The politics of asylum: How local exposure reshaped UK voting

The UK has seen a sharp rise in asylum applications since the Covid-19 pandemic. This column examines how exposure to asylum seekers affects voter behaviour. The authors find that increases in dispersed asylum seekers – who are provided with accommodation but cannot choose their destination – lead to a sizeable rightward shift in local election outcomes. While migration has become more salient for both of the UK’s main political parties, only Conservative MPs appear to effectively channel local discontent, helping explain why these rightward electoral shifts occur despite little observable change in public attitudes.

The UK has seen a sharp rise in asylum applications since the Covid-19 pandemic, reaching about 100,000 annually in 2023–25, compared with 30,000–40,000 in the previous two decades. Around half of applicants now enter irregularly, mostly via ‘small boats’ coming across the English Channel – a recent development that has drawn intense media and political scrutiny. Together with concerns over hostel accommodation, these arrivals have gained substantial salience within an emboldened anti-immigration movement, culminating last September in one of the largest xenophobic demonstrations in recent UK history. While such movements routinely scapegoat recent arrivals – especially if undocumented or seeking asylum – it remains unclear whether their electoral success stems chiefly from making the issue salient or from voters’ own negative reactions to direct contact with asylum seekers and other migrants in their local areas.

In other countries, influential studies have examined how exposure to asylum seekers and refugees shapes voting behaviour, particularly during the 2015–16 European refugee crisis (Dustmann et al. 2019, Hangartner et al. 2019, Steinmayr 2021, Campo et al. 2024). As in the broader literature on immigration and electoral outcomes, results are mixed: some studies document shifts towards right-wing and far-right parties, whereas others find no effects or even gains for pro-immigration parties (Cools et al. 2021, Alesina and Tabellini,2024). We contribute to this debate in a new paper (Fasani et al. 2025), offering the first assessment of how local exposure to asylum seekers affects electoral outcomes in the UK – a major destination for refugee and migrant flows and a context in which migration and asylum feature prominently in public discourse (Hatton 2017).

Politics and the dispersal policy in the UK

The UK’s plurality electoral system operates within a multi-party landscape dominated by Labour (centre-left) and the Conservative Party (centre-right), which have alternated in government for the past century. Over our study period (2004–19), Labour governed until 2010, followed by a Conservative–Liberal Democrat coalition, and from 2015 a single-party Conservative government.

Since the early 2000s, asylum seekers have been allocated across the country through a dispersal policy in which the central government provides accommodation, but claimants cannot choose their destination. Those with sufficient resources or networks may opt out, forfeiting housing support; fewer than 20% did so during our period. This opt-out margin enables us to compare the electoral effects of two groups – dispersed and non-dispersed asylum seekers – subject to different placement regimes within the same national context.  Figure 1 illustrates their contrasting spatial patterns. Dispersed asylum seekers (left) are concentrated in a limited set of Local Authorities (LAs), whereas non-dispersed claimants (right), free to choose their residence, are more widely distributed and more likely to settle in the wealthiest areas of southern England.

Figure 1  Dispersed and non-dispersed asylum seekers, 2004-2019

Source: Fasani et al. (2025)

Identifying causal effects: Beyond the dispersal policy

In our empirical analysis, we examine whether within-LA changes in the number of dispersed asylum seekers affect voting outcomes. A simple fixed-effects approach cannot yield causal estimates due to two main endogeneity concerns: (i) unobserved local trends that influence both asylum placement and voting behaviour, and (ii) potential reverse causality if asylum seekers avoid areas exhibiting rising hostility.

Although dispersed asylum seekers are assigned housing on a no-choice basis, we show that the allocation is not orthogonal to local political conditions. LAs with Labour administrations were disproportionately selected when the scheme began, and these patterns persisted over time – as often suggested by the British media. Our identification strategy, therefore, does not simply rely on the dispersal policy. We develop an original instrumental variable strategy that predicts the local presence of asylum seekers using predetermined public housing characteristics interacted with national inflows of asylum claims. Specifically, we use the pre-existing stock of low-quality public housing – commonly used to host asylum seekers – to forecast subsequent dispersal. Importantly, the instrument is orthogonal to baseline electoral performance. For non-dispersed asylum seekers, whose residential choices are endogenous, we employ a standard shift–share instrument based on historical settlement patterns of migrants from the same origin countries (Card 2001).

