Political ads: What should Facebook have done?
Facebook’s Mark Zuckerberg has announced the company are going to make political advertising on their platform “more transparent”. In many ways, this is a welcome move, albeit one made under duress and lacking specific detail.
Zuckerberg’s announcement promises:
- Disclosure of who paid for a political ad
- Publishing all ads placed by an organisation’s page
But it’s not enough.
We think Facebook should go further to help users and researchers. People should be able to see:
- The full targeting of the ad, as specified by the advertiser and the goal they have selected for it
- How much the advertiser has budgeted to spend on it
- The current and predicted reach of the ad
- When ad was placed, and the predicted date the budget will be exhausted
- Data about the locations the ad has been viewed, ideally on a map
- The proportion of paid vs. non-paid interactions on the ad post
- All alternative versions of the ad running now or in the past (e.g. A/B tests, or duplications with revisions)
- Ads in perpetuity — not just those that are currently running — and include deleted posts in the archive
Further to these, Facebook should require advertisers to mark whether their ads contain political content and use this information to routinely tag/graphically treat political ads so users are easily able to distinguish between normal posts, non-political ads and political ads.
Users can then report ads with political content that aren’t properly categorised. The ads should be suspended, pending investigation by moderators. This would add a further safety valve to the (probably algorithmic) solution Facebook will adopt when identifying political posts and pages.
Facebook should also help regulators and researchers by providing real time raw data feeds of political advertising, allowing instant oversight of political ads. This will prevent delays in getting hold of relevant data (in the case of the 2016 US Presidential Election) or the complete absence of data (in the case of the UK Brexit Referendum) which, in each case, may have had their outcomes affected by Facebook advertising. It will allow more trained eyes a view of what’s out there and should mean more misinformation and cheating is caught.
Over time, we hope and expect all internet platforms to adopt a higher degrees of transparency, like those we recommend.
Zuckerberg’s announcement may well fix political “dark ads”, but it’s a limited concession. There is much more work to do to reduce concerns about dark strategy — the microtargeting of voters and the fragmentation of the political offer in order to divide and conquer society — which has become something for political campaigns, foreign powers and data-driven consultancies, to exploit.
The additional visibility we propose will mean voters can see the issue from three angles — first, what an advertiser wants to achieve; second, what Facebook users are seeing and third, that regulators have the full access they need in order to protect democracy in the 21st century.
As much as companies like Facebook taking voluntary steps to make things better are welcome (note that the other platforms haven’t yet made similar offers), these issues really need wider public debate, legislation and proper enforcement. Unless as a society we start to proactively consider the problems thrown up by unregulated targeted ads at vast scale, challenges like these will skip from platform to platform, issue to issue and from one new technological development to another.
Ultimately therefore, this is an issue for governments, and citizens, to grasp, to ensure the integrity of elections.
In the meantime, we’ll continue to launch Who Targets Me? in more countries, working to help voters understand how social media ads are being used to try to influence their political views.
