Twitter CEO Jack Dorsey announced Wednesday that the social-media platform he runs will no longer allow political advertising. The company has reached the judgment that “political message reach should be earned,” as when people decide to follow or retweet a politician, rather than “bought,” as with targeted political ads. Commercial ads are fine, he argued. But political ads, which “can be used to influence votes to affect the lives of millions,” present new challenges to civic discourse, including “machine learning–based optimization of messaging and micro-targeting, unchecked misleading information, and deep fakes. All at increasing velocity, sophistication, and overwhelming scale.”
It isn’t clear that political advertising affects the lives of millions any more profoundly than commercial advertising for unhealthy drinks or more effective birth control or pornography or guns or fad diets or Hollywood films or video games or credit cards does—but let that pass. What about political ads advocating for an issue rather than a candidate? “We considered stopping only candidate ads, but issue ads present a way to circumvent,” Dorsey explained. “Additionally, it isn’t fair for everyone but candidates to buy ads for issues they want to push. So we’re stopping these too.”
Political advertising is still allowed on Facebook, where its accuracy is not policed. When reaffirming that policy recently, Facebook CEO Mark Zuckerberg asserted a need to be careful about adopting more rules that restrict what people can say. He doesn’t think that censoring politicians is right. “Although I’ve considered whether we should not carry these ads in the past, and I’ll continue to do so … so far I’ve thought we should continue,” he said. “Ads can be an important part of voice—especially for candidates and advocacy groups the media might not otherwise cover so they can get their message into debates.”
One reason Twitter and Facebook are diverging on this question may be that Twitter is a lesser platform for political advertising compared with Facebook. So far this year at Facebook, Wired reports, “the Democratic candidates for president have spent a collective $32 million on Facebook advertisements; Donald Trump has spent $13.7 million on Facebook ads in 2019 alone.”
By contrast, “political ad spend for the 2018 US midterms was <$3M” on Twitter, the company’s CFO tweeted. The Guardian notes that Twitter’s political-advertising operation “had just 21 advertisers across the entirety of the EU during the parliamentary elections this year,” and posits that Dorsey’s announcement “turned a weakness into a strength, cutting off a minuscule revenue stream to heap pressure on his main competitor.”
Business implications aside, would democracy be better off if Facebook followed Twitter and banned political ads or if pressure to do so led it to police ads for accuracy? I’m baffled by everyone who is confident in either position, given how little we know about key variables.
First, remember that Facebook is global.
Facebook has 270 million users in India, 130 million users in Indonesia, 120 million users in Brazil, 82 million users in Mexico, 68 million users in the Philippines, 58 million users in Vietnam, 46 million users in Thailand, and 37 to 38 million users each in Egypt, Turkey, and the United Kingdom. A ban on political ads might be good for some countries and bad for others.
Even looking narrowly at the United States, it is hard to know exactly how much Facebook political ads affect any campaign; whether their effect is similar or different in federal, state, and local elections; where campaign dollars now spent on Facebook would go if the platform banned political advertising; whether a ban would advantage or disadvantage Republicans or Democrats; or whether eliminating political Facebook ads would advantage incumbents or challengers.
“Incumbent politicians have tons of money and a huge megaphone to spread their message,” Senator Ted Cruz argued recently. “If you ban political advertising from social media, how on earth is any upstart challenger supposed to beat an incumbent? If you think America would be better with more career politicians, in both parties, entrenched in power for life, then Twitter’s proposed ban is a good idea.”
Perhaps Cruz is right, but I don’t know. Before weighing in, I’d want to better understand if incumbents or the best-funded campaigns are able to exploit Facebook’s ad targeting to greater effect than challengers or less monied candidates, whether because they hire more sophisticated ad managers to handle a complex system or because Facebook targeting is exponentially more powerful when used in combination with other data on voters that are in the possession of only well-resourced candidates.
As for issue ads, do they help activists for marginalized causes reach larger audiences? Do they help stoke division and anger? If they do both of those things, to what degree and with what effects? On even a single issue such as climate change, or health-care policy, or a local minimum-wage law, do we truly understand what effect Facebook issue ads have had or will have on public opinion? If not––and I can find no reliable research––how can we be confident in the effects of an overall ban?
As for whether Facebook should start policing accuracy in its ads, questions include whether the typical Facebook ad is more or less accurate than ads in other media; what percentage of past and present Facebook political ads would run afoul of whatever standards were adopted; whether inaccurate Facebook ads have a significant effect on public opinion; and how accurate, inaccurate, and evenhanded the ad checking would be.
We have clarity on none of those questions.
I do know that the politicians who most concern me, personally, would prove perfectly able to drum up fear and animus against immigrants, religious-minority groups, the free press, and more without running afoul of fact-checks. During the 2016 campaign, Donald Trump urged “a total and complete shutdown of Muslims entering the United States until our country’s representatives can figure out what the hell is going on.” He called for a judge’s recusal based on his ancestry. Fact-checking can’t stop bad ideas.
I dislike inaccurate political ads as much as anyone. But I am not at all certain that inaccurate ads are a major problem for democracy or that whatever ill effect they have would shrink absent Facebook ads.
What I can say for sure is that I’d like to know more about the effects of Facebook political advertising. The company could give the public, researchers, journalists, and policy makers access to more information, though it’s worth noting that it already expanded ad transparency in 2018 and 2019, going far beyond what many other platforms provide.
Before critics weigh in on whether Facebook should follow Twitter, then, they should carefully study what the company has already committed to making public during and after the 2020 election cycle.
Meanwhile, Facebook researchers, if you could persuade the company to give you anything it isn’t now providing to afford greater insight into the effects of political ads, what would it be? I’d be curious to hear via email, at firstname.lastname@example.org.