Facebook and the UK General Election

Facebook was a key battleground in the UK’s 2017 General Election. Whether it was Jeremy Corbyn’s Labour party using it to motivate new voters or Theresa May’s Conservatives trying to target swing voters with aggressive ad messages, everyone wanted to get their message out in channels under their own control.

In this post I’m going to compare the use of Facebook by the parties, and in particular, the Conservatives and Labour.

With around two billion monthly users worldwide and around 32 million accounts in the UK, Facebook’s size makes it an attractive target for political actors to communicate with voters. As the former Executive Director of the Stronger In campaign, Will Straw, pointed out in the Financial Times the social network is: “on a par with the BBC for getting your message out there ….The difference is with the BBC you don’t have control of the end product.”

The ability to disintermediate political journalists and deliver your message untainted and direct to voters provides a strong motivating factor for campaigns to engage with the network. It is also clear, from figures published by Enders Analysis, that demographic groups who are more likely to vote are less likely to use Facebook. Despite the high level of usage of Facebook among UK voters, it is more prevalent among the under-40s.

Measuring the campaign

The growth in the number of followers of parties and party leaders on Facebook during the course of the campaign gives an indication of the impact and engagement with the digital campaigns.

To measure this I recorded the number of people who had liked the Facebook pages for the main political parties and their leaders at the start of the campaign and again at 10pm on the 8th of June.

I defined main parties as those contesting seats in Great Britain but not Northern Ireland; that is the Conservatives, Labour, the Liberal Democrats, the Scottish National Party, the Green Party, UKIP, and Plaid Cymru. This matched the parties invited to take part in televised debates through the campaign. It should be noted in the below table that The Green Party has two leaders, Caroline Lucas and Jonathan Bartley.

Name Likes on 19/4/17 Likes on 08/06/17 Change  + Change % +
Caroline Lucas 72,907 81,347 8,440 11.58%
Conservatives 565,915 629,277 63,362 11.20%
Green Party 278,732 303,168 24,436 8.77%
Jeremy Corbyn 839,332 1,138,239 298,907 35.61%
Jonathan Bartley 7,661 9,082 1,421 18.55%
Labour Party 543,241 956,915 413,674 76.15%
Leanne Wood 31,071 32,447 1,376 4.43%
Liberal Democrats 161,513 185,049 23,536 14.57%
Nicola Sturgeon 293,704 299,346 5,642 1.92%
Paul Nuttall 46,802 49,389 2,587 5.53%
Plaid Cymru 26,426 28,912 2,486 9.41%
Scottish National Party 276,253 286,798 10,545 3.82%
Theresa May 343,562 419,094 75,532 21.98%
Tim Farron 30,823 37,458 6,635 21.53%
UKIP 582,364 596,109 13,745 2.36%

Figure 1 – Change in Facebook likes during the campaign

This demonstrates the dramatic growth in Facebook likes for Jeremy Corbyn, up by more than 35%, and the Labour Party, up 71%. This is despite a strong starting position.

While there was growth for all the parties and leaders across the campaign it is notable that the performance of Labour and Corbyn considerably outstripped their rivals; The Conservatives, for example, rose 11%, Theresa May gained almost 22%.

The table also shows that UKIP’s digital penetration had stalled. Despite starting as the most popular party account, by the end of the campaign both Labour and the Conservative accounts had surpassed it.

During the seven week short campaign I recorded all the posts by the seven main parties. The below figure shows total Facebook posts between the announcement of the election and closing of polls at 10pm on 8th of June, it includes reposts and shared posts.

amiretreat17 presentation

Figure 2 – Total Facebook posts during short campaign

This figure clearly demonstrates the considerable lead in Facebook page activity shown by the Labour Party over its competitors.

This trend is also replicated in the number of posts made by party leaders, with Labour leader Jeremy Corbyn far exceeding his competitors. Although Plaid Cymru leader Leanne Wood uses Facebook far more consistently than her party’s main account.

leader posts

Figure 3 – Total Facebook posts by party leader

The pattern of activity can also be seen by examining the number of posts by day across the course of the campaign. The below figure is the number of posts per day made by the Conservative Party’s main Facebook page.

