and Society concluded that “today’s digital advertising infrastructure creates disturbing new opportunities for political manipulation and other forms of antidemocratic strategic communication.” Sally Hubbard, director of enforcement strategy at the Open Markets Institute, an anti-monopoly think tank, takes the argument further in a forthcoming book. “I honestly believe we are not going to solve any of the problems that we’re worried about, like election interference and disinformation, unless we ban targeted advertising,” she told me recently. Facebook, she said, has created a “manipulation machine” that can be used to discourage black voters just as easily as to sell sneakers. (Facebook didn’t reply to requests for comment.) “It’s the business model that’s the problem.”

If this nascent movement had a motto, that would be it: the business model is the problem. The task of regulating an increasingly out of control digital environment often looks like a multifront war against various enemies: privacy breaches, hate speech, disinformation, and more. What if we had a weapon that could bring all those armies to their knees?

LET’S PRETEND IT really happened. Imagine Congress passed a law tomorrow morning that banned companies from doing any ad microtargeting whatsoever. Close your eyes and picture what life would be like if the leading business model of the internet were banished from existence. How would things be different?

Many of the changes would be subtle. You could buy a pair of shoes on Amazon without Reebok ads following you for months. Perhaps you’d see some listings that you didn’t see before, for jobs or real estate. That’s especially likely if you’re African-American, or a woman, or a member of another disadvantaged group. You might come to understand that microtargeting had supercharged advertisers’ ability to discriminate, even when they weren’t trying to.

It’s true, the ads you came across while browsing might be for things you’re less inclined to buy. But a ban on targeted advertising wouldn’t mean the end of personalization. Spotify could still suggest Marvin Gaye based on your enjoyment of Sam Cooke. Bumble could still monitor your swipes to figure out your type. Netflix could still surmise that your life has felt empty ever since you finished season 7 of the Great British Baking Show, and suggest the appropriate spinoffs. (For example.) What companies couldn’t do anymore is share their dossiers about you with adtech companies and advertisers. The geyser of behavioral data currently gathered for marketing purposes would slow to a trickle. As a result, a lot less of your personal information would end up in the hands of data brokers and, from there, third parties like insurance companies, potential employers, or law enforcement agencies.

With the market for our behavioral data mostly shut off, social media platforms and other free apps would have to come up with ways to replace lost ad revenues. A hard paywall for Facebook or Instagram would be unlikely, though—social networks can’t afford to kick off too many users. Instead, they might decide to add a premium option next to the existing free version. Meanwhile, if theorists like Gary and Soltani are right, your feed would slough off its most offensive, hateful content.

You’d probably be able to read about all these changes in a revived news media. The past decade has been devastating for journalism, with waves of job losses year after year. The rise of behavioral advertising isn’t the sole culprit, but it’s a big one. Newspaper ad revenue, steadily climbing until 2006, has plunged ever since. Where have advertisers taken their budgets instead? Overwhelmingly, to Facebook and Google and the advertising infrastructure they control. Take away their targeting advantage, and marketers would have to shift back to paying publications to reach their audiences, so-called “contextual advertising.” Getting rid of microtargeting wouldn’t singlehandedly restore journalism to its glory days, but it could help—a lot.

The proponents of a behavioral advertising ban paint a rosy picture: less discrimination, better civic discourse, a rejuvenated news media. What’s not to like?

Other defenders of the status quo go further, suggesting that we would be forlorn without the steady companionship of our “relevant” online ads. “I think that relevant ads enhance the consumer experience,” said Dave Grimaldi, executive vice president of public policy at the Interactive Advertising Bureau, the leading industry trade group. “I think they help serve the right ads to the right people at the right time.”

It’s possible that consumers are happy to have the most minute details of their lives surveilled and monetized in return for seeing ads they might want to click on. This is a hard theory to test, because very few people even know they’re making the trade. However, one organization recently tried to find out. After the European Union’s landmark privacy law, the General Data Protection Regulation, went into effect in 2018, a Dutch public broadcasting agency started prompting all visitors to its website to choose, in a clear and straightforward manner, whether they wanted their data shared with advertisers. The result? Ninety percent opted out, and the agency abandoned behavioral advertising altogether. (A Google spokesperson notes that all users can opt out of personalized ads, and that Google has long prohibited personalized advertising based on sensitive information.)

Meanwhile, according to Google and others in the industry, microtargeting actually helps publishers to survive. Instead of siphoning off money, behavioral ads boost publishers’ bottom lines by providing their advertisers with more expensive, higher-value opportunities. “Data shows personalized ads are valuable for the entire digital advertising ecosystem: users prefer relevant ads and publishers make significantly more money from personalized advertising,” the company said in an emailed statement. Elsewhere, Google has argued that publishers would lose more than half their revenue if they stopped using the technique. Deighton has made similar findings, in a study commissioned by the Interactive Advertising Bureau.

But recent academic research suggests that the effect is actually quite modest. Professors Veronica Marotta, Vibhanshu Abhishek, and Alessandro Acquisti compared a major online publisher’s revenue from ads served to users who had cookies enabled—meaning they could be targeted—against revenue from ads served to users who couldn’t be targeted. (Their paper is still in draft form.) After controlling for a variety of other factors, they found that the presence of the cookie alone accounted for a 4 percent average increase in revenue, or 0.00008 cents per click. That’s not nothing, but it’s a lot less than what Google claims—and it has to be weighed against what publishers could be making if the market weren’t dominated by the Facebook-Google duopoly. And here’s a piece of anecdotal evidence: in 2019, the New York Times stopped running behavioral ads in Europe. Revenues were unaffected.

