The perils of such ads far outweigh their benefits.
Eric Goldman (@ericgoldman) is the co-director of the High Tech Law Institute and Supervisor of the Privacy Law Certificate program at Santa Clara University's School of Law, where he specializes in internet law and intellectual property. Irina Raicu (@IEthics) is the director of the Internet Ethics program at the Markkula Center for Applied Ethics. Views are their own.
Note: This essay was originally published in The San Francisco Chronicle on December 8, 2020. We wrote this piece at a time when Google and Facebook had both temporarily stopped publishing political ads (Twitter had earlier announced a permanent ban on such ads). Since then, Google has resumed accepting such ads, and Facebook has announced that it will accept ads related to the Georgia runoff election, to be run only in Georgia. For the reasons listed below, we continue to believe that Twitter made the right choice, and that other platforms should follow suit.
Political advertising on social media played a major and well-documented role in the 2016 elections. Since then, the major social media services have reconsidered their approaches. Twitter banned political ads outright. YouTube restricted ad targeting and subjected the ads to some fact-checking. Facebook initially did little to restrict political ads, but then imposed a temporary moratorium on them. Google followed suit.
We think Twitter reached the right conclusion, and we hope Google and Facebook will make their moratoriums on political ads permanent, too.
Undoubtedly, banning political ads favors incumbents, who can reach potential voters more easily than challengers can. For example, incumbents often get more press coverage and have more social media followers. For that reason, the Democratic Senatorial Campaign Committee has urged Facebook to accept political ads again to bolster the two Democratic senatorial candidates challenging Republican incumbents in Georgia
While we’re sympathetic to those concerns, we believe the perils of online political ads far outweigh their benefits.
Online political ads pose two significant structural challenges for social media services: transparency and fact-checking.
Regarding transparency: Politicians can target their ads to deliver customized messages to niche audiences. This helps politicians express inconsistent (and likely false) messages to different communities without being held accountable for any hypocrisy or disingenuousness.
To curb this risk, Facebook and Google created databases that allow third parties to monitor politicians’ advertising practices — a remarkable resource that virtually no other media publishers provide. These databases help with transparency, but they may not be enough. Researchers at New York University’s Ad Observatory have shown that Facebook’s database is missing a significant number of political ads, and Facebook does not disclose the targeting information for the ads it does include.
Even if services do more to address the transparency issue, however, the fact-checking problem is essentially unfixable.
Politicians often lie to potential voters. Some lies are relatively harmless while others are brazen contradictions of well-known facts or promises that the politician never intends to keep. Because factual statements come in many shades, services struggle with how to best navigate the gray area between absolute truth and pure fiction.
Regardless of the severity of lies, social media services aren’t well-positioned to police them. Social media services don’t have the internal expertise to adjudicate the truth; they often lack the facts and context sufficient to recognize lies. Outsourcing fact-checking to third parties only partially solves the problem; the choice of those third-party sources introduces new biases and generates its own disputes.
Most importantly, even when a politician’s lie is apparent, social media services are nervous about fact-checking people with the power and motivation to regulate them. It is far easier for the services to acquiesce to lies from powerful people and hope that someone else — an opponent, the media, voters — will do the much-needed fact-checking instead. The services’ reticence to intervene creates an ecosystem where politicians routinely get away with lying.
Twitter’s political ad ban has not been flawless. There are difficulties deciding what constitutes a political ad, and blocked advertisers will feel frustrated and “censored.” Some political activists have found it harder to reach their audiences.
Still, banning political ads is the best outcome. Without effective fact-checking and transparency, online political ads degrade into a cesspool of propaganda and lies. Google and Facebook have already recognized the value of imposing a moratorium on political ads. Additional time won’t help them fix the underlying problems. A permanent ban will.