Why Do Mainstream Parties Lose Either Way on Accommodating the Far Right?

Moving right on immigration can win voters, but the gains may not survive the losses

Today’s post is a guest essay by Hanno Hilbig, a political scientist at UC Davis who does some of the most interesting work on political behavior. I’m excited to publish his assessment of the accommodation literature, which has become especially controversial lately. Too often, the debate gets flattened into either “moving right on immigration never works” or “mainstream parties should do it.” Hanno’s piece gives the more nuanced useful answer: accommodation can move some voters, but it also creates hard trade-offs around credibility, base defection, and issue salience.

---

In May 2025, Britain’s Keir Starmer warned of an “island of strangers” and rolled out a sweeping immigration white paper. A few months earlier, Friedrich Merz had pushed a five-point migration motion through Germany’s parliament with the votes of the far-right Alternative for Germany (AfD). Both leaders reached for a familiar playbook: move closer to restrictive immigration voters, reduce the appeal of the radical right, and hope to gain back voters drifting away from the mainstream.

The temptation is easy to understand, as mainstream parties often do not align with their own voters on the issue of immigration. As Laurenz Guenther documents, large representation gaps exist on cultural issues across Europe, with immigration among the clearest examples. In many countries, majorities of voters favor lower immigration, while mainstream politicians, including those in center-right parties, hold more permissive views. Such a gap looks like an opening, and so far the radical right has often been the party to exploit this gap.

Whether this kind of mainstream “accommodation” wins voters back remains an open empirical question. The research has not settled it yet. Existing evidence suggests that restrictive shifts can attract some anti-immigration or radical-right voters. But those gains can be offset by losses among pro-immigration voters and the mainstream party’s own base. Accommodation can also make immigration more salient, look less credible than the radical right, and make once-fringe positions seem more normal.

Parties usually turn to accommodation only when they are already losing. That is part of why it is so hard to judge: the studies finding it rarely works are looking at parties that were sinking anyway. Accommodation can win some voters back, but only when the shift looks credible and the party keeps its own base. When it can’t, the party just ends up boosting the rival it set out to beat, which is why mainstream parties so often lose either way.

Accommodation as spatial voting

So what is accommodation? In practice, accommodation usually involves one or more elements: policy movement, cooperation with other parties on legislation, and rhetorical signals. Merz’s case involved substantive movement on migration and a controversy over the norm not to cooperate with the radical-right AfD party. Starmer’s case combined an official policy white paper with rhetoric that made immigration highly salient.

The basic case for accommodation stems from what political scientists call “spatial voting.” In the simplest version, it means that we can imagine that voters and parties sit on a certain policy line. Voters prefer the party closest to them. Parties can convert some voters by moving toward them but may lose voters that they are moving away from. If voters want more restrictive immigration policy than mainstream parties currently offer, this creates an opening for radical-right parties. In turn, spatial voting predicts that a mainstream party may win back some voters by moving toward those voters in policy terms. Since mainstream politicians are relatively distant from some voters in terms of immigration positions, this suggests an opportunity to regain voters by closing the representation gap.

But several processes are not captured in a simple spatial model. First, accommodation on issues like immigration can make these issues more salient or important to voters. This could then benefit radical-right parties who “own” the issue of immigration.

Second, accommodative positions by mainstream parties may not be seen as credible. Voters may consider such positions to be “cheap talk.” This could be the case when accommodation is proposed by some party officials but criticized by others. It could further occur when accommodative rhetoric directly contradicts a long track record of policy positions of the party that proposes it.

Third, accommodation may reduce the stigma around radical-right parties or illiberal political ideas. This can happen because mainstream cues signal that positions, parties, or forms of rhetoric previously treated as outside ordinary democratic competition are now acceptable options for ordinary voters. There is a tradeoff between short-term electoral gains and long-term normalization of radical positions.

What We Should Expect

My discussion focuses on programmatic accommodation: a mainstream party’s substantive movement toward more restrictive positions on immigration, asylum, border control, integration, or immigrant welfare.

The spatial-voting expectation. The pro-accommodation case starts with a “representation gap.” If mainstream parties are more liberal on immigration than many voters, radical-right parties can attract voters who are to the right of the mainstream. Recent work supports this possibility. Guenther (2025) argues that representation gaps are especially large on cultural issues and that right-wing populists fill those gaps. In Germany, the immigration gap is especially relevant to AfD support among voters who consider immigration important. Alizade (2025) makes a related point: concerns about immigrant crime can create a mismatch between left voters and left parties, pushing some voters toward the center right.

This logic gives accommodation its strongest theoretical foundation. If some voters are close to mainstream parties on most issues but more restrictive on immigration, a mainstream party might gain by reducing that distance. In this view, accommodation is a response to underrepresentation rather than an attempt to pander to radical-right voters.

Salience and issue ownership. Spatial voting is useful, but it leaves out several things that matter in campaigns. Movement on immigration can make immigration more central. That matters because radical-right parties often benefit when politics is organized around immigration, national identity, or crime.

Immigration was already highly salient when Starmer made his 2025 pivot. In May 2025, Ipsos found that 49% of Britons named immigration as one of the biggest issues facing the country, the highest level since the 2016 Brexit referendum. At the EU level, immigration was also among the top concerns in recent Eurobarometer data.