Con post frequency

Figure 4 – Conservative post frequency by day

 

The broad trend of this activity is reflected in many of the accounts activity by day. The trend was to increase the activity across the course of the campaign, with the most intensive activity in the final week. Most party accounts had their most active day on the 8th of June, when followers were exhorted to vote and also encourage others to do the same by sharing supportive messages.

All the party accounts reflected the pauses in campaigning made after the 22nd of May suicide bombing at the Ariana Grande concert in Manchester and the 3rd of June ISIS inspired attack at London Bridge.

However, when comparing all the parties’ Facebook posts by day the trend for Labour to outperform its competitors is clear.

All party post frequncy

Figure 5 – Party Facebook posts by day

Having recorded the posts by the political parties, I decided to conduct a content analysis of the Facebook activity.

The diagrams below illustrate the share of media type by the Conservative party and Labour party main Facebook accounts.

Con media type

Figure 7 – Conservative Facebook posts by media type

 

There are a number of trends illustrated here. Firstly, the Conservative party account made heavy use of political advertising, using both graphics and video. This, alongside the relatively sparse number of posts, gave the account a quite static, although polished, tone. Some of these political adverts were attack ads, adverts produced with the sole purpose of attacking the policies or personalities of their opponent, and some were supportive of Conservative party policy.

There was also a significant concentration on the figure of the leader of the party, Theresa May, during the campaign. In the run up to the short campaign Mrs May had outperformed her chief rival, Jeremy Corbyn, on measures connected to leadership. As such the party chose to concentrate on her in both native campaign videos and in videos lifted from broadcast television.

The Labour party’s posts reveal a different set of trends.

Lab media type

Figure 8 – Labour party Facebook posts by media type

 

Labour party election strategists had two clear drivers for digital media strategy. Firstly, they believed that when voters saw Jeremy Corbyn during the short campaign, unfiltered by the perceived biases of the media, that the electorate would find him an engaging and sympathetic figure. To that end much of the focus during the campaign posts was on Corbyn as a leader and personality.

The second strategic driver was to counter negative reporting of Labour’s policy positions. Labour communications team took the view that they had to aggressively counter the media’s framing of their policies. As such, a significant proportion of the Labour party posts were animations that explained policy positions on topics such as student tuition fees or the NHS.

There was also a significant policy difference between the Conservatives and Labour in the areas highlighted by Facebook posts. The Conservative leader, Theresa May, made it clear in her Downing Street speech that she viewed Brexit as the clear motivating factor for the election saying “Division in Westminster will risk our ability to make a success of Brexit and it will cause damaging uncertainty and instability to the country. So we need a general election and we need one now, because we have at this moment a one-off chance to get this done while the European Union agrees its negotiating position and before the detailed talks begin.”

While sceptics may have felt that there was a significant element of calculation involved in kicking what appeared to be a weak opponent while they were down, the posts in the short campaign were clearly dominated by Brexit. When the campaign was rattled by the misfiring of the manifesto and the backlash over social care policy, it was to Brexit that it returned after the pause in campaigning triggered by the Manchester bombing.

Con policy areas

Figure 9 – Conservative Facebook posts by policy area

 

Other election themes became objects of ridicule given the oft-repeated formulations of “strong and stable” leadership, often contrasted with the “coalition of chaos”  that Jeremy Corbyn would seek to lead. This was reflected in the significant number of Facebook posts that focused on leadership and the horse race, the jockeying for position and coalition deal-making the Conservatives predicted would happen if there was a hung Parliament or narrow Labour victory.

What is also noticeable is the lack of focus by the Conservatives on areas the party had previously identified as election winning themes, in particular tax, the economy and business. During the 2015 General Election campaign, the Conservatives concentrated on a few key themes. Of the 65 videos the party posted, 19 were about the economy and another 17 were about the horse race. This reflected two of the big campaign themes for the Tories – that they had a long-term economic plan and could be trusted over Labour to deliver on it; and that Labour would bring parties such as the SNP into government in order to form a coalition. Labour too concentrated on its key messages 18 of the 84 videos it posted were about the NHS.

In 2017, the Conservative campaign tried not to pin down policy specifics, partly as a result of feeling hemmed in by the policy commitments of David Cameron and George Osborne in the run up to the 2015 election. The desire for flexibility combined with a sense hubris about their inevitable victory, led to a lack of clarity around the detail of policies and a defensive approach to a policy agenda at a time when voters were growing weary of austerity policies and more receptive to a change narrative.

