LSE - Small Logo
LSE - Small Logo


May 2nd, 2017

Practice makes perfect: how the Leave campaigns dominated Twitter in the EU referendum


Estimated reading time: 5 minutes


May 2nd, 2017

Practice makes perfect: how the Leave campaigns dominated Twitter in the EU referendum


Estimated reading time: 5 minutes

Twitter was a crucial campaigning platform ahead of Brexit. Simon Usherwood looks at its use by campaigners and places their efforts in a broader context. He explains that since the 1970s, politicians did not actively promote the benefits of being in the EU, leaving it to eurosceptics to promote their own agenda. This focus meant that in 2016, those for Leave had well-rehearsed arguments to push forward – this time through a very modern medium.

With the outcome of last year’s EU referendum still very much with us, it is useful to return to that campaign to consider how the different sides talked about the issues involved and how they sold their messages. Could the roots of current confusion lie in the obfuscation of the referendum?

In our recently-published paper, Katherine Wright and I look at the Twitter campaigns from the two official groups – Stronger In and Vote Leave – as well as Leave.EU between February 2016 and the vote in June. At the time, we produced weekly updates on our methodology and findings, as part of our work with the UK in a Changing Europe programme, but it has only been in retrospect that some of the broader findings have come into clear relief.


The first major finding is that all groups used Twitter primarily as a platform for mobilising and organising pre-existing supporters, rather than as a conversion tool. Given the nature of the medium, this is perhaps unsurprising, since it requires users to decide to follow accounts and it is relatively hard to push content out to non-followers in any significant volume, unlike Facebook.

With this in mind, there are two things to look for: the volume of your followers and the size of your footprint. The former is easy to check and shows the two Leave groups well ahead of Stronger In (over 94,000 for Leave.EU and 69,000 for Vote Leave against 48,000 for Stronger In by the time of the vote). The latter was determined by checking the mean number of retweets per tweet per follower: this gives a sense of how much sharing of content went on between the different groups’ followers. Here it was Vote Leave that was clearly ahead of the other two groups, out-performing them both in all but two weeks of the period (see Figure 1).

In short, Leave not only had the numbers but also had more of the engagement, spreading its message much wider than Remain. This is further underlined by the existence of two major Leave bodies, where no other Remain group came anywhere close to a comparable profile. Even if we assume a high degree of overlap between Vote Leave and Leave.EU, the absolute size of the online constituency for Leave appears to have been much larger.

Negative campaigning

Our second set of findings point to a broadly comparable approach to the use of negative campaigning by all three groups. As Figures 2 and 3 show, there was a gradual decline in the percentage of positively-framed tweets over time, especially once the official campaign began in April, although still tending to be more positive than not on balance. At the same time, there was a more marked rise in the use of ad hominem tweets in the final weeks, as groups made ever more attacks on their opponents: even the call for more civility following the murder of Jo Cox in the final week only had a limited impact on this.

Arguments used

The final area of interest for us was the kind of arguments used, both in terms of the themes and the language. The simplest way to visualise this is by looking at the most common word-stems found in the tweet texts: Table 1 shows the top 20 for each group.

We might start with the relative absence of substantive themes or ideas: the focus is much more on mobilising language (e.g. ‘voting, ‘support’, etc.) and self-referencing (all three groups have many tags of themselves, usually from their retweeting of supporter’s content).

Where substance does come through, it is the familiar themes: economy and business for Stronger In, control and immigration for Vote Leave. This suggests reinforcement of their core arguments, rather than pushes into each others’ territory. Leave.EU’s more free-floating position in the campaign is reflected by the presence of only one substantive word (‘trading’) in their top 20, on a weighting that would have only just made onto the lists for the other groups.

However, there are two much more striking observations to be made of these lists.

The first is the use of individuals. While Stronger In has no names in its top-20, both the Leave groups do. Usage followed one of two patterns. When it was a Leave supporter – Michael Gove, Nigel Farage, Boris Johnson – then framing was very positive and reinforcing of their leadership roles. When it was David Cameron, then it was equally uniformly negative, using his words and actions against him, and reminding non-Tory voters than this was an opportunity to hurt his government.

This imbalance of ‘faces’ in the campaign clearly advantaged Leave, both in their range of options and in their ability to make the most of voters’ willingness to make a second-order election of the referendum, voting on issues not on the ballot.

And this is perhaps the most noticeable finding to be taken from our analysis: the campaign on Twitter was fought very much on Leave’s terms and using its frames. We might not be surprised to see the Leave groups using ‘leaving’ or ‘Brexit’ very frequently in their tweets, but that it is the same for Stronger In points to an inability to communicate a vision of what EU membership could be like on their own terms.

Root causes?

Taken together, all this points towards a Remain Twitter campaign that failed to find its feet. And here we might suggest that this has been a function of the broader trajectory of UK-EU relations. Once membership was secured in the 1970s, politicians and other groups did not engage in ongoing promotion of the benefits of being part of the European Community/Union. Instead, it fell to critics and eurosceptics – especially from the 1990s onwards – to set the agenda for public discussion of ‘Europe’. Indeed, in our study we find very little in the Leave campaign’s arguments that had not been made at several points in the preceding decades.

And this is the point. Leave was able to draw on a deep well of arguments and pitches, honed over the years and colouring the frame of public discourse. By contrast, Remain was starting largely from scratch, trying to find resonances among a public unaccustomed to hearing such views.

Certainly, there were many other factors at play in the campaign, but the importance of this long-term context is brought through in this very modern medium.


Note: this draws on the article “Sticks and stones: Comparing Twitter campaigning strategies in the European Union referendum“, published in The British Journal of Politics and International Relations.

About the Author

Simon Usherwood is Reader in Politics at the University of Surrey.




Print Friendly, PDF & Email

About the author


Posted In: Brexit | Featured


Creative Commons Attribution-NonCommercial-NoDerivs 3.0 Unported
This work by British Politics and Policy at LSE is licensed under a Creative Commons Attribution-NonCommercial-NoDerivs 3.0 Unported.