Tuesday, November 19

Soon after the 2016 election concluded, Facebook and Twitter understood they needed to make changes to their political advertising tools. Their services had been exploited by bad actors — including Russians posing as Americans pushing ads discouraging voting — and their sales teams operated essentially without rules, leading to some regrettable outcomes.

Twitter’s sales team, as I reported in 2017, offered the Kremlin-sponsored television network RT 15% of its total US election ad space. Facebook, for its part, shared a diagram of the US electorate, breaking it up into 14 segments, from left-leaning "youthful urbanites" to pro-gun "the great outdoors" — a blueprint for dividing the country.

Granular political ad targeting wasn’t new in 2016. But in that cycle, tactics grew dirtier while tools grew more sophisticated and exploitable, creating a mess. The systems needed to evolve.

Over the past three years, Facebook and Twitter have made a series of changes to restrain bad actors. They both increased transparency by listing political ads in public ad libraries. They banned foreign ad targeting of political ads to people in the US. And Facebook worked to cut down fake activity by requiring political advertisers to confirm they lived in a US address via a mailed code.

Even after putting up these defenses, Facebook and Twitter knew their ad systems were powerful machines that above board political groups could use to sow division — and a potential PR nightmare too boot. Advertising increases awareness, but it also manipulates. So both had to consider different options for handling the upcoming 2020 cycle:

Ban All Political Ads

Twitter announced earlier this month that it would ban all political ads, including candidate and issue ads. This ban, the broadest possible, would protect Twitter’s ad system from being overrun with divisive and false ads. But as some critics pointed out, it would also limit cause-based advertisers. Pro-climate advertisers, for instance, would be held back under such a ban, while fossil fuel companies could promote their businesses.

Worth noting: Politicians and causes thrive on unpaid reach on Twitter, unlike Facebook. So a total political ad ban wouldn’t make a major difference there.

Twitter, in the end, did not roll out the total ban.

Ban Political Ads That Mention Elections And Legislation

This is where Twitter netted out. If you’re advertising generally about an issue, that’s okay with Twitter. This policy allows cause-based advertisers to push out their message, while keeping Twitter free of candidates and political causes that distort the truth for their own ends. Though it still doesn’t address distortions inside Twitter more broadly.

Allow Ads, Fact Check

Permitting political ads and fact checking them would allow political ads to flourish, while cutting down on distortions. In an ideal world, this could work — to a point. Remember, this is politics we’re talking about. Fact checking political ads would be a nightmare and a running controversy for anyone who tries to do it. Mark Zuckerberg emphatically said that Facebook wants no part of this. Snapchat, which is not a particularly important political platform, is taking this approach.

Allow Ads, Limit Targeting

Here’s a question to consider: Should politicians be able to carve up the electorate into tiny groups, and then run ads based on their interests? Internet advertising is largely built on this practice, and social media is its top enabler. Facebook was considering a targeting rollback, according to rumors flowing around DC, but Facebook marketing VP Carolyn Everson said on Monday it’s not going to happen. Facebook’s refusal to consider this option remains puzzling given its wide array of targeting criteria. Twitter, meanwhile, recently said it would ban microtargeting for issue ads.

Allow Ads, Limit Ad Placements

Restricting the number of ads a political campaign could run might help with the “carve up the electorate” problem, and do so with a lighter touch than broader ad targeting restrictions. Could a platform do this? Yes. Facebook will. “In mid 2020, we are implementing a limit on the number of ads each Page can run at the same time,” the company said in a blog post last month. This will impact all advertising, not only political. But let’s just say the timing is interesting.

Allow Ads, No Restrictions

There are few indications that Facebook or Twitter are considering approaching 2020 with no restrictions whatsoever. But it’s certainly an option for them. Regulation could put some restraints on what these companies can offer, but there’s no indication anything meaningful is on the way.

For now, Facebook and Twitter will take their best shot at balancing the policy that keeps them relevant and, they hope, doesn’t lead to democracy-shaking disaster. For Facebook, this means letting political advertisers get their message in front of voters while limiting placements and staying away from determining which political advertisements are true. For Twitter, the plan is to force politicians and causes to mostly rely on unpaid reach, minimizing voter manipulation via ads to the extent possible. As we head into 2020, this is where both platforms are netting out.