Skip to content
The UK's only independent think tank devoted to higher education.

So did students have any impact on the election?

  • 18 December 2019
  • By Nick Hillman

Many people thought students could boost the election results for Jeremy Corbyn’s Labour Party. Over time, as the proper academic studies of the election drip out, we will come to know for certain if that happened or not.

In the meantime, we have undertaken a quick-and-dirty exercise to test the proposition:

  • we have taken the 20 constituencies thought to have the most full-time students; and
  • we have then calculated the swing between the top two parties in these areas (and the Labour Party is in the top two in every seat).

This should provide a rough sense of whether students played the role assigned to them in advance.

In every one of these seats, there was a swing away from Labour to the other party placed in the top two, which was generally but not always the Conservatives. Yet it was not enough to make a difference: none of these seats changed hands, even though some were targeted by other parties (such as by the Greens in Bristol West and the Conservatives in Canterbury).

The psephological concept of an electoral swing is contested for lots of good reasons, and there are some other huge caveats with this exercise too.

  1. These 20 seats are unrepresentative of the country as a whole. They include no seats in London, no seats in Northern Ireland and no seats in rural areas.
  2. The data for student numbers in each constituency are old, poor and not brilliantly well suited to this purpose. Some seats will have more students than we think; some may have less.
  3. There is no perfect way to measure the electoral impact of students, and this way is far from perfect. In particular, comparing the results between one election and the subsequent only shows extra swings between elections. If students were already impactful in 2017 and still impactful in 2019, it wouldn’t show up as a change in the numbers.

Nonetheless, the results are intriguing because they generally show a lower Labour-to-Conservative swing than occurred between the two parties overall (4.6%). In one seat, Canterbury, there was even a swing from the Conservatives to Labour. This suggests that once a constituency has broken a deeply engrained habit of voting for one party, you cannot assume it will be a one-off. It also confirms that you need to look at seat-by-seat results to understand any general election fully.

While none of these 20 seats with high numbers of students changed hands, it is unclear how much this was down to the presence of students. Other factors, like the tendency for people in large cities (which are more likely to have universities) to vote for Labour, may be just as – if not more – important. As we pointed out before the election, for students to matter in an election, lots of boxes need to be ticked, such as whether the seat is a marginal one.


The embers of Labour’s defeat are now being pored over for clues on how they might do better next time. It would be wrong to assume that appealing even more to students is likely to boost Labour significantly at the next election, at least with regard to these seats. This is because, despite the general swing away from Labour, Labour held on to all 18 out of 20 that they already held, with the two Scottish seats staying in the hands of the SNP. When you already hold 90 per cent of the most student-dominated seats, there isn’t much further room for improvement.

Indeed, if anything, our tentative results support the idea that Labour’s problem is among less well-educated older people than it is more well-educated younger people.

Postscript: After posting this, David Kernohan of Wonkhe, whose statistical abilities are in a different ballpark to mine, very kindly got in touch with an early Christmas present of a Tableau for the numbers across the UK. It is worth looking at in detail and David summarised the findings as ‘your conclusions hold up across the UK’.

Leave a Reply

Your email address will not be published. Required fields are marked *