UK election: Did social media analysis predict a hung parliament?

Jun 9, 2017 | Social media

As the UK General Election results in a hung parliament, analysis by data science and media technology company, 4C Insights, revealed sentiment across social media towards the Conservative Party dipped to just 37% in the final days prior to the election. With no political party in the UK getting enough seats to form a majority […]

As the UK General Election results in a hung parliament, analysis by data science and media technology company, 4C Insights, revealed sentiment across social media towards the Conservative Party dipped to just 37% in the final days prior to the election.
clouds%20ben.jpg


With no political party in the UK getting enough seats to form a majority government, the analysis of social media sentiment in the run up to the vote seems to have indicated a dissatisfaction amongst Brits that ultimately led to hung parliament.
Top line findings:
· Tracking from the 1st May through to 5th June, Theresa May has managed an average engagement sentiment of just 43.5% on Facebook and Twitter.
· At 309,234 engagements, the busiest day on social media for May and the Conservative Party took place on the 22 May, the day the Prime Minister announced a cap on lifetime care costs.
· By contrast, Jeremy Corbyn drove the most engagements for the Labour Party (304,976) on 5th June, when he called for the Prime Minister to quit over police cuts in the wake of the terrorist attacks.
Tracking social media activity on Facebook and Twitter towards parties and leaders from 1st May, social data has revealed sentiment towards the Conservative Party bottomed out on 5th June at 37%, shortly after it was announced Theresa May’s lead over the opposition Labour Party had narrowed to just 1 percentage point. Overall, the Conservative Party lost 6 percentage points from the start of the period and 21 points from its peak. Through the course of the campaign, May has managed an average engagement sentiment of just 43.5%.
Key moments
· At 309,234 engagements across Twitter and Facebook, the busiest day on social media for Theresa May and the Conservative Party took place on 22nd May, the day the Prime Minister announced a U-turn on lifetime care costs, following widespread protests that more families would be forced to sell the homes of pensioners paying for their care.
· By contrast, Jeremy Corbyn drove the most engagements for the Labour Party (304,976) on 5th June, when he called for the Prime Minister to quit over police cuts in the wake of the terrorist attacks.
· Sentiment towards the Green Party peaked on 9th May, just after the party complained to the BBC about ‘disproportionate’ coverage of UKIP following local elections.
Correlation between social media and election results
4C’s analysis of social media data as proven to be highly correlated to outcomes in recent elections including the leave vote majority in the EU Referendum and Donald Trump winning the electoral college in the race for U.S. President. Aaron Goldman, Chief Marketing Officer at 4C, commented: “Traditional polling methodologies are based on self-reported behaviour that increasingly fail to predict election outcomes. Social media engagements more accurately reflect the sentiment of the populace and enable real-time assessment of the momentum behind political leaders, parties, and policies. Not all social media analsyes are created equal though. 4C removes passive engagements such as page likes and account follows to focus on active engagements including comments and retweets. These actions demonstrate stronger affinity and recency and thus are better indicators of voter intent.”
Methodology
From May 1st to June 5th, 4C analysed social media engagements and sentiment for the UK political parties and leaders across Facebook and Twitter. Social media engagements include mentions, retweets, comments, and post likes from candidate and party Facebook pages and Twitter handles, as well as related keywords and hashtags. Sentiment is a percent of engagements which are positive and not negative.
Source: www.4CInsights.com

All topics

Previous editions