High salience does not necessarily doom accommodation. If immigration is already unavoidable, a mainstream party may prefer to compete on it rather than ignore it. But if accommodation makes immigration more salient, it can change the payoff of the strategy. Issue ownership means that voters see one party as more competent or authentic on an issue. If voters see the radical right as more credible on immigration, a mainstream party may end up making the radical right’s strongest issue more important.

Credibility and policy delivery. Accommodation also requires credibility. A mainstream party with a long record of pro-immigration policies and visible internal divisions on the issue may struggle to convince voters that its pivot is real and durable. The problem is especially severe for rhetorical accommodation. Voters may hear a restrictive speech and conclude that the party is merely opportunistic, especially if the party’s record points in a different direction.

A U.S. example is Kamala Harris’s 2024 pivot on border security. Harris emphasized border security and tougher asylum rules while still supporting comprehensive immigration reform. Substantively, this moved her campaign toward a more restrictive border position. Politically, it also created a credibility problem, because the shift came after years in which the Biden administration was associated with a more permissive border and asylum posture. Voters may therefore have perceived the move as late electoral repositioning rather than a durable commitment.

Policy change may be more credible than slogans, but delivery creates its own problems. Visible change can keep immigration in the news. Invisible change may not be noticed. Failed delivery can reinforce the perception that mainstream parties talk a lot but do little.

Where the Lost Voters Go

Similar vote movements can have different consequences across party systems. In a bloc-based proportional system, a social-democratic party might lose pro-immigration voters to Greens, liberals, or other left parties while gaining anti-immigration voters from the radical right. If these voters remain inside the left bloc, accommodation may improve the bloc’s chance of governing even if the main party loses some supporters. This is the logic behind the more favorable Danish case in Hjorth and Larsen (2022).

But this logic does not always apply. If alienated voters abstain or move across blocs, losses are not contained. In a majoritarian system, a voter leaving Labour for the Greens or Liberal Democrats may still hurt Labour’s ability to win seats, even if Labour gains some voters who favor more restrictive immigration policies. Radical-right parties can also respond. They can emphasize mainstream parties’ lack of credibility, shift the debate to stricter positions, or argue that their own ideas have been vindicated.

Accommodation may also change what counts as legitimate politics. The concern is not only that mainstream parties lose or gain voters in the short run. It is also that their cues reduce the stigma around radical-right parties or illiberal ideas. Evidence for this first-stage effect appears in several settings. Valentim et al. (2025) show that anti-immigrant statements by mainstream-right politicians erode anti-prejudice norms more than similar statements by radical-right politicians. Daur (2025) shows that mainstream legitimization can increase the perceived legitimacy of pariah parties. Ekholm et al. (2022) show that mainstream recognition can increase sympathy toward the Sweden Democrats.

A U.S. example is the erosion of the distinction between legal and illegal immigration. On some accounts, Trump did not create the underlying restrictionist constituency, but he recognized it and pushed it beyond prior boundaries. The result was not only a tougher stance on illegal immigration, but a broader weakening of the old norm that kept legal immigration politically protected. This is the normalization mechanism in a different setting: once the boundary of acceptable restrictionism moves, later political entrepreneurs can push farther at lower reputational cost.

So, the net effect is hard to predict. Accommodation can work through several channels at once. It can persuade voters who see the mainstream party’s shift as credible. It can make immigration more salient for voters who trust the radical right more on the issue. It can also fail if voters view the pivot as opportunistic.

What Does the Evidence Say?

Researchers have made progress on the relationship between accommodation and electoral returns. But the evidence is genuinely hard to read, for a simple reason: parties rarely accommodate at random. They usually move right on immigration when radical-right parties are already gaining, when immigration is already salient, or when public opinion has shifted. If radical-right support then changes, it is hard to know whether accommodation caused the change or simply followed the same trend.

The cross-national evidence is mixed and does not support a pro-accommodation story. Krause et al. (2023) find little evidence that mainstream accommodation reduces radical-right support, and some evidence that it may increase movement to the radical right. Dahlstrom and Sundell (2012) find, in observational evidence from Swedish municipalities, that tougher average mainstream positions are associated with higher Sweden Democrat support. Down and Han (2020) do not directly measure legitimacy, but their voter-level evidence is consistent with a legitimization channel: restrictive mainstream positioning is associated with more radical-right voting among voters who do not already see the radical right as competent on immigration.

Other studies are more favorable. Spoon and Klüver (2020) find that mainstream-left parties benefit electorally from going tougher on immigration, while mainstream-right accommodation is largely null. Thesen (2025) reanalyzes part of this literature with controls for immigration news and radical-right media visibility, and finds that once those media variables are included, accommodation appears to hinder radical-right success.¹

The strongest causal evidence comes from designs that separate accommodation from the trends that usually precede it. Conjoint designs ask respondents to choose between profiles whose attributes are randomly varied. Survey experiments randomly assign information about party positions. Unexpected-event designs use real political shocks that occur during survey fieldwork. These designs are narrower than cross-national studies, but they are closer to the central question: whether a mainstream party’s restrictive movement changes vote choice, vote intention, or propensity to vote.

Chou et al. (2021) study German voters using a conjoint design that varies candidate positions on immigration and other issues. Restrictive mainstream positions can attract some AfD voters, which supports the spatial argument. But the same positions can alienate mainstream-party