This was perhaps most clearly illustrated by the reaction to a change in social care policy, quickly branded the dementia tax, which would see voters pay for care using money defrayed against the value of their home, which would then be sold upon their death. This key policy was briefed to the press the night before the manifesto launch and the furious reaction to it dominated the coverage, eventually prompting an apparent volte face just four days later in a press conference where Theresa May repeatedly claimed that “nothing has changed”.

The Labour party, on the other hand, tackled a range of policy areas, notably leading with a hardy perennial, the National Health Service. The NHS usually features highly in Labour campaigns, either trumpeted as the recipient of extra investment under a Labour government or with voters being warned of the dangers of privatisation or collapse under a Conservative one. This campaign was no different.

Labour policcy areas

Figure 10 – Labour party Facebook posts by policy area

What is perhaps more significant is the amount of time the Labour party Facebook feed spent on issues connected to its campaign and voting. Labour strategists wanted to energise voters who had drifted away from the party since 1997’s New Labour landslide. Either by voting for other left-wing parties such as the Greens, or because they had stopped voting altogether. They were confident the campaign and Jeremy Corbyn could also energise younger voters, whose turnout had steadily declined since the 1990s.

As such, a significant proportion of the Facebook posts were dedicated to encouraging people to register to vote and to getting the vote out, celebrity endorsements of Labour, and campaign activity by Mr Corbyn. Events and speeches were planned in many safe Labour seats and Mr Corbyn was given a hero’s welcome at events in places such as Tranmere. Sceptics suggested he was preaching to the converted but videos of these events built into a powerful narrative of a social movement gaining widespread popular support. A Facebook live video of the 6th of June Corbyn election event in Birmingham compered by the comic actor Steve Coogan and featuring a set by the dance act Clean Bandit was watched an astonishing 2.3 million times by the time polls closed.

The aggressive use of feel-good videos, policy explainers, and shareable media helped drive forward the Labour party’s social media strategy. Unlike other parties Labour repeatedly reposted material, rather than merely posting once on a topic and moving on. This meant that in total just 35% of Labour’s 545 posts were original, whereas 65% of them were reposts. This strategy of repeatedly reposting content in order to reach the maximum number of users means that Labour content outperformed Conservative party content by a factor of at least two to one. According to social engagement analytics of News Whip in the month leading up to the vote, the Labour page pulled in 2.56 million engagements on 450 posts, while the Conservative page saw 1.07 million interactions on 116 posts.

The Facebook accounts of the leaders tended to have a similar but not identical policy mix to the official party accounts. Sometimes this was strategic; Theresa May’s account was far more positive than the Conservative account. All attack ads were posts on the Conservatives account whereas Mrs May’s tended to concentrate on positive messages about Brexit.

May policies

Figure 11 – Conservative party leader Theresa May’s Facebook posts by policy

 

Jeremy Corbyn’s account featured more campaigning material than the Labour party’s one, as well as piloting celebrity endorsements aimed at younger voters, which were picked up some days later by the official party account.

corbyn policies

Figure 12 – Labour leader Jeremy Corbyn’s Facebook posts by policy

At other times this reflected the character of the leader. Plaid Cymru leader, Leanne Wood, clearly operated her own Facebook account and its posts reflected her interests and thoughts. She also regularly engaged with voters in comments.

Use of video

While a significant proportion of the Conservatives’ video content featured attack ads, Labour’s tended to focus on the campaign and policies. Subsequent reporting has made it clear that Corbyn himself urged his team not to resort to negative campaigning, although he was able to do this at least in part because other political actors were doing it for him.

The videos proved to be highly effective. Viewing figures jumped enormously on the 2015 general election where a Labour party election broadcast, again featuring Steve Coogan, was the most popular video, watched 1.2 million times on Facebook. This time there was significant increase in watching of videos by both Conservative and Labour. It should be noted that a view is recorded by Facebook after a user has consumed three seconds of video.

