Brexit: national identity and ethnicity in the referendum

In the referendum on EU membership held on 23 June, local authorities with higher proportions of people identifying as ‘English only’ recorded higher vote shares to leave the EU, while polling data suggest that ethnic minorities were more likely to vote Remain. Is there a relationship between national identity and ethnicity on the one hand, and votes in the referendum on the other?

‘English only’ and ethnic minorities

National identity and ethnicity are not the same: ethnic identities refer to smaller cultural groups within countries, while national identities refer to countries as a whole. But they are both community identities that tend to summarise something about the groups people feel part of, such as ideas, traditions, institutions and relationships.

While nationality is an objective legal category, national and ethnic identities are both self-defined. They are commonly measured in social surveys. A question about national identity was first introduced in the Census in 2011. People in the UK can select from a range of (combinations of) national identities, but high level categories are English, Welsh, Scottish or Northern Irish only, British only, or Other. In England, 60% identified as English only and 19% as British only in 2011. Ethnicity in the UK is measured in different levels of detail, but high level categories are White, Mixed, Asian, Black and Other.

National and ethnic identities are probably related: people from ethnic minorities are more likely to identify as ‘British only’, while white respondents are more likely to identify as ‘English only’.

National identities by ethnicity
British only English only Other UK
White 14% 64% 17%
Indian 58% 12% 5%
Pakistani 63% 15% 6%
Bangladeshi 71% 8% 5%
Black African 43% 10% 6%
Black Caribbean 55% 26% 7%

Source: A portrait of modern Britain, Policy Exchange, 2014, p34 (Census 2011)

Local authorities: national identities and Leave votes

The charts below show that English local authorities with higher proportions of people who gave their national identity as ‘British only’ in the 2011 Census were more likely to record lower vote shares for Leave. English local authorities with higher proportions of people who classified their national identity as ‘English only’ were more likely to record higher vote shares for Leave.

National identity

Source: Electoral Commission, EU Referendum results; Census 2011, available from Nomis

Note that these charts do not necessarily support the conclusion that identifying as English indicates a form of nationalism that desires the sovereignty and controlled borders the Leave campaign proclaimed. After all, they only show a relationship between English identities and aggregated votes at the local authority level: they do not offer individual level explanations.

And even at the local authority level, the national identity effect is not as strong as it appears at first sight. Comparing the relative contribution of eight factors to the proportion of Leave votes at local authority level in England in a multiple regression model shows that English national identity does make a small contribution, as do age and the proportion of people working in the public sector. The biggest predictive factor by far, however, is the proportion of non-graduates. Economic and other factors were not found to have any predictive value.

Ethnicity and Remain votes

At the same time, available data at individual level does suggest some relationship between ethnicity and voting patterns. As the poll is anonymous, there is no official data recording how individuals have voted, but opinion polls give insight into voting intentions among different ethnic groups.

Most opinion polls published before the referendum do not include information on voting intention by ethnicity. However, such information is included consistently in 8 ORB polls released since January 2016. The sample sizes of these polls were small, so the chart below shows the average across all 8 polls. When asked how they would vote, regardless of whether they thought they would actually vote, all ethnic minorities were more likely to vote Remain than white respondents.

Voting intention by ethnicity

Source: ORB polls on voting intentions in the EU referendum, available from www.whatUKthinks.org

The British Election Study carried out after the May 2015 election also found ethnic minorities were more likely to vote to stay in the EU (although it found different support rates for different minority groups).

If there was a referendum on Britain’s membership of the EU, how would you vote?
White Mixed Black South Asian* Other Prefer not to say
Stay in the EU 46% 59% 61% 49% 44% 41%
Leave the EU 35% 23% 17% 22% 30% 28%
I would not vote 3% 4% 7% 8% 11% 7%
Don’t know 16% 14% 15% 21% 16% 25%
N        18,423 220 343 506 244 239
* Indian, Pakistani, Bangladeshi
Source: British Election Study, 2015, panel

Again, this does not prove that ethnicity is causally related to voting behaviour. But it does suggest some relationship, particularly as other factors that could be thought to influence votes to leave the EU among ethnic minorities do not operate as we might expect. For example, while Pakistani and Bangladeshi groups are more likely than white groups to have no degree level qualification, they were more likely to vote Remain.

National identity, ethnicity and values

While national and ethnic identities (and education) probably do not exclusively determine how people voted, they may have played a role in as much as they are themselves associated with a set of social values and attitudes.

Lord Ashcroft’s polling found that Leave voters and Remain voters took opposite views on the desirability of issues like multiculturalism, immigration, social liberalism, feminism and the internet. Similarly, Eric Kaufmann suggested in a recent blog that personal values related to order and openness help explain Brexit votes. This raises the question to what extent national and ethnic identities (as well as other factors such as education and socio-economic background) are related to such values, and whether this relationship is stronger for some identities than others.

However, it seems at least plausible to suggest that not all ethnic identities are strongly associated with positive attitudes towards feminism and social liberalism, and therefore with Remain votes. A more simple explanation for higher levels of support for Remain among ethnic minorities could be that immigration became an issue of contention during the referendum campaign.