Why is Google hiding one of the UK’s biggest political advertisers?

One of the UK’s top ad-spenders is nowhere to be found in Google’s ad archive.

One of the UK’s top political advertisers is not featured in Google’s political ad archive, exposing the pitfalls of the search giant’s limited approach to transparency about online campaigning.

Best for Britain, the pro-Remain campaign launched by Gina Miller in 2017, has consistently ranked among the UK’s highest spenders on online ads. In April 2019, it was revealed that Best for Britain had paid £317,000 to run 890 Facebook ads over the previous seven months, a figure that positioned it as the platform’s third largest ad-buyer in the UK.

In the last 30 days, Best for Britain has spent £186,417 on 128 Facebook ads, many of which link to the group’s tactical voting website getvoting.org – whose declared aim is to keep the Conservatives out of office and thwart Brexit.

But while Facebook’s ad library provides a full breakdown of the group’s expenditure and messaging, Best for Britain’s ads are nowhere to be found in Google’s ad archive. That is not because Best for Britain did not buy any Google ads: on the contrary, Google search results for “tactical voting” will often feature an advert promoting getvoting.org at the very top. Yet, Google omits to include those ads, and the relative spending figures, in its archive. Of course, this is not about Best for Britain – it is about Google’s transparency failure.

A company source (and in line with Google’s official policy), says that Google’s transparency tools only includes political ads that explicitly feature a candidate for office, a political party, an officeholder, or a choice in a referendum. According to Google, partisan tactical voting tools or so-called “issue-based” ads – which promote political positions without attaching them to any particular party or candidate – do not fall in any of those categories and can therefore escape closer scrutiny. The source says that this is partly down to the difficulty to define what an issue-based ad is, adding that there is no clear consensus on that matter among politicians, researchers or technology companies.

That defence flies in the face of Facebook’s much more transparent – if imperfect – arrangement. Mark Zuckerberg’s social network has repeatedly come under fire for tolerating the presence of pseudonymous, opaquely funded campaigns – officially unaffiliated with any party – promoting a political agenda through targeted ads.

One of the most glaring examples was Mainstream Network, a pseudo-news outlet that spent over £250,000 in Facebook ads promoting a no-deal Brexit until it was outed by a parliamentary enquiry in October 2018; it was later found that a former aide of Boris Johnson’s had been running the page. But at least underhanded campaigns are simpler to spot and probe on Facebook, due to the company’s ad library displaying every advert about “social issues, elections or politics”.

Can you actually trust tactical voting websites?

In contrast Google’s archive leaves voters and media alike in the dark about the dealings of all non-party campaigns – from legitimate groups like Best for Britain itself, to, possibly, much murkier groups. In an email, a Best for Britain spokesperson said that the organisation is “all for greater transparency to prevent the misuse of ads.”

Back in September 2018, Facebook, Google, Twitter, Mozilla and other social networks and advertising bodies signed up to a self-regulatory Code of Practice on Online Disinformation written by the EU Commission. It’s a set of voluntary standards that include pledges by the big social networks to be more transparent about political advertising, as well as a larger effort to remove fake news and bot accounts.

“There’s a clear commitment in the Code of Practice to improve transparency on political issue advertising. We saw Facebook make an effort to try and include that in their advertising library. It certainly wasn’t perfect, and there are a lot of glaring gaps and false positives and false negatives, but there was an attempt by Facebook to do that,” says Chloe Colliver, lead researcher at the Digital Research Unit at the Institute for Strategic Dialogue.

“We didn’t see that from Twitter or from Google, who currently release only limited transparency around very specific political ads, and they haven’t broadened that out in the run up to the UK election at all.” (Twitter has recently announced it is going to stop all political advertising on its platform).

According to the Electoral Commission, non-party campaigners need to register with the electoral commission if they spend over £20,000 during an election campaign period. “But a lot of third-party campaign groups don’t register with the electoral commission, and so are able to spend whatever money they wish without any transparency to the public about where that money is coming or going in support of certain issues or parties or candidates,” Colliver adds.

According to Carl Miller, research director at the Centre for the Analysis of Social Media at Demos, the issue can only be solved by taking these kinds of decisions out of the hands of technology companies: something like political ad transparency should be defined in law by democratic authority.

“As soon as you move beyond advertising sent specifically by the party campaigns, it can be very hard to know where to draw the line,” he says. “We should never have left it to the platforms to make such difficult, thorny distinctions alone.”

In an emailed statement, a Google spokesperson said: “Our focus for the upcoming election is transparency to show voters who is purchasing election ads about candidates and political parties.” The spokesperson added the company “remain[s] committed” to different approaches to ad transparency, including ads about issues, and said the company “will have more to share in the future.”

Last week, Google announced that it would no longer allow political advertisers to micro-target voters based on their political leaning or affiliation. Political advertisers are still able to target voters based on age, gender or postal code, and on the basis of their search’s keywords.

https://www.wired.co.uk

SHARE THIS POST