 

Top 5 Facebook videos Content Party Views (per million)
1 Attack ad on Corbyn’s record on national security Conservative 7.9
2 Corbyn’s question to May on ITV’s Facebook live Labour 4.6
3 10 reasons to vote Labour animation Labour 4.4
4 Attack ad on May’s record on national security Labour 3
5 Attack ad on Abbott’s record on national security Conservative 2.9

Figure 13 – Most watched Facebook videos

 

It is notable that despite Jeremy Corbyn’s intention to resist the black arts of negative campaigning, Labour’s third most watched video was an attack ad on Theresa May’s security record, both as Prime Minister and Home Secretary. With a dramatic background music track and a reprise of Police Federation complaints on the impact of cuts on policing, its tacit allegation is clear, that through unthinking, uncaring and unnecessary austerity measures Mrs May allowed the Manchester and London Bridge terror attacks to take place.

Labour’s policy of maximising social engagement through organic mechanisms such as repeated posts delivered clear success in getting content shared and engaged with – News Whip analytics suggest that in the month preceding the election, the average Jeremy Corbyn video had 23,081 reactions versus the average Theresa May’s one with 11,842. However, organic sharing was not the only mechanism for driving users to content. Both Labour and the Conservatives spent a considerable amount of money on targeted advertising.

Otherer political actors

There were two other significant factors that helped support the Labour campaign on Facebook and both were from political actors that sat outside the party’s traditional structure.

The first was Momentum, the campaign group that had grown out of a youth focussed support group for Jeremy Corbyn’s leadership bid. Having fought two leadership elections in two years, it had developed a strong sense of how to use social media, and in particular, video to deliver to its followers positive messages with viral attributes. Humour was a key tool, the emotional response adding to the shareable nature of the content. With 24,000 activists across the country, it was able to mobilise support for Labour in previously hard to reach environments, for example southern cities with large student populations, such as Canterbury which returned its first ever Labour parliamentarian . According to the Momentum activist, Adam Peggs, during the final week of the election, the group’s Facebook videos were watched more than 23 million times by 12.7 million unique users.

A significant part of this was the due to the ‘Dad, do you hate me?’ film, which was watched around 7 million times in the final week of the campaign.

Momentum also had significant social media impact in areas Labour needed to win, including Cardiff, Derby, Sheffield, Canterbury and Plymouth. In the final week of the election, the group says 42.2% of Facebook users in Canterbury viewed its videos, while in Sheffield Hallam, where the former Deputy Prime Minister Nick Clegg was unceremoniously ejected, the percentage was 55.9%.

The other significant factor in Labour’s success was the growth of hyperpartisan political blogs. Sites such as The Canary, Another Angry Voice and the London Economic presented a relentlessly positive view of Jeremy Corbyn, an irredeemably damning view of his opponents, both outside and inside the party, and an intensely hostile view of the mainstream media. While partisan political blogging is not new, Guido Fawkes, Left Foot Forward and ConservativeHome have all been in existence since the New Labour years, the new hyperpartisan sites have achieved a level of success outside the wonkish circles of their forebears. As Buzzfeed’s UK Political Editor, Jim Waterson, has pointed out in a seminal article on the sites “The rise of the Alt-Left”, the use of graphics, clear and understandable writing and consistent tone often mirrors the tabloid press they hold in such contempt.

Electoral impact

It was clear from the moment that the exit poll was published at 10pm on 8th of June, that Theresa May’s electoral gamble had failed to pay off. An election she called to strengthen her hand had weakened it, almost certainly terminally, as she lost her majority.

The final result saw her lose 13 seats, while Labour gained 32. Mrs May was still able to form a minority government propped up by Northern Ireland’s Democratic Unionist Party, but far from being the electoral liability some of his MPs feared, Jeremy Corbyn again proved himself to be a formidable campaigner. In opinion polls published since the election his party has been running neck and neck with the Conservatives.

An examination, by the pollsters YouGov, of voting behaviour by age demonstrates the split in the electorate. The Conservatives only outperform Labour with voters aged more than 50.

Turnout was high at 69%, the highest since 1997, but lower than the 2016 European referendum when it reached 72%. There was considerable speculation in the immediate aftermath of the election about the size of the youth turnout, with some commentators attributing Labour’s success to a “Youthquake”. According to a survey carried out by the British Election Study and Ipsos Mori, youth turnout (those aged 18-24) was the only demographic to increase on the European Referendum with some 64% voting, up from 60% at the referendum and 43% at the 2015 election. This was clearly a substantial increase, but Labour’s message resonated further than Millennials.

