Google allows misleading ads on mail-in voting to remain

The ads “stand to promote confusion for the public at a time when clarity is needed,” said Kristen Clarke, president and executive director of the National Lawyers’ Committee for Civil Rights Under Law. “And so it is deeply troubling.”

A Google spokeswoman, Charlotte Smith, declined to answer specific questions about the ads or explain how they complied with the company’s policies, saying only in an emailed statement: “We have zero tolerance for ads that employ voter suppression tactics or undermine participation in elections. When we find those ads, we take them down.”

The ads are the latest in a campaign to discourage voters from using the U.S. Postal Service to deliver their ballots in the upcoming election by suggesting there is a difference between “mail-in voting” and the way absentee ballots have been cast for decades in most states. While some states prefer one term over the other, experts say, there is no meaningful difference. All ballots delivered by mail are verified before they are counted.

The ads were sponsored by a little-known group called Protect My Vote and appeared in response to Google searches for “mail-in voting” in numerous battleground states. These included Florida, Michigan, Iowa, Arizona, Texas, and Georgia, according to the University of Washington researchers.

Most voters, one common variation of the ad said, “think mail-in voting and absentee voting are the same. Think again! There are different safeguards for each.”

But in Texas, for example, there are not distinct processes for absentee voting and voting by mail, making the claims in the ad inaccurate. “Absentee balloting is when a voter who is eligible to do so requests a ballot by mail and votes that way,” Stephen Chang, a spokesman for the Texas secretary of state, confirmed in an email.

The ads direct users to a website, protectmyvote.com, that warns that mail balloting results in “lost votes and lost rights.” It attacks the U.S. Postal Service, saying the institution is “Steadily Getting Worse,” and argues, “Rushing to fundamentally transform voting processes just a few months before a critical election puts your vote at risk.”

The website, which lists no owner or contact information, has been promoted extensively by FreedomWorks, a tax-exempt nonprofit that helped launch tea party protests a decade ago and is now aligned with causes central to Trump’s reelection.

FreedomWorks is the sole organization to have repeatedly promoted the website, according to data from CrowdTangle, a social media analysis tool, and even purchased an ad on Facebook earlier this month to boost the campaign. Peter Vicenzi, a FreedomWorks spokesman, said a “partner group” was responsible for the website and the associated ads. Elements from the FreedomWorks site appear extensively in the Protect My Vote site’s source code, illustrating the extent of the overlap, though no connection is publicly disclosed.

“We are proud of the work we have done and continue to do on election security,” Vicenzi said. “We’ve spent at least six figures in the past year and plan on spending even more than that going into November.”

The decision by Google not to remove the material and the length of time it took to reach that decision raises questions about whether the technology giant is prepared to respond swiftly to election-related misinformation, said Shannon C. McGregor, an assistant professor of political communication at the University of North Carolina at Chapel Hill.

“This is a really worrying warning sign,” McGregor said. “On its face, this is misleading users about the voting process. And if that’s what these platforms want to protect against, then this is the type of ad they should remove.”

The days-long period Google took to review the ads, she added, deepens worries that Silicon Valley may not be equipped to respond adequately in the event of a chaotic post-election period, in which politicians or others may seek to make unfounded claims about the outcome of the vote.

Because of the expectation that large numbers of voters, concerned about voting in person during a pandemic, will prefer to vote by mail, experts anticipate that tabulating the results of the Nov. 3 election will take longer than usual.

“So how long is misleading information about the outcome of the election going to be left up?” McGregor asked. “And what damage is that going to do to the legitimacy of the election writ large?”

While the ads speak to minor differences in the voting process in some states, experts said, the sweeping and extreme language they use — similar to the Twitter salvos issued by Trump — are clearly intended to stoke fears about mail balloting and to bait users into visiting the associated website.

“We’ve not been taking the position that there’s a meaningful distinction between absentee voting and voting by mail, and deem President Trump’s statements as fully intended to cause chaos and confusion for the public,” said Clarke, of the Lawyers’ Committee.

In the case of Iowa, for instance, there is an opportunity to vote absentee in person, as opposed to the all-mail option. But the additional information about voting in Iowa provided on the website is plainly false. “Is mail-in voting in place?” it asks. The erroneous answer, provided in all caps and blue lettering, is “NO.”

Facebook, in explaining its decision to remove the material appearing on its service, pointed to the use of “voter suppression tactics” by the Protect My Vote page.

Google’s divergent decision reflects the patchwork of election-related rules in Silicon Valley. Facebook and Google allow political ads. Twitter does not. Facebook allows narrow behavioral targeting of these ads, while Google has limited that capacity. Twitter has taken to hiding posts by Trump that violates its rules on “civic and election integrity,” while Facebook has appended a link to its Voting Information Center.

In the case of the ads from Protect My Vote, Google did not even classify the posts as political — a category limited to ads mentioning candidates or elected officials, as well as political parties and ballot measures — meaning they did not appear in the company’s Transparency Report. That made it difficult to discern the size of the ad buy or the precise geographical scope of the campaign. It was not even clear whether any of the ads were still active.

Google has taken a harder line against other kinds of political material. Last month, the company rejected a 30-second spot called “Police State” that showed police officers beating, pushing, and Tasering protesters and other unarmed Americans — clips set against Trump’s description of himself as “your president of law and order.”