How did the EU referendum affect population movements?

New research shows that, after the EU referendum, people were less likely to move if they were aligned with the Brexit preferences of their district.

Major “polity-shaping” events, like the UK’s 2016 decision to leave the European Union (EU), can have important consequences on people’s attitudes and behaviour. The Brexit vote has deepened existing divisions in the British society over key issues, such as national identity, globalization and multiculturalism, and generated salient ‘affective polarization’, with individuals segregating themselves socially and distrusting people from the opposing side. At the same time, the referendum outcome caused a negative impact on people’s life satisfaction and mental distress (Powdthavee et al., 2019), and has even led to fallouts with family and friends (Edelman 2019 Trust Barometer).

In a recently published study in the journal Political Geography, Harry Pickard, Vincenzo Bove and I investigate whether such a hostile culture of ‘othering’ political rivals can affect broader social relationships by changing individuals’ propensity to migrate internally. Existing research has largely focused on the geographic sorting of the American electorate, and little attention has been given to other countries. And while one of our recent works shows that internal migration choices are strongly affected by political preferences also in the UK — and districts with the same political preferences exhibit higher migration flows — there is no evidence of the effect of Brexit as such. The unique circumstances of Brexit, with a near 50-50 vote in the referendum, and the fractious discussions around its future, make the EU referendum a particularly suitable test-bed to examine the consequences of polarizing politics on internal population movements.

There are two complementary behavioural explanations for our research question. First, we could expect the referendum outcome to make people more polarized. In fact, the national split revealed and reinforced by the Brexit vote has popularized the notion of a “divided Britain”, an expression frequently used to capture a growing sense of social and political polarization (Duffy et al., 2019). As Brexit identities have reinforced themselves in the aftermath of the referendum, often surpassing traditional party identities, political differences became more salient, affecting in turn internal migration decisions. Second, the referendum returns could have made more visible and consequential latent divisions that already existed in the population or provided new information about the aggregate political preferences of the district of residence. This could be particularly important in light of the high degree of uncertainty around the Brexit outcome and the forecasting errors of financial markets and opinion polls, which did not anticipate the victory of the Leave campaign.

We leverage comprehensive survey-based data for around 18,000 individuals from the UK Household Longitudinal Study (UKHLS), and examine whether the change in relocation patterns from before to after the referendum date of 23 June 2016 was different for different groups of people, depending on their ‘political alignment’ status – with the latter defined as having the same preferences over EU membership as the majority (at least 50%) of people residing in the same district. Our estimation strategy is designed to address ‘self-selectivity concerns’; that is, unobserved factors affecting both relocation decisions and people’s preferences towards Brexit.

Our analysis reveals that individuals’ migration choices are indeed influenced by deviations from community preferences. In particular, we find that UK citizens were significantly less likely to move to another district after the referendum when they were strongly aligned to their district of residence. At the same time, we find that this ‘alignment-induced’ effect is mostly driven by Remain supporters. On the one hand, as Remainers found themselves on the losing side, they were more likely to value the alignment to their district, given their ‘misalignment’ to the country as a whole. As those who preferred to remain in the EU became also worse off in terms of mental health (e.g., Powdthavee et al., 2019), they were less inclined to leave neighbours who share similar political values. On the other hand, as Leavers won nationally, we find that local alignment matters relatively less for them, and only when their district’s share of Leave votes is particularly high.

Our analysis also provides evidence that the main channel underpinning our results is the desire for ‘political homophily’; i.e., living in areas with political views similar to your own can satisfy your need for belonging and thus reduce the likelihood to relocate. Using information on the destination districts, we show that, after the referendum, non-aligned individuals were more likely to move to a district to which they could then feel aligned.

If the referendum served initially to polarize attitudes towards the EU, the subsequent Brexit process has served to ensure that the legacy of the referendum is still present despite the fact the polling day has long since come and gone. In fact, scholars and political commentators have increasingly warned about the increasing “tribalization” of British politics. Existing divisions do not only mirror divergences over the consequences of Brexit as such, but also in terms of people’s sense of identity and the values they uphold, given the implications of EU membership for cross-border migration and issues of sovereignty (Curtice et al., 2019). As such, in the long run, the effect of political preferences on geographic polarization could contribute to exposing and exacerbating divisions across the British society, and disrupt the efficient functioning of the labour market.

Increasing the local homogeneity of citizens’ political preferences can also increase the number of areas where election outcomes are not in doubt. A key concern in this respect is that by reinforcing the presence of politically homogeneous communities, deepening political divides do not only discourage the discussion of opposing viewpoints; they can also promote intolerance which can ultimately damage the social fabric of the country (Bishop, 2009).