A flurry of voter registrations by people under-35 in the final few days before the deadline on May 22nd has got everyone talking about how the influence this group could have on the election result.
The pollsters are at odds about how large a role young people will play, as Peter Kellner, the former president of YouGov, explained in the Evening Standard. At the time of writing different organisations were giving a wide range of final predictions for the Conservative lead, after various adjustments to their raw data. When simply weighting for demographics and political interest, the data from these different organisations gave a narrow range of 5-7 points. But once adjustments for turnout were made, the range of predictions for the Conservative lead widened to 6-14 points.
So, why the disparity? After the failures of the polls in 2015, the pollsters are taking very different approaches to accounting for turnout across age groups. In short, this time around more young people are saying they will vote than did in 2015 - some pollsters believe it (increasing the Labour vote) and others don’t.
So who is right? Are we likely to see a rise in turnout among younger voters? Looking back over 25 years of data from the British Social Attitudes survey gives us a good grasp of longer-term trends in voting habits among this demographic.
General elections - the historic record
Before examining turnout figures, it is important to note that the estimates generated by surveys for turnout are typically higher than the true figures (in the case of British Social Attitudes, by an average of 7 percentage points). This is because people who agree to take part in surveys are also more likely to vote in elections. While these estimates should be treated with some caution, they provide reliable way of looking at trends and comparing voting behaviour between different age groups.
Looking at the historical trends, turnout dropped across all age groups between 1997 and 2001 where it has remained relatively stable ever since, with one exception. Among 18-34 year olds, since 2005, reported turnout has increased by six percentage points.
Voting – a civic duty?
We have been asking the public whether they think it is their civic duty to vote for several decades. Eighteen to thirty four year-olds have always been the least likely to consider voting a civic duty. In 2015, 56% said it was, compared with 76% of over-55s. However, there has been an upturn in political engagement among younger voters since the low of 2008/9 in the wake of the financial crisis and the parliamentary expenses scandal. But there has also been a similar upturn among older age groups.
Interest in politics
Another attitude that is strongly associated with turnout is general interest in politics. Among 18-34 year olds in 2015, this figure was at its highest level since 1991. Twenty seven per cent of this age group have quite a lot or a great deal of interest in politics. However, like attitudes to voting, interest in politics is at a high levels across all age groups, so any associated rise in turnout may be mirrored in other demographics.
What will happen on polling day?
Younger people are still way behind older generations in their commitment to voting. According to our estimates, the gap in turnout between the youngest and oldest age groups was still a chasmic 27 percentage points in 2015.
While 18-34 year olds are the only age group to have seen an increase in turnout since 2005, other measures of political engagement do not appear to have shifted any faster for this group than for older voters.
However, the most recent data presented here were collected in summer 2015, before Corbyn’s election as Labour leader, and of course, before the vote to leave the EU and everything that has happened since. A lot has changed.
In recent history, we have witnessed shifts in youth turnout of as much as 12% between general elections (the fall between the 1997 and 2001), so it would be irrational to rule out a similar swing entirely. As always, with our first past the post voting system an increase in turnout by as little as 1%, in the right constituencies, could have a significant impact on the result.
Follow NatCen on Twitter: @NatCen.