ronjohnstonDr David Manley

Kelvyn Jones

In their recent analysis, Ron Johnston, Kelvyn Jones and David Manley used a large body of YouGov polling data to explore which social groups are most likely to vote Leave in the upcoming EU Referendum and where they live, producing a clear geography of support for Brexit. Here, they explore how differential turnout rates across those groups could alter the pattern of support for the Leave campaign, which suggests where the Remain campaign might most effectively target its campaigning resources.

The pattern displayed in our initial study on the EU referendum assumed that everybody would turn out to vote – see Map 1. That is, of course, extremely unlikely and the experience of recent British general elections suggests that turnout will probably be no more than two-thirds of those registered to vote – unless, that is, the electorate is as motivated by this issue as were the Scots regarding the possibility of their country becoming independent two years ago. Further, in general turnout is lowest among the young and those with fewest qualifications. Because the former group tend to be those most likely to support Remain, their level of turnout could have a crucial influence on the overall outcome.

To establish the extent of this possibility we modified the data deployed in our earlier exercise by applying differential turnout rates to the 25 age-by-qualifications groups. The rates applied were derived from the 2015 British Election Study face-to-face survey, in which turnout was verified by checking the marked-up electoral rolls to see whether each of the 2,987 respondents actually voted (overall 6 per cent of those who said they had voted did not according to the registers, for example).

The table shows the percentage turnout within each age group according to their qualifications – which increase across the columns from those with no qualifications to those with Level 4 (those with degrees or the equivalent). These data show clear differences by age group; among those with no qualifications, for example, the turnout rate among the oldest age group was double that for the youngest. Furthermore, within the youngest age group those with Level 4 qualifications were twice as likely to turn out at the general election as those with none.

Table: Turnout by Age and Qualification (per cent)

table 1

In our previous exercise, we calculated how many people in each age-by-qualifications group reported that they would vote either Leave or Remain, or Did Not Know how they would vote. For this updated exercise we took those estimated numbers and then applied the turnout percentages in the above table. (If, for example, we estimated that 100,000 individuals aged 18-24 with no qualifications would vote to Leave, we reduced this to 35 per cent of that total, or 35,000.) These reduced totals were then summed to identify the total number of anticipated votes for Remain and Leave in each local authority.

If everybody voted on 23 June, we estimated from the YouGov data that Leave would have a majority over Remain in 61 of the 380 local authorities. With application of the turnout estimates, Leave would have a majority over Remain in 74 authorities – an increase of over one-fifth of the places.

Our second map shows both the greater number of authorities where Leave would ‘win’ compared to the situation if everybody were to vote on 23 June and the pattern of change in the Leave vote. On average, across Great Britain as a whole, the reduced turnout increased support for Leave by 2.6 per cent – largely because of higher than average abstention rates among young voters. But that percentage does not  apply equally across the country. Instead, some of the largest increases in support for Leave are in London and central Scotland, where in general support was low. But the change was also high elsewhere, as in Ceredigion – Aberystwyth having been identified in other research as the town giving Remain its greatest level of support. In other words, where support for the Remain campaign appeared strongest, low anticipated turnout rates – especially among the young – resulted in the greatest increases in support for Brexit.

Given that our estimated turnout rates were lowest among the young, it is not therefore not surprising that the places with the greatest increases in support for Brexit when turnout was added to our models were in local authorities with large numbers of students. This is made very clear by the following list of the eighteen places with the greatest percentage increases in estimated voting to Leave the EU between the two models:

table 2 cities

All have large student populations, relative to their total size. In the past, students living in properties managed by universities and colleges (such as halls of residence) were registered by their ‘landlords’. This no longer happens, and those students must register individually. Many undoubtedly have not – and will not in the short time in which they can before registration for the referendum closes unless they are mobilised to do so – and because of this it may well be that we have overestimated the turnout rates by young people in those towns and cities.

In the referendum every vote counts, and each of the campaigning organisations is working hard to get its supporters out. But much recent research has shown that however powerful the media campaigns, contact between the protagonists and the individual voters is most likely to ensure high turnout. The main campaign referendum organisations have limited resources and so will want to target their groundwork on places where their position is most likely to prevail. For the Remain campaign the message from the analyses reported here is very clear: focus on the university and college towns and cities where there are large numbers of potential Remain votes among students and other young people. With limited resources for a ground campaign, focusing on getting them to register and then vote would be the most efficient strategy. For Leave, on the other hand, their supporters are more likely to vote in any case – and negative campaigning in places where they don’t want a high turnout is likely to be counter-productive!

Geography, operating through differential support and differential turnout, could therefore be crucial to the referendum outcome.

Map 1. The geography of the percentage likely to vote Leave is everybody turns out.

Brexit Map 1

Map 2. Change in the percentage likely to vote Leave with reduced turnout rates.

Brexit Map 2

Note: This arcitle first appeared on British Politics and Policy blog. The authors are grateful to Joe Twyman and Anthony Wells of YouGov for facilitating this experiment in modelling voting outcomes by providing the polling data. More details on the modeling procedure can be obtained from the authors: R.Johnston@bristol.ac.uk

Ron Johnston is Professor in the School of Geographical Sciences at the University of Bristol.

Kelvyn Jones is Professor in the School of Geographical Sciences at the University of Bristol.

David Manley is a senior lecturer in the School of Geographical Sciences at the University of Bristol.

 

 

Print Friendly