If the 2015 UK General Election was – in the words of Alex Salmond – about the roar of the Scottish Lion – The EU Referendum is essentially all about England, and more specifically, Englishness. Between 80 and 90% of the votes cast in this Referendum will be English ones, and on this vital issue England is a house divided, not so much between red and blue, but between young and old, the university educated, and school leavers. Demographic markers such as age and education offer useful rules of thumb about what groups will vote remain or leave, but what these statistics don’t really tell us is why older voters tend towards leave, or why the better educated tend towards in.

The growing divide in British politics, and certainly the ground on which the Referendum will be won or lost, is between those who stand to gain from a Britain that is globalised, interconnected, and whose economy is market-driven and those who feel threatened by these same trends. People without a graduate education, the majority of whom don’t work in white-collar jobs, are the people most likely to have their wages undercut by foreign labour. Those who describe themselves as being only of “fair health” worry about access to healthcare and other public services in the face of large scale immigration. More tellingly many who describe themselves as Christian feel they are culturally distinct from a Britain that has become increasingly multicultural and secular. Likewise for those that describe themselves as English only, rather than British or any other mixed national identity. What unites groups leaning heavily towards Brexit is a sense that their interests are under threat, whether cultural or economic. That Brexit is often presented as a panacea for those feeling the pressure of modern Britain should therefore come as no surprise.

Chart 1

The British Election Study, a gold standard face-to-face survey carried out in late 2015, and which correctly estimated the eventual Conservative lead over Labour at the General Election, corroborates the theory of how people’s sense of national identity relates strongly to their attitudes to Europe. Those who consider themselves “British not English” are three times more likely to say they will vote remain rather than leave, while those who consider themselves “English not British” are almost twice as likely to vote leave than remain.

Chart 2

Information on national identity is one of the few non-demographic pieces of information collected about the UK population via the National Census. As such it’s an important part of the arsenal of data that political strategists and academics use to create models which predict where potential voters might be. At a constituency-level, you might try working out how Eurosceptic it is by looking at the UKIP vote share as a proxy, but if you want a more accurate read of the political composition of a smaller area, for which there’s no election information, non-political data, primarily about age, education, national identity, health and ethnicity is needed to build your model.

Building an algorithm to identify the principal component or “X Factor” which best underpins the variables we have identified as being most closely associated with voting UKIP is much more efficient than trawling through the UK Census looking at every demographic variable available for every area. We have used such an algorithm to produce the component which does the best job at explaining the differences in education level, the number of professionals, Christians, Whites, those of fair health, The results are below.

Chart 3

Chart 4

Chart 5

When analysed by the seat winner a number of interesting patterns emerge. Firstly, when the seat is a Labour one, the level of UKIP support is very closely related to the demographics of the area, as represented by whether there is a high/low component score. The only exception to this clear relationship are a cluster of Merseyside seats which show very low UKIP support, even when their demography would indicate much higher support for Mr Farage’s party. Liverpool gives a sense of what Labour’s support would be like in the rest of the country in a political landscape absent UKIP.

Chart 6

The pattern is more complicated in Conservative-held seats. The relationship between how “Ukippy/Eurosceptic” an area is (as indicated by the component), and the actual UKIP vote share is much more disparate than it is in Labour seats. There are a number of plausible reasons for this. First of all, it could be that the Conservatives were at least partially successful in preventing more defectors to UKIP than the demographics of certain seats would otherwise indicate, particularly in the marginals. Secondly, it could suggest that the politics and appeal of certain individual Conservative MPs had a real effect on potential UKIP defectors. Either way the Conservatives have done a better job of protecting their flank against the appeal of English nnationalism than Labour. This though could be a temporary advantage.

In Scotland, Labour’s collapse has been relatively sudden – since 2010 – and largely at the hands of the SNP. In England, Labour’s decline has been more gradual and has benefitted both UKIP and the Conservatives. Labour’s support among those who describe themselves as “English only” has significantly eroded since 2005 in places like the South West, Kent, Essex, East Anglia, market-towns in the Midlands and vast tracts of rural Cumbria, Northumberland and North Yorkshire. In fact no other variable explains the decline of Labour in England better. Movements such as Blue Labour, and the rallying cries from Labour figures such as John Denham and Tristram Hunt imploring their party to embrace Englishness have come 10 years too late. Even as Labour has solidified its hold in more urban, diverse areas, in which a far greater proportion of people describe themselves as British, its support has declined almost everywhere else.

Chart 7Chart 8

The English Question won’t be resolved by the EU Referendum vote on the 23rd June. Up to 60% of England’s population identify as being “English only”, and will in all likelihood vote for Brexit in far greater numbers than not. Despite this, they are politically fluid and haven’t to date voted for any one party en masse, as such, which party they migrate to next, if at all, will be key in determining the future contours of British politics.