The analysis of Facebook activity by the parties and other political actors leads me to draw a number of conclusions about its impact on the 2017 general election.

  1. Labour outperformed other parties on Facebook. More content was produced by the party and its supporters, and it generated more engagement with Facebook users than its rivals.
  2. Labour galvanised voters online and offline supported by radical media and activist groups. The Labour party’s strategy of promoting Jeremy Corbyn and his policies successfully engaged voters, including youth voters who had not voted in a General Election in such large numbers since 1992. The party was able to successfully extend its message beyond Facebook into real world events, which could be filmed, transmitted live, and used to engage more people via social media. Corbyn’s supporters in activist groups such as Momentum and the hyperpartisan media were able to successfully amplify these messages.
  3. Labour was able to neutralise negative mainstream media coverage by speaking directly to voters. It is notable that Labour felt no need to share supportive messages in the mainstream media with its followers. By curating large communities of interest online and using these to promote real-world events, Labour was able to bypass a mediated and predominately hostile press.
  4. Labour used Facebook to deliver positive, organic messages and tackle negative messaging. By concentrating on policy explanation and positive messaging about the political movement it was creating, press negativity about Corbyn was successfully neutralised. As the Conservatives struggled to explain complex policies to concerned voters on issues such as social care, Labour was able to cut the Gordian knot, with simpler messages that addressed voter concerns, for example the abolition of student tuition fees.
  5. Labour used paid ads effectively in marginals to target voters concerned by Conservative social care plans. Labour’s targeting of ads seems to have been more effective that the Conservatives, in both defending its own marginal as well as target seats. Conservative over-confidence may have played a key role in the failure to adjust expectations as the campaign developed, particularly after the unsuccessful manifesto launch.
  6. Those least likely to use social media were most likely to vote Conservative. Age was the key indicator of voting intention, as it is for likelihood to use social media.

This is a short version of a paper that I presented last month at the Journalism, Society and Politics in the Digital Media Era conference in Cyprus. 

Advertisements

Media predictions for 2016

2016 desktop calendar

What does the coming year hold for the media and what will the impact be on news and journalism?

2015 was a year of enormous change, and there’s no reason to expect 2016 to be any different. So, here are five trends I think will define the next 12 months.

1. Ad-blockers will go mainstream

The current status quo for digital advertising in media cannot continue.

Advertising is too intrusive. Splash screens, auto-playing video, and ads that scroll the screen are ruining the user experience.

On a desk-top this can be annoying but for mobile users it can destroy the user’s relationship with the publisher.

I’ve lost count of the number of times I have left an Independent article without reading it on my iPhone because the splash screen can’t be easily removed.

And now I don’t bother reading them at all.

This is no longer a niche experience – mobile is how people consume digital content. If your mobile experience is poor, you will lose audience.

The uptake in ad blockers has increased during the past couple of years. Early adopters have been using them to tailor their internet experience. 2016 will be the year their usage goes mainstream and the impact for publishers will be immense.

Business plans at almost all major publishers are based on delivering eyeballs to ads – even at those who also have a subscription strategy.

If the relationship between content and advertising breaks, then a new settlement will have to be found. Either by denying access to users of ad blockers, finding new forms of advertorial, or by publishers finding a way to manage their advertising in a way that is meaningful and useful for users.

If they can’t they risk being swept away.

2. A choice to make: destination vs distribution 

The content business is going to get a lot tougher in 2016.

The watchword of 2015 has been scale. Those that have it, want to retain it. Those that don’t, want to get it.

As the content bubble continues to deflate some publishers will go to the wall, others will sell up. Consolidation will be a key factor in 2016.

But the key decision for publishers will be whether or not to pursue a destination or distribution strategy.

Here’s the thing: when I talk to students about where they get their news, they invariably say from Twitter or Facebook. Of course, when you dig into this they actually get it from a publisher who is using social media to disseminate their content. But users don’t necessarily distinguish between the publisher and the distributor.

Brand and brand values don’t appear to survive the transition to social media. And that means users treat all information as being of equivalent value.

Look at the rise of fake news sites that publish stories simply to get eyeballs to their sites – how often are people taken in by them? Even journalism students, who should know better, are sometimes fooled.

