How Britain voted in the 2017 election

As we have for every general election since 1979, Ipsos has produced estimates of how the voters voted in 2017.  As always, it is important to note that these are estimates only, based on people’s answers to pre-election surveys during the campaign.  However, the figures have been calibrated to match the final actual results and turnout at a regional level, which should make them a more accurate guide to how different sub-groups voted. 

Note on turnout:

As we have noted before, estimating turnout is one of the hardest challenges when relying solely on survey data.  Despite the changes we have introduced to our methodology since 2015, which we believe have made our samples more representative, polls may still be more likely to interview politically engaged people than those who are disengaged, people may over-estimate their likelihood of voting, and they may think they are registered when in fact they are not.  For post-election analysis we have the advantage of knowing how many people turned out in the end, but we still have to identify them in our data on the basis of the answers they gave before the election.  This means that the turnout estimates given below should be treated with particular caution, including taking into account the voter validation results from the British Election Study when these are published (in previous elections these have shown a similar pattern to our estimates).  Two turnout figures have been provided.  The first expresses turnout as a % of all resident adults, which is what our sample is based on.  This gives a figure lower than the “official turnout” figures usually used, which is a % of the number of names on the electoral register (as not everyone in Britain is registered, and some people may be registered twice even though they can only vote once). The two measures should not be confused¹.  We have also provided a figure for all those registered, however for the reasons expressed here we believe the first figure is both more reliable and more meaningful. 

Here are ten key findings from the results:

1. Age was even more of a dividing factor than in 2015 (and the biggest we’ve seen since our records began in 1979).  All the swing to Labour was among under 44s (and highest of all among 25-34s), while there was a swing to the Conservatives among over 55s.  This is the biggest age gap we’ve seen in elections going back to the 1970s.  Although (as in previous elections) the swing among women and men overall was similar, there was a difference between young men and young women.  Among 18-24 year olds, Labour increased its vote share much more among women than men.
2. The middle classes swung to Labour, while working classes swung to the Conservatives – each party achieving record scores.Although the Conservatives maintained a six-point lead among ABC1s, Labour increased its vote share among this group by 12 points since 2015.  Similarly, while Labour had a four-point lead among C2DEs, and increased its vote share among this group, this was eclipsed by the 12-point increase for the Conservatives.  This is simultaneously Labour’s best score among ABC1s going back to 1979, and the Conservatives’ best score among C2DEs since then.
3. Education represents another clear divide.  The Conservatives had a large 17-point lead among those with no qualifications, and a smaller seven-point lead among those educated to below degree level.  Among graduates, though, Labour had a 15-point lead.
4. The Conservatives fell further behind Labour among BME voters.  Labour’s lead among BME voters increased to 54 points, a swing of six points since 2015 – mainly because the Conservative vote share among BMEs fell (Labour’s vote share increased even more among white voters, but the Conservative share also rose among this group).  Turnout among BMEs also increased six points since 2015. 
5. The Conservatives and Labour both held on to around nine-tenths of their 2015 vote.  However, only half of 2015 Liberal Democrat voters stuck with the party this time (30% went to Labour, 15% to the Conservatives), while six in ten UKIP voters from 2015 switched to Theresa May’s party (only two in ten voted UKIP again).
6. Labour had a lead among 2016 Remain voters, the Conservatives had a lead among 2016 Leave voters.  Labour led the Conservatives by 54% to 26% among Remainers; the Conservatives beat Labour by 65% to 24% among Leavers
7. Not many people who didn’t vote in 2015 or 2016 voted in the 2017 election, but those who did mostly chose Labour.  Only around one in eight of those who did not vote (excluding those who were too young) in 2015 or in the 2016 referendum voted this year, but around six in ten of those who did voted for Labour.
8. The Liberal Democrat share was mostly stable at the topline (although this hides significant churn), but the Greens lost popularity among young people.  Around half the Liberal Democrat vote came from their 2015 supporters; another three in ten roughly equally from 2015 Labour and Conservative voters.  Despite this, the profile of the Liberal Democrat vote is little changed – they did best among graduates (12%) and remain voters (13%), and worst among those with no qualifications.  The Green Party’s vote share among 18-34s, though, fell five points since 2015.
9. UKIP’s vote share fell across the board, mostly among its biggest supporters from 2015. Only among male DEs did its share reach 5%, while UKIP’s biggest falls were among older people and the working classes.
10. Compared to 2015, turnout rose most among young people, to match their estimated turnout levels in the EU referendum, while it fell very slightly among older people – but older people were still much more likely to vote overall.  As noted above, caution should be taken in interpreting these turnout estimates, but our estimates suggest just over half of all resident 18-34 year olds voted², very similar to our estimates of turnout among young people in the 2016 EU Referendum.  Compared with the last general election in 2015, turnout among the population as whole rose by 16 points among 18-24 year olds, and by eight points among 25-34 year olds.  Turnout fell slightly by three points among those aged 65+ since 2015.  Turnout also rose among social classes C1 and C2, but fell slightly among men in social class AB. 

(Use arrow keys or fingers to scroll left and right in the table)

(Use arrow keys or fingers to scroll left and right in the table)

Note: The EU referendum figures were updated at 4pm on 20/06/2017.

Technical note

Base: 7,505 GB adults aged 18+ (of whom we have classified 5,255 as likely voters, using the same definitions as in previous election estimates), interviewed by telephone and online between 21 April – 7 June 2017. The proportions of voters for each party and non-voters was then weighted to the actual results by region. The data were also weighted to the population profile of Great Britain.