Demography undermining the legitimacy of the 2016 referendum

Support for Remain was hightest among young voters and lowest amongst elderly in 2016. Young people coming onto the register and older voters dying is gradually changing the balance. Even if nothing else changes, these demographic changes will remove the majority for Leave towards the end of 2019. This erosion of the legitimacy of the 2016 vote is a strong argument for a People’s Vote.

Support for Remain was highest among young voters and lowest amongst the elderly in the 2016 referendum. Young people coming onto the register and older voters dying is gradually changing the balance. It will take only a few years to reverse the result. In among all the reasons for a People’s Vote, this erosion of the legitimacy of the 2016 vote is a powerful reason to ask voters again.

There’s been lots of discussion of whether the 2016 referendum result makes sense as “the will of the people”. Much of that has centred on lies from the Leave campaign, their apparent bending of election law and the possibility of Russian interference. But YouGov found that 71% of those under 24 who voted, voted for Remain, where among the over 65s that figure falls to 36%. The implication is that, even if no-one changes their mind and there’s no change in voter participation, Remain will be ahead of Leave by the end of 2019.

Writing soon after the referendum, David Howarth asked if it is reasonable for the elderly to bind the young like this.

In May 2018 The Express took up the story, and also naming serious concern over the long term legitimacy of the result.

The numbers

The poll in 2016 simply records the numbers voting Remain and Leave. Various organisations undertook polls soon after the referendum in which they asked people whether they voted, and if so, how. For the purposes of this post, I am drawing on a YouGov poll, offering some fine-grained data. I’m making four assumptions:

  • The number of people age 18 in the middle of 2017 is an adequate approximation for the number of people turning 18 in each of the next few years;
  • The number of over 65s whose deaths were registered in 2017 is an adequate approximation for the number who will die in each of the next few years (this includes people whose deaths were registered but who were not eligible to vote);
  • That among young people coming onto the register, support for Remain, and voter turnout will be the same as for the under 24s in 2016 (as support for Remain is highest among the young, this may under-estimate future support for Remain);
  • That among the over 65s the level of support for Leave will be the same among those who die each year as among this age group in 2016 (arguably this underestimates the highest number of deaths is among people in their mid 80s, who are likely to be more inclined to leave than people at the younger end of this age group);

The last two of these both suggest my numbers will be underestimates, but I don’t have data to quantify this.

  1. Number of people aged 18 in the middle of 2017: 766000 [source]
  2. Turnout among under 24s: 64% [source]
  3. Proportion of under 24s voting Remain: 71% [source]
    Proportion of under 24s voting Leave: 29% [source]

  4. Number of additional Remain votes from young people coming onto the register each year (766000 × 0.64 × 0.71): 348,070
  5. Number of additional Leave votes from young people coming onto the register each year (766000 × 0.64 × 0.29): 142,169
  6. Number of over 65s who died in 2017: 241885 men 261227 women, total: 503112 [source]
  7. Proportion of over 65s voting Remain: 36% [source]
  8. Proportion of over 65s voting Leave: 64% [source]
  9. Turnout among over 65s: 90% [source]
  10. Reduction in Remain vote from over 65s dying each year (503112 × 0.9 × 0.36): 163008
  11. Reduction in Leave vote from over 65s dying each year (503112 × 0.9 × 0.64): 289792
  12. So, each year, the Remain vote goes up by 221062, and the Leave vote goes down by 147623, making the total switch from Leave to Remain 368685.
  13. In 2017 the result was Remain: 16,141,241, Leave 17,410,742, giving a difference of 1,269,501. That would reverse in three years and five months — by December 2019.

Things undermining participation by young people

Two things worked together to reduce participation by young people in 2016.

One was the change in voter registration in 2014 so that individuals have to register rather than being registered by the place where they live. The reason for this was that individual voter registration is less susceptible to fraud. But an unintended consequence was that around 800,000 people dropped off the electoral register. There was a particularly sharp impact on students, with various University towns showing a drop in registration above 10%. The reason was that student accommodation had been registering all the students who were eligible to vote. The drop in registrations suggests that the message that students had to register themselves had not got out sufficiently.

Given the extend of disappointment among young people at the result, it seems likely that there would be more energy to register ahead of a People’s Vote.

The other problem was the poll took place soon after the end of University terms. Many students were registered where they were at university or college but had gone home before the vote. This meant they had to get a postal vote in order to vote in the referendum. Many hadn’t.

Taken together, this meant referendum messaging among students and to be “register to vote and get a postal vote”. To avoid recreating this problem, a People’s Vote would need to happen during term time.

To avoid disenfranchising young people, this means a People’s Vote would need to take place during term time. In its turn, this means it would be reasonable to anticipate a higher turnout amount the under 24s in a People’s Vote, leading to an increase Remain vote.

Rumours of low participation by young people

Soon after the referendum there were claims that there had been low participation by young people, with some claims that the figure was as low as 36%. That figure seems to have come from polling that confused the referendum and the 2015 General Election.

Among Leave voters on twitter I have seen claims that young people didn’t bother to vote, and that their apathy is to blame for not getting the result they wanted. The actual data points to a different conclusion, but the persistence of the rumour asks to be heard in terms of inter-generational rivalry. In effect, that is a claim that young people can be ignored. From the other side, a friend’s son, in his mid 20s responded to the referendum result by saying to his mother “your generation has stolen my generation’s future”. However it is seen, this is dangerous in terms of good connections across the generations

I know people in the 80+ bracket who have argued that they shouldn’t vote in a referendum of this sort, because they won’t be alive to face the consequences. While I don’t share the view that they should be excluded, this position is a reminder of the electorate’s responsibility to those who will have to live for longest with the consequences of a referendum decision.

Overseas and immigrant voters

Two other groups were excluded from the 2016 referendum: the majority of citizens of other EU nations living in the UK, and UK nationals who had been living abroad for more than 15 years. Both of these came about because the referendum used the rules for General Elections. While both cases can be argued for General Elections, those cases are not watertight. For a referendum they leak rather more.

It is possible to argue that a UK national who has been living abroad for a long time has ceased to be sufficiently involved in life in the UK to vote in General Elections, but their future is profoundly affected by whether they still have EU passports, which is a strong case for them to be able to vote. Denying the vote to citizens of other EU nations who are makeing a valuable contribution to the UK is also problematic — at its bluntest, if they responded by returning to their home countries, that would do serious damage to the UK.


Extending the voting to citizens of other EU nations living in the UK and to UK nationals who have been living abroad for more than 15 years can be argued. It would be controversial to Leave voters, as most of these would vote Remain, but there is a case.

The demographic trends suggest that the Leave vote will evaporate by the end of 2019 even if no-one changes their mind and voter turnouts don’t change. It would be sooner than that if a People’s Vote took place in term time. This means the legitimacy of the 2016 vote, already strained by stories of elector irregularities from the leave campaign, will have evaporated by the end of 2019. Some are claiming that a People’s Vote is somehow undemocratic because it challenges the 2016 result. In fact a People’s Vote is essential so that the decision can be said to be legitimate.

One thought on “Demography undermining the legitimacy of the 2016 referendum”

  1. The point about overseas voters, away for more than 15 years is an interesting point.

    They were promised to be refranchised in the Tory Party election manifesto of 2015 but nothing was done, until now.

    It has been included in the Elections Bill, due for its second reading on 07.09.2021 and includes a number of other controversial issues.

    The cynic in me makes me wonder if this particular part was included as a sop to attract parliamentary votes from those who otherwise would vote against it.

Leave a Reply

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