British voters shifting to the right

We find that increases in dispersed asylum seekers lead to a sizeable rightward shift in local election outcomes: a one standard deviation rise widens the Conservative–Labour vote-share gap by 3.1 percentage points in favour of the Conservatives (Figure 2, blue bar). Party-specific estimates indicate that this pattern reflects a significant electoral setback for the Labour Party, accompanied by gains for Conservative candidates. Electoral gains are observed also for the Green Party and UKIP, although these do not fully translate – unlike from the two major parties – into council seats, owing to the majoritarian electoral system. We detect no effect on turnout. The political response intensifies after 2010, when the Conservatives entered government: effects become even more favourable to the Conservatives, suggesting a genuine rightward shift in voter preferences rather than an anti-incumbent backlash. Importantly, we find no electoral effect of non-dispersed asylum seekers (Figure 2, red bar), plausibly because they are less visible – living in private rather than concentrated public housing – and typically better resourced, attracting little local attention.

Similar rightward shifts appear in the five general elections held during our study period and in British Election Study panel data, which allow us to track within-individual changes in voting intentions. We also estimate higher support for ‘Leave’ in the Brexit referendum in areas more exposed to dispersed asylum seekers, complementing existing evidence on other drivers such as trade shocks (Colantone and Stanig 2016) and austerity (Fetzer 2019).

Figure 2 Effect of asylum seekers on the Conservative-Labour vote share gap in local elections

Note: Two-stage least squares estimates.
Source: Fasani et al. (2025)

Mechanisms: More hostile attitudes or heightened salience?

To understand the mechanisms underlying our main results, we examine both demand-side and supply-side channels.

On the demand side, we consider voters’ attitudes. Using longitudinal survey data on natives’ economic, cultural, and welfare-related concerns, we test whether local exposure to asylum seekers heightens anti-immigrant sentiment. We find no significant effects on average attitudes and only slight declines in the most positive cultural and welfare evaluations. These results align with evidence that Europeans’ views on migrants and refugees are highly stable over time (Kustov et al. 2021, Bansak et al., 2023).

Figure 3 Salience of migration/asylum issues in MPs’ speeches, 2002-2020

Source: Fasani et al. (2025)

We next examine political supply using a dataset of all parliamentary speeches from 2004–19. After identifying migration-related speeches and linking them to MPs’ constituencies, we construct a time-varying measure of local migration salience. While migration has become more salient for both parties over time (Figure 3), our estimates show that only Conservative MPs react to local exposure: those representing areas with more dispersed asylum seekers place greater emphasis on migration, whereas Labour MPs do not (Figure 4, blue bars). Notably, our findings for Conservative MPs’ speeches closely mirror those for British citizens. First, we find effects for dispersed but not for non-dispersed asylum seekers (Figure 4, red bars). Second, content analysis reveals no systematic shift in speech content (measured by sentiment, pro-immigration stance, and universalism), which does not become systematically more hostile to immigrants and refugees. We interpret these findings as evidence that Conservative MPs effectively channel local discontent, helping explain why relatively modest increases in asylum exposure generate sizeable rightward electoral shifts despite little observable change in public attitudes.

Figure 4 Effect of asylum seekers on asylum/migration salience in MPs’ speeches (2SLS estimates)

Note: Two-stage least squares estimates.
Source: Fasani et al. (2025)

Conclusions

Although our study concludes in 2019, recent UK developments can be read through its lens. Since then, the Conservative Party has entered a marked decline – cycling through three prime ministers in four years and losing the 2024 election to Labour. Yet Labour’s resurgence has proved short-lived: by late 2025, Starmer’s approval ratings had collapsed, and latest polling show the two historically dominant parties commanding only about 35% of the vote combined. Support has partially shifted to the Greens and Liberal Democrats (at 13%-15%), but far more dramatically to Reform UK under Nigel Farage’s strongly anti-immigration platform (now above 30%). Whether this shift will prove durable is unclear, but it suggests that efforts by mainstream parties to channel anti-immigration discontent – previously effective for the Conservatives and recently attempted by Labour – are no longer succeeding, leaving hostile narratives as strong as ever.

Source : VOXeu