A post mortem on truth
The 2019 General Election was marred by what leading fact checking organisations and some journalists have characterised as unparalleled disinformation. It took place against the backdrop of acute political instability and uncertainty over how to deal with Britain’s looming departure from the European Union, an unravelling health crisis, and the climate emergency. It was described by all parties as a seismic election in terms of the significance and urgency of the issues at stake.
In spite of this, there has been relatively little post-election scrutiny as a result of the ensuing pandemic. Most of the available data and evidence on disinformation was collected in real-time by fact checking organisations, journalists and activists. As a result, this evidence is inevitably snapshot, selective and in some cases unsubstantiated.
We took a comprehensive look at patterns and examples of disinformation across the full duration of the campaign, focussing on the two major parties as well as the two main routes by which campaign messages could reach beyond their core vote: television news and online advertising. Specifically, we examined thousands of online ads across Facebook and Google, cross-referencing key messages with claims that were debunked by Full Fact during the campaign. This was supplemented by a qualitative analysis of television news coverage in relation to these claims, as well as an in-depth case study of a fake news controversy that surfaced during the final days of the campaign. Finally, we investigated Facebook ads run by non-party campaigners, of which the 2019 election saw a record number registered with the Electoral Commission.
Overall, the evidence collected strongly suggests that disinformation was an endemic feature of the Conservative party campaign. Although it does not make comfortable reading for the Labour Party either, the extent and frequency of misleading online ads between the major parties was incomparable. Over the course of the campaign, the Conservatives ran a total of 167 adverts across Facebook and Google which were either subsequently removed due to breach of the platform’s advertising policies and/or featured misleading or inaccurate claims. These ads ran for a cumulative total of 1,038 days which is the closest proxy measure for exposure and reach (give that both Facebook and Google only provide indicative ranges for the number of impressions generated by each ad). The equivalent figure for Labour was 139.
The extent of false online advertising by the Conservatives was therefore seven times that of Labour. It was also much more heavily skewed towards the final week of the campaign, during which the Conservatives pushed a particularly egregious fake news story ‘organically’. This suggested that a photograph of a 4-year old boy with pneumonia forced to lie on the floor of a hospital waiting room was ‘staged’. Amidst the fallout from the story, ‘senior Tories’ then falsely claimed that a special advisor to Matt Hancock was ‘punched’ by a Labour activist.
Although Google was considerably more likely to remove ads for breach of its policies than Facebook, these still ran for an average of 7 days and a cumulative total of over 350 days. More importantly, Google failed to act on dozens of ads based on claims that had long since been exposed as false by fact checking organisations. These included 13 ads that also directed users to the URL ‘labourmanifesto.co.uk’; a website run by the Conservatives.
Unlike Labour, false or misleading claims by the Conservatives were further reinforced by a string of non-party campaign groups advertising on Facebook as well as flyposting, and amplified by much of the mainstream press.
On the whole, television news journalists did challenge and scrutinise these claims. But in some cases there was a significant time lag between when the claims were first simply reported and matched against Labour’s reply, and when their veracity was questioned by journalists themselves. Even then, Conservative candidates and government ministers were often given a platform to both repeat and defend the claims. And during the last three days of the campaign, a fake news story being debated and debunked on television news ensured that a much more damaging story for the Tories quickly receded from the spotlight.
In the end, perhaps the most serious failing of television news coverage was its tendency to frame the issue of fake news as a problem implicating all political parties. This narrative served to obscure the vastly disproportionate role played by the Conservative Party in fermenting and circulating falsehoods.
In summary, the evidence suggests that the current regulatory framework for political advertising and campaigning during elections is not fit for purpose and wholly inadequate for the digital terrain on which elections are increasingly fought. There is no greater threat to democracy than disinformation, especially when it is produced and disseminated by an incumbent government and reinforced by the bulk of the mainstream press. As well as fact checking claims in real-time, broadcasters should provide viewers with regular updates and data on the scale of falsehoods put out by rival parties. As for online advertising, it seems clear that the current policies and enforcement operated by the platforms is nothing more than minimal and feeble given the scale of the problem. Unless and until the major platforms are able to implement a robust fact checking approval system before publication, there is an unanswerable case to ban all political advertising online during election periods.