Facebook, Google political ad bans not effective, researchers say

Justin Hendrix / Jul 21, 2021

Two researchers at the Duke University Center on Science and Technology Policy conclude that bans on political advertising put in place by the tech platforms just before and in the period after the November 2020 U.S. elections were not necessarily effective, and had a number of negative side effects.

In September 2020, Facebook CEO Mark Zuckerberg announced Facebook would not accept new political ads after October 27th. Zuckerberg wrote that while he generally held the view that "the best antidote to bad speech is more speech," "in the final days of an election there may not be enough time to contest new claims. So in the week before the election, we won't accept new political or issue ads." That meant campaigns could post new ads to run through election day only until that a week prior to it; afterwards they would be rejected.

Then, fearing the possibility of a contested outcome in the Presidential race, Google and Facebook subsequently announced bans on political ads that would extend after election day. Other platforms also instituted bans of varying severity, including Amazon, LinkedIn, Spotify and TikTok. While some advocates lauded the bans, Democratic and Republican operatives suggested at the time they may be ineffective or worse- that they might drive campaigns to spend money in venues even less transparent than the tech platforms.

“The thing that was so striking to us is that these were historic interventions by the tech platforms, and we haven’t seen any analyses about whether these interventions worked or not. Before we go into 2022, we need to know more about what worked and what didn’t," said one of the report's authors, Matt Perault, the Director of the Center on Science and Technology Studies and a professor at Duke’s Sanford School of Public Policy, in an interview. Before his appointment at Duke, Perault was a director of public policy at Facebook.

Working with data from political advertising archives on Facebook and Google, information from the Federal Election Commission, and interviews with political candidates and advertising strategists, the researchers set out to understand how the bans impacted candidates' ad spending, the strategy and content of their ads, and the impact of the bans "across different campaigns, parties, or committees," and the implications on the special election for two Georgia Senate seats that took place on January 5th, 2021.

Among its findings, the report says that there is "little evidence that the bans reduced misinformation." It is, of course, difficult to prove the counterfactual, since there is no way to know what types of messages would have been posted had campaigns retained the ability to post new ads in the final days or hours ahead of the elections. But while there is some evidence that the bans pushed campaigns to choose more generic creative content, the researchers saw little sign that any more fact-checking of ad copy was going on in the run up to the deadlines set by the platforms to post new advertisements. And, candidates could not use paid advertising to address misinformation from opponents spread through other channels.

The impact of the bans was asymmetrical in a number of ways. The report concludes that "the bans likely hurt poorer candidates more than wealthier ones" by putting the onus on campaigns to make decisions and find alternative means to reach voters, something that was easier to do for campaigns with more resources.

Honestly, I think it’s bullshit. I don’t think anybody is going in there [Facebook’s ad library] in the seven days before looking through all of those ads and fact checking them … I don’t think any of that honestly happened and I truly do believe that Facebook was doing it to kind of save face and say: “Look, we did something, you can’t blame us any longer because we did a thing”

A political advertising strategist verbatim. Source

Another finding concerns the possibility that the bans hurt Democrats more than Republicans.

"Acknowledging that the overall impact on election outcomes was likely small," write Perault and his co-author, senior research associate J. Scott Babwah Brennen, "there was some evidence that the bans hurt Democratic campaigns and committees more than Republican campaigns and committees." One reason the report offers is perhaps that conservative content has more organic reach on platforms such as Facebook, so reducing the ability to use paid advertising to achieve similar levels of impressions reduced the chances for Democrats to match the reach of Republicans.

A second reason for the likely disparate impact offered by the report is the practices of Republican and Democratic digital strategists, and their "willingness to follow the rules set by the new policies". The report includes a comment from a Democratic strategist suggesting that Democrats are more willing to follow the rules set out by the platforms than their Republican counterparts. A Republican operative is quoted acknowledging that “it’s a cat and mouse game. So we’ll keep trying to figure out how to game the system. And it’s a game to them [Facebook], right?”

With an eye toward the 2022 U.S. midterm elections, the authors suggest that the platforms should permit political ads without restrictions, give advertisers more lead time to deal with policy changes, and do more to study and report the impact of the decisions they take on elections. It also suggests that "platforms should apply to paid content at least the same standards and
enforcement practices as to organic content," rejecting or applying labels to false claims just as they might to a typical post. "Paying for content should not exempt an advertiser from content restrictions or moderation," they say.

Finally, Perault and Babwah Brennen argue that more broadly "Congress should prohibit misinformation intended to suppress voting," pointing to the potential criminalization of deceptive practices related to voting as a solution.

There is language related to deceptive practices and voter intimidation in the For the People Act of 2021, but Republicans in the Senate recently blocked its advance.


Justin Hendrix
Justin Hendrix is CEO and Editor of Tech Policy Press, a new nonprofit media venture concerned with the intersection of technology and democracy. Previously, he was Executive Director of NYC Media Lab. He spent over a decade at The Economist in roles including Vice President, Business Development & ...