Yesterday BBC aired the last in its series of EU Referendum Debates from the Wembley Arena in London. The two hour event saw Boris Johnson and Sadiq Kahn go head to head for leave and remain respectively. The debate was watched by a studio audience of 6000 and televised live within the UK. While it aired, Jennifer Jackson-Preece collected a random sample of over 50,000 tweets with the hashtag #BBCDebate using the Discovertext platform.
A word frequency analysis of this Twitter sample strongly confirms the Twitter trends that I have been observing over the past couple of months. It is also broadly consistent with the Brandwatch meta data released on 8 June and with the YouGov polling data analysed by my colleagues Sarah Hobolt and Christopher Wratil.
The big issues which dominated the #BBCDebate conversation were migration and the economy, with migration references outpacing those to the economy by a ratio of almost 2:1. Leave hashtags appeared far more frequently than remain hashtags in both tweets referencing migration and the economy. In fact, contrary to the Hobolt and Wratil YouGov findings, those concerned about the economy were even more likely to identify with the leave position (by a ratio of 3:1) than those concerned about migration (where leave outstripped remain by a ratio of 2:1).
From the Twitter reaction to the big issues, you would expect Boris Johnson to have gained more Twitter mentions than Sadiq Khan. But in fact the reverse is true. Sadiq Khan appears in almost 11,000 tweets while Boris Johnson is cited in just over 9000 tweets.
So which side won on Twitter? Analysis of the #BBCDebate suggests the leave campaign won the final referendum debate on Twitter. Of the more than 21,000 Twitter references to leave and remain, leave appeared in 56% while remain appeared in 44%. Of course this figure is only a proxy indicator of support for either side. Just because an individual uses leave or remain in a tweet does not necessarily mean that they are pro-leave or pro-remain. And crucially more than 30,000 tweets did not include a direct reference to either leave or remains. So there is a large margin for error in this Twitter sentiment analysis.
Twitter is not an opinion poll. Individuals choose to tweet. Those who participate in political debate on social media are more inclined to have firm opinions, and they tend to interact most with those who share their opinions. As a result, politics on social media is less likely pick up undecided positions.
A snap YouGov poll for the Times conducted just after the debate saw leave winning, with 39% of respondents, while 34% supported remain and a further 17% remained undecided.
This blog represents the views of the author and not those of the BrexitVote blog, nor the LSE. Image CC0 Public Domain.
Jennifer Jackson-Preece is Associate Professor of Nationalism, with a joint appointment in both the European Institute and the Department of International Relations, LSE.