"Why not just ban all political advertising?" is the common question lobbed at social media companies when each controversy arises.
It’s a reasonable query, given how small the income they provide to Facebook (at most 0.5% of total ad revenue, according to chief executive Mark Zuckerberg) and Twitter ($3m, it estimated yesterday, which is 0.9% of its 2018 revenue).
And the answer to this question has not changed since it was asked two years ago, when the Facebook/Cambridge Analytica data breach scandal broke. Social platforms that do not vet content (paid for or not) are still open to abuse from nefarious actors bent on damaging democratic elections by spreading misinformation.
So that's why many in the industry have welcomed Twitter chief executive Jack Dorsey's announcement that it will ban political ads from next month.
In a thoughtful thread of comments, Dorsey said: "Internet political ads present entirely new challenges to civic discourse: machine learning-based optimisation of messaging and micro-targeting, unchecked misleading information, and deep fakes. All at increasing velocity, sophistication, and overwhelming scale."
In what appears to be an eminently responsible move (far from the "move fast and break things" culture that has pervaded Silicon Valley in recent years), Twitter appears to be saying: "We can't control this stuff, so we're going to take a step back."
The proximate reason is publicity. Dorsey timed the announcement to coincide with the release of Facebook's third-quarter financial earnings. Knowing that Zuckerberg may not otherwise face tough questions from investors, Dorsey presumably sought to put his rival on the defensive by getting him to defend Facebook's recent announcement that it will not fact-check political ads.
But the deeper strategic reason is that the 2020 US presidential election campaign is about to ramp up significantly in the next few weeks, with the marathon primary season beginning with the Iowa caucus in January. The election in 2016 was targeted by a sophisticated Russian disinformation campaign that helped Donald Trump land crucial swing states by narrow margins to win the presidency.
Neither a polarised Congress nor Trump, who is now facing an impeachment investigation on top of woeful polling numbers, shows a willingness to pass new laws or guidelines to stop similar misinformation campaigns during the 2020 vote.
Dorsey said as much in the announcement: "We need more forward-looking political ad regulation (very difficult to do). Ad transparency requirements are progress, but not enough. The internet provides entirely new capabilities, and regulators need to think past the present day to ensure a level playing field."
And yet, by imposing a blanket ban on paid-for "political" ads, Twitter has put itself in the position where it must decide what is political versus what is an issue ad versus what is a non-political ad that is slightly political versus what is not a political ad full stop.
Dorsey said full details will be revealed on 15 November, one week ahead of the ban coming into force. When Twitter does this, it must be clear about how it hopes to enforce such a policy.
For a start, Twitter does not have a good track record when it comes to enforcing its own policies on how content is monitored. In June, it came under immense criticism for not taking action against hateful messages posted by Trump, in contravention of Twitter's hate-speech rules.
First, Twitter said it would downrank or hide tweets that broke its policy on hateful speech. Then it said, because Trump is a verified public official with millions of followers, there is a public interest in having his tweets remain, but hateful ones should have a disclaimer attached. Then earlier this month, Twitter said it would allow hate speech from politicans but would remove content that promotes terrorism, election misinformation or includes private information about people.
Mobbie Nazir, chief strategy officer at We Are Social, says: "[Banning political ads] is a bold move from Twitter, which sets the bar high for the rest of the industry. But the key for Twitter when it comes to political content is more organic than paid. Twitter is a news platform and has come under pressure recently to censor content from political leaders that violates their platform guidelines. However, they have resisted this citing 'newsworthiness' as the rationale."
Many will also try their best to get around whatever enforcement regime Twitter builds around banning political ads. Take it from someone who used to work for Cambridge Analytica: Patrick Fagan, co-founder of behaviour science consultancy, Capuchin, warns that Twitter's ban is likely to have "ironic effects" because of 'reactance theory': "People need to feel they have autonomy. "No smoking" signs have been shown to prompt smoking, for instance. Telling people, 'You’re not allowed to see this,' is only going to drive them to seek it out."
You may not agree with Facebook’s policy to not fact-check political ads (some of Facebook’s own employees feel that way), but it’s a position that is clearly (if perhaps radically) rooted in an aggressive defence of free speech. While Zuckerberg told Congress earlier this month that he thinks "lying is bad" (thanks!), he went on to say: "I think if you were to run an ad that had a lie, that would be bad… in our position, the right thing to do to prevent your constituents or people in an election from seeing that you lied."
This "sunlight is the best medicine" ethos is behind the launch of Facebook Ads Library last year – a treasure trove of information about what political parties are saying, to whom they’re saying it and how much they are spending.
Yes, Facebook may have become a source of ridicule in this regard after several deliberately false ads have been bought in the last month by politicians and activists testing the platform to its limits.
But, ultimately, Facebook may end up having the last laugh if Twitter fails to enforce its ban properly. How many human moderators are needed? How big of a role can automated systems realistically play in determining the nuance of content?
Coming up with convincing answers to these questions should be what the industry and Twitter's user base should look for now.