Meanwhile, the ability to track users wherever they go tends to shift ad revenue from higher quality sites to less reputable ones. “The way the adtech system works is, it follows the reader from Wired.com all the way down to the cheapest possible place, the basement bottom-feeders on the internet, and will serve you the ads there,” explained Nandini Jammi, a former product marketer and co-founder of Sleeping Giants, which pressures brands not to advertise on sites that promote hate or bigotry. Jammi pointed me to worldlifestyle.com, whose homepage features a random jumble of years-old articles on celebrities, self-help, and cute animals. It’s a content farm: a site designed not for human eyes, but to make money by harvesting ad clicks from bots.

Many small businesses, especially direct-to-consumer, do use behavioral ads to build their customer base. David Heinemeier Hansson told me his company, Basecamp, had success with a Facebook ad campaign in 2017. “Compared to everything else we did online, they were the most effective,” he said. “Targeted advertising works.” (Hansson added that he gave up on Facebook advertising anyway because he finds it objectionable.)

And yet, if behavioral advertising were such a boon to entrepreneurship, you might expect it to have spurred a wave of startup growth. Even more than a decade since the recession, though, both the startup rate and the share of Americans working for small businesses are at historic lows—in large part thanks to the rise of monopolistic companies like Facebook and Google, according to many experts. Microtargeting might help some small enterprises get ahead, but that doesn’t mean it’s a boon overall. As with any business strategy, there are both winners and losers.

Entrepreneurs aren’t the only small fish who use targeting. Political candidates who rely on a high volume of small donors also benefit from being able to target their appeals to the people most likely to respond. They could have a harder time campaigning in a post-microtargeting world. On the other hand, by far the biggest spenders on digital political ads this year are a trio of billionaires—Michael Bloomberg, Tom Steyer, and Donald Trump—the first two of whom probably would never have even made it to a televised debate if not for their war chests. Which illustrates a fundamental truth: Digital platforms, like most technological advances in history, often end up delivering the biggest benefits to the already rich and powerful.

ONE AFTERNOON IN January, I sent a Twitter DM to an Irishman named Johnny Ryan. All I knew about him was that he worked on policy issues for Brave, a privacy-focused browser; another source had suggested I reach out. Ten minutes later, we were on a video call. Ryan, an academic turned tech-worker turned privacy crusader, talked me through the inner workings of the online ad industry from his home office in Ireland, and I came to realize why his perspective is so important. In the U.S., a handful of folks are suggesting that the behavioral advertising business model should be illegal. In Europe, a coalition of activists including Ryan are saying that it already is.

But the GDPR generally forbids companies from processing user data without consent. It’s simply impossible, the activists argue, for users to consent to real-time bidding when there’s no way to know which companies are involved in the auction. Ryan calls it “the largest data breach ever recorded.”

The law does offer a potential loophole: Companies can process user data if it’s necessary for their “legitimate interests . . . except where such interests are overridden by the interests or fundamental rights and freedoms of the data subject.” The fate of the legal assault on real-time bidding may hinge on how European courts interpret that vague clause.

“If we are successful in our complaint, this indiscriminate sharing of data will have to end,” said Finn Myrstad, director of digital policy at the Norwegian Consumer Council, one of the other groups involved in the effort. “And if it ends in Europe, we’re also hoping that it will end around the world. Because you will create other business models that are more privacy-friendly.”

The U.S. doesn’t yet have a national privacy law, let alone a substantive debate over whether we should ban all behavioral advertising. But that could change faster than you might think. Just a few years ago, for instance, hardly anyone was talking about enforcing antitrust laws against the tech giants; now that idea motivates multiple investigations at the state and federal levels.

Perhaps the Overton window on behavioral ads will see a similar shift. Fordham law professor Zephyr Teachout helped put anti-monopoly issues on the map in her upstart 2014 primary challenge to New York Governor Andrew Cuomo. Now she is part of the pushback against microtargeting. In early February, she co-authored a paper arguing that the dominant internet platforms should be treated as public utilities and prohibited from using behavioral ads. If telephone utilities aren’t allowed to eavesdrop on our conversations and sell the details to marketers, then Amazon or YouTube shouldn’t be able to do the same with our browsing history.

And while there’s no GDPR-style legislation in the works here, some proposals would move us in that direction. In California, an initiative on this year’s ballot would expand the state’s new privacy law to give users broader rights to opt out of both the sale and sharing of their data. That’s not the same as a prohibition, of course; but it does suggest that the era of behavioral advertising as a fact of life may be coming to an end. So does the fact that internet browsers have been moving to block third-party cookies. Members of Congress may now take it as a given that your tee-shirt shopping habits should be tracked and traded. Will they feel the same in 2022, or 2024?

Even the staunchest critics of the present system recognize that killing microtargeting wouldn’t be a panacea. The internet’s worst pathologies—its effects on discourse, media, privacy, and so on—defy any single remedy. Shoshana Zuboff, author of The Age of Surveillance Capitalism, cautions that the practice of mining and monetizing user data has migrated to sectors like insurance, finance, and even automobiles—not to mention law enforcement, as the revelations about the facial recognition company Clearview AI remind us. And Roger McNamee, a former Zuckerberg confidant who is now a zealous big-tech antagonist, argues that microtargeting is only half the problem. Even a subscription-based social network would want to engage its users, he said, and what engages users is sensationalism and filter bubbles. “I do not think it is enough to address the damage of microtargeting if you don’t also deal with algorithmic amplification,” McNamee told me.

“Could all these [businesses]still exist? I think yes,” said Tim Libert, a computer science professor at Carnegie Mellon and a harsh critic of the surveillance economy. “The thing that won’t exist is this weird anomaly where some advertising companies—which is what they are—are the biggest, most powerful companies on the planet.”