Publishers have to decide where they want to focus their attention – destination or distribution?

The received wisdom of the past few years has been to emphasise the importance of being where your audience is.

But Netflix didn’t build its business on YouTube.

Why should publishers, so keen to take up Facebook Instants, build their business on other people’s platforms? Give up your brand, give up your revenue streams, give up your platforms, and you give up your business.

3. Innovation will move to the heart of the newsroom

Media businesses are generally pretty poor at innovation – as this excellently argued article by the Wall Street Journal’s Neal Mann explains.

All of us who’ve worked in jobs where we’ve tried to shake-up the existing way of doing things have encountered the same issues. Resistance to change, a culture that demands instant results driven by delivery to daily deadlines, a veneration of tradition at the expense of experimentation.

That has to change and I think newsroom managers will have to bite the bullet in 2016 and create bespoke innovation units.

Journalism and journalists will survive but if the media businesses of today want to have a future they need to embrace innovation as a process rather than always trying to copy ideas from others.

And that means experimenting with everything. Cool stuff can be done with virtual reality, 360° video, and automation. And who knows what else is around the corner? Who would have predicted in the first internet boom that mobile would be key technology of the future? The key question is how can these things be turned into innovative systems that will deliver consistently for users in the future?

If publishers take a structured and strategic approach to this, (experimenting, implementing ideas, measuring success and failure, focussing on the outputs not the processes, spreading success) they can build new products and new revenue streams that at the moment seem like pipedreams.

Of course that may mean setting up things like the Second Life bureau – but learning the lessons of failure is just as important as reaping the rewards of success.

4. Twitter will face an existential crisis

I love Twitter. For me, and for many journalists, it’s a fantastic tool. Filled with ideas, ever-changing, and rumbustious.

But it’s got a problem. It’s just not growing. And with the failure to match other social networks in scale comes a problem with attracting advertising. And that’s not going to solved by adding Moments.

In 2016 I think Twitter will face an existential crisis. What is it for? What is its future? How will it grow? How will it make money for its investors? I don’t pretend to have the answers to these questions, and I suspect no-one knows the answers, but unless Twitter is able to find a new strategy it faces a future of ossification, stagnation and decline.

5. Managing the decline in TV news

2015 was a pretty good year for TV news.

Levels of trust remain high, viewership of the main bulletins has been broadly stable, the election was still fought on TV rather than online and some high profile transfers and relaunches have garnered positive headlines.

But….

There is a long term problem for TV news. It’s not breaking through to younger viewers and, as older ones die off, it faces decline.

I expect that process to gather pace in 2016. Audiences will drop, perhaps not precipitously, but steadily. As viewers drift away from linear TV towards on-demand, the point of having an appointment to view TV news bulletin becomes less and less clear. And if younger viewers don’t pick up the habit of watching at 10pm – the format is doomed.

The TV companies recognise this risk. If James Harding was employed at the BBC in order to bolster its digital coverage, the always impressive Jonathan Munro was brought in to manage the TV coverage; to ensure that quality was maintained during a period of sustained decline in audience as the BBC experiments with different formats for its journalism.

This isn’t a prediction of imminent collapse but I think 2016 will be the year that the declining trend in audience and relevance for TV news becomes more clear.

So, five trend predictions for 2016. But, as they say, the only things certain in life are death and taxes.

Book review – Media talk and political elections in Europe and America

A book review I wrote for The European Journal of Communication here

The book is an edited collection using discourse analysis to examine everything from The Daily Show to Cleggmania and Obama’s email communications.

It’s edited by Mats Ekström and Andrew Tolson and published by Palgrave Macmillan.

I’m only just getting used to academic publication cycles. I wrote this six months ago and it has just been released – hence the reference to Ed Miliband having more time on his hands!

That’s a lesson learnt about time specific references dating your copy…

First attempts at media punditry

A couple of interviews done for BBC Radio Northampton’s Helen Blaby programme

Were the media right to use images of the murdered U.S. journalists? (Phone quality)

Favourite programmes on ITV as Radio Times launches a new poll

Content marketing for beginners

UN_TLIC_Logo_Black

A couple of posts done for the University of Northampton as part of its new social media marketing campaign

Should the paparazzi be allowed to stalk Prince George?

Wait before sharing the murder video of two U.S. journalists.