Old advertising pitches nearly always boiled down to a cliche — “Half of advertising works, half doesn’t, and we don’t know which half is which. So, you’d better advertise.” Then the Internet came along and changed that equation, promising precise data on which ads do work, through highly targeted ads, for a price. That price has turned out to be hefty: democracy, free speech, even reality have all suffered at the hands of the targeted online advertising bargain.
Now, there’s a movement to ban targeted ads, taking direct aim at the business model that has made Facebook one of the richest companies in the world, and helped Google and Twitter gain strangleholds on their markets. The recent executive order signed by President Trump threatening the end of so-called “Section 230” liability protection for digital companies has added more fuel to the fire, and that became one focus of a discussion I attended Wednesday called “Making Facebook & Google Safe for Democracy,” sponsored by the American Economic Liberties Project.
“The Internet is a dumpster fire right now,” said Matt Stoller, a reseacher at the D.C.-based think tank. A ban on targeted ads is “one way to fix a lot of the problems we’re seeing.” Stoller is also author of Goliath: The 100-year War Between Monopoly Power and Democracy.
Google, Facebook, and Twitter make money through user engagement, and have cracked the code on keeping users glued to their services: outrage. By tracking user habits and hacking into their primal instincts, consumers are dragged into ratholes that lead to widespread spreading of misinformation hate speech. Like-minded — or at least like-behaving — groups are then sold to advertisers. This is a profitable enterprise — Stoller pointed out that YouTube recommended conspiracy theorist Alex Jones’ videos 15 billion times.
“That kind of divisive content is profitable…(tech firms) make money from illegal and racist content,” he said. “You can’t fix that by telling (Mark) Zuckerberg to censor better.”
Can it be fixed by eliminating Section 230 immunity for Internet services, making them more liable for negative content spread on their services? Perhaps — Stoller suggested that immunity acted like a subsidy to the tech firms, saving them the expense of worrying about negative or even dangerous content spread on their networks. But Karen Kornbluh, Digital Democracy director at the German Marshall Fund, wasn’t so sure.
“You’d get a lot of private lawsuits, she said. “But it’s not clear that disinformation is defamatory in a way that would hold up in court.”
The problem isn’t ineffective content moderation, the thinking goes: It’s the business model. Many companies are still profiting off spreading lies, hate, disinformation, and division.
Banning targeted advertising is a far more bold move, a step that was first widely advocated by David Dayen, inthe New Republic, in 2018. Dayen is executive editor of American Prospect.
“Instead of leaving regulation to Facebook, or devising one Rube Goldberg scenario after another to try to protect consumer data, the U.S. can take one simple, legal step to roll back this dystopian nightmare: ban targeted advertising,” he wrote.
What would non-targeted ads look like? The ads you saw before Facebook arrived, when Google was merely selling search terms. Or even the world before digital advertising, when companies that wanted to reach audiences would work with entities that had hard-earned, trusted reputations. These ads didn’t require splitting users into groups that were open to messages which limited hid housing access from minorities, or traded in hatred of religious groups, or zealotry around vaccines, or pseudoscience that the Earth is flat.
“In a targeted advertising model, misinformation and conspiracy theories are often the product, not an accident,” wrote Jeff Gary and Ashkan Soltani in a paper for Columbia’s First Amendment Institute calling for a ban on targeted ads. Such a regulation would be far a more productive and First-Amendment-friendly solution than forcing fact-checking on private companies or requiring them to ban speech.
Banning targeted ads would also eliminate many of the spookiest experiences that Internet users currently suffer; being stalked by ads for products they’ve searched for, or recommended disgusting videos because they’ve accidentally clicked on an unusual news source. It would end the business model which requires hoovering up every data point possible about mobile phone users and web browsers.
Resistance to the idea is sure to be dramatic. While Google was plenty profitably before the wide proliferation of targeted advertising — and still makes good money from search term ads — Facebook’s business is built on the idea. Digital advertising was worth $130 billion in 2019, much of it targeted ads.
Meanwhile, it’s unclear that a targeted ad ban would stop the Internet’s bad habit of favoring crazy content. After all, publishers have done desperate things to get readers’ attention since the beginning of….media.
“Even a subscription-based social network would want to engage its users, he said, and what engages users is sensationalism and filter bubbles,” wrote Gilad Edelman in Wired earlier this year. Roger McNamee, the Facebook refugee, told him “I do not think it is enough to address the damage of microtargeting if you don’t also deal with algorithmic amplification.”
On the other hand, the current situation is untenable and existentially threatening. It’s also unclear that targeted ads really work for any of the parties involved, except the ones cashing the checks. I’ve long thought this: When I’ve just bought a set of drums, the worst ad ever invented in the history of humanity is an ad for the same drumkit.
Carnegie Mellon professor Alessandro Acquisti recently studied quantified my anecdotal observation.
“We find that when a user’s cookie is available, publisher’s revenue increases by only about 4%. This corresponds to an average increase of $0.00008 per advertisement,” he wrote in a paper called Online Tracking and Publishers’ Revenues: An Empirical Analysis.
A $0.00008 premium in exchange for hate speech, the death of democracy, and the end of knowledge as we know it? Seems like a bad deal. There must be a better way. It’s a big problem, one unfit for a small solution.