Political advertising can’t go on in the way it was executed in the 2019 UK election. If you like the work we’ve been doing over the election please consider donating to our campaign here. We rely on donations to continue our work.
- The research conducted using one of the market leaders in panel and census internet data appears to demonstrate that heavy Facebook users tend to have more extreme political views.
- We propose that the findings have implications for how disinformation and extreme views can spread on Facebook. For example, due to the way algorithms on social platforms work this can often mean that more extreme views get a disproportionate share of impressions.
- We argue that this data reinforces the need for there to be rules for claims in political advertising and content.
- Facebook is often criticised for not fact-checking political ads and content. However, the solution for political advertising are similar rules being set up to those that govern all other advertising via the ASA – not in individual media owners applying their own rules.
Alex Tait, Co-Founder of the Coalition for Reform in Political Advertising comments:
“The debate in the UK election has never been more polarised. It is also proving to contain an unprecedented amount of misinformation and disinformation from the political parties. It has often been said that false information travels faster than truth in social media. That’s partly because it is fertile ground for spreading misleadingness in political information as its heavy users are very engaged. This research supports that and emphasises further the urgent need to modernise the rules for political ads.”
We have taken the reach of newspaper websites, and used that as a proxy for internet users’ political opinions using a methodology developed by YouGov.
These appear to show that heavier users of Facebook are more likely to hold political views towards the extremes of the political spectrum.
The chart below shows the percentage of heavy, moderate and light Facebook users visiting each of the websites along the x axis.
When we look at indices, we also see heavier Facebook users exhibiting the more extreme behaviours, as the charts below demonstrate.
So, for example, heavy Facebook users are 7.25 times more likely to visit Labour.org.uk or 9.3 times more likely to visit Conservatives.com than the average internet user (625 and 930 respectively against an index of 100 representing the average internet user).
This chart shows (against an index of 100 for the average Facebook user) how much heavy, moderate and light Facebook users visit each of the websites along the x axis. We have again attempted to rank the websites from left to right in their political persuasion. To give another example, a heavy Facebook user is 7.25 times more likely to visit Breitbart.com than the average Facebook user (725 against an index of 100 representing the average Facebook user).
‘Heavy’ Facebook users are defined as the top 20% by time spent, the ‘moderate’ the next 30% and the ‘light’ the bottom 50%.
We used one of the market leaders in online panel and census methodology data. Their approach combines person-level measurement from a global panel with census-informed tonnage of consumption to account for 100 percent of a property’s audience.
Participating companies place tags on all their content – web pages, videos, apps and ads, and these calls are recorded by the measurement providers servers every time content is accessed. The provider is able to view these calls on its global panel in addition to measuring the census tag calls. This perspective allows the measurement provider to validate that the tags are measuring activities consistent with its audience measurement methodology. Additionally, the company has developed a proprietary methodology to combine panel and server-side metrics in order to calculate audience reach in a manner that is not affected by variables such as cookie deletion and cookie blocking/rejection.
For more information about the research or the Coalition for Reform in Political Advertising’s campaign contact firstname.lastname@example.org.