ELECTION 2017: Fact Check - if 30% more people under 25 vote, could the Conservatives lose the election?

An article on The Independent is being widely shared on social media suggesting that a 30% increase in turnout among 18 to 24-year-olds would make the election of a majority Conservative government on June 8 rather unlikely. It followed a tweet by Alan Firth, a linguist at Newcastle University, commenting on an article by the vice president of the National Union of Students, Shelly Asquith. When contacted by The Conversation, Firth said that the calculations made in The Independent article reflected his own.

For the purpose of this fact check, we will look at whether the Conservatives could lose if turnout of those under 25 were to increase by 30 percentage points. To check whether this really could happen, we need to answer four questions.

First, what are the turnout rates among younger and older voters? The 2015 British Election Study (BES) shows that 57% of 18 to 24-year-olds claimed to have voted in 2015, compared to 76% of people over 25. These figures overestimate actual turnout rates because some people say they voted when they did not (and all surveys tend to over-sample voters).

Because of this, the BES “validates” votes for some people: it checks whether people actually voted. When it did this for 2015, it found that 10% of 18 to 24-year-olds claimed to have voted but did not, while 3% of people over 25 did the same. So if we correct the turnout figures to reflect this, then 47% of 18 to 24-year-olds voted at the last election compared to 73% of those over 25. This gives us a good indication that there is a real difference, a bit shy of 30%, between older and younger voters’ likelihood to vote.

The second question is whether young people and older people vote very differently. Again, the 2015 BES is helpful for showing the longstanding differences between age groups in Britain in vote choice. The first graph shows that younger people are less likely to vote Conservative than older people.

The third question is whether those who did not vote would behave the same way as those that did? The second graph shows the parties that non-voters in 2015 said they would have supported if they had voted, excluding those who said “don’t know”. It appears that the Conservatives actually have less potential support among young non-voters than they do among young voters.

On the face of it then, increasing turnout rates by 30 percentage points among 18 to 24-year-olds should bring voters into the electorate who are quite unlikely to vote Conservative.

But the impact of increasing turnout for those under 25 depends on what proportion of the electorate they make up – my final question. According to the 2011 census, 18 to 24-year-olds make up less than 12% of the electorate.

Imagine that turnout did increase for this group by 30 percentage points, and imagine that only 16% of those previous non-voters voted as they said they would after 2015 and opted for the Conservatives. The Conservatives would get a lower share of the vote, but this effect would be fairly small: the party’s overall share of the vote would fall by slightly over one per cent. This is because anything that affects only 3.6% of the electorate (30% of the 12% of the electorate under 25) can never lead to large shifts in aggregate vote shares.

On the whole, younger non-voters, like everyone else, are also probably more likely to vote Conservative than they were in 2015. A sizeable minority would also pick a party which will at best win one or two seats (according to the 2015 BES, 19% of under-25s voted for UKIP, the Greens or another small party). Taking this into account, the actual effect in first past the post constituency contests would be even smaller.

Verdict

Even on the generous assumptions here, there really is no way that increasing turnout among such a small number of people – however distinctive their party preferences – can make much difference to an election in which the Conservatives have a poll lead of nearly 20 percentage points over their nearest rival.

Review

Ben Bowman, Teaching Fellow in Comparative Politics, University of Bath

Predicting elections based on polls is a tricky business, but the calculations and arguments made here are logical. Like The Independent, I’m not entirely sure where Alan Firth has got his 30% claim from because I don’t see it in the original article by Shelly Asquith.

Is such a rise possible? About 64% of young people voted in the EU referendum, up around 20% on recent general elections. A further increase would require grassroots organisation to include the most marginalised: 25% of school leavers have fallen off the electoral roll since registration rules were changed in 2014, with black and minority ethnic communities hit hardest.

A word of caution is required, then: a 30% rise would require a groundbreaking social movement, but it would bring along older voters, too. I do not see such a movement yet, and so I agree with the author’s conclusion nationally, and locally as well. Labour will gain about twice as many young votes as the Conservatives but has not done enough to organise this support. For instance, a lot of students vote (69% of students voted in the 2015 general election) and their votes could swing many marginal seats. However, 60% vote in their home constituencies rather than at university and so their vote is diluted. If young people are to be Labour’s base, organising students should be bread-and-butter stuff for party organisers. They might still do it, but the clock is ticking.

Until then, this fact check holds water. The big swing in this election has been UKIP voters to the Conservatives, and there just aren’t enough young people to counterbalance that without a broader movement to Labour and organisation in key marginals.

This article originally appeared in The Conversation

The Conversation is checking claims made by public figures and in the public domain. Statements are checked by an academic with expertise in the area. A second academic expert then reviews an anonymous copy of the article. Please get in touch if you spot a claim you would like us to check by emailing us at [email protected]. Please include the statement you would like us to check, the date it was made, and a link if possible.