Google will stop targeting political ads based on behavior. That’s a start. I call on every platform to end all targeted political ads (except by geography) — and for Congress to make this the law.
Here’s what happened with political ads so far in 2019:
- Twitter banned all political ads.
- Google just announced that for election ads, the only targeting it will continue to allow is by age, gender, adjacency to content, and location (at the postal code level).
- After relaxing its fact-checking of political ads, a Facebook spokesman said “we are looking at different ways we might refine our approach to political ads,” leaving the door open to possible restrictions.
I called for Facebook to end ad targeting (except for geographic) in my recent Boston Globe op-ed, and now I want to go further.
Why political ad targeting is a problem
You may imagine that the solution is to ban all political ads, as Twitter did. That’s too extreme. Candidates and parties deserve the chance to make their cases. While Twitter, as a private company, has the right to make any policy it wants, a blanket ban on all political advertising would violate the First Amendment. Of all the people to whom freedom of speech applies, candidates are among the most important.
But if we don’t ban ads outright, why ban targeting? Because everyone should see what political advertisers are doing. I don’t want political advertisers showing one message to Democrats and another to Republicans, or one to men and another to women, or one to people who clicked on a website about abortion and another to those who visited a site about guns.
(Geographical targeting is an obvious exception. People in Nebraska don’t need to see an ad from a candidate in Maine; that’s just wasteful and annoying.)
If we ban political ad targeting:
- Candidates will run a limited collection of ad formats, and watchdogs and competitors can check, evaluate, or respond to all of them.
- Lies in ads will decrease, since a candidate won’t be able to tell different stories to different voters.
- Candidates will focus on messages that resonate with the largest collection of voters, which is healthy for democracy.
A targeting ban is broader than you think
Here’s what I’m envisioning for this targeting ban:
- No targeting allowed, except by geography. This means even Google’s new policy doesn’t go far enough, since it still allows targeting by age, gender, or proximity to content.
- Applies to all media and social platforms, including Facebook, Google, Twitter, Instagram, Tumblr, LinkedIn, Bing, TikTok, YouTube, and anything else you can think of.
- Applies to all media companies.
- Applies to all ad networks. Most of the ads you see, including on media sites, are placed by ad networks based on behavioral information. If we don’t include ad networks, there’s a huge loophole in the ban.
- Applies to all political advertisers. (Google’s ban is specifically for election ads.) So this would apply to political parties, PACs, advocacy groups, and issue advertisers as well — anything intended to influence how voters vote.
- Applies to email as well. Targeted email is another environment where secret lies intended for specific groups will flourish. Shut it down. Political email lists shouldn’t be targeted based on personal data.
As it stands now, any remaining platform that allows targeted political ads has an advantage. Advertisers who can’t target on Twitter and Google are just going to go to Facebook and ad networks.
Moves like Google’s are intended to stave off legislation — “We got this, stand down.” But until the ban is on all digital ad formats in all platforms, deception will rule. The ban on targeting must be total. Which means it must be a federal law.
I’m betting you think I’ve gone too far. Fine. What’s your argument for why politicians should be able to sneakily tell different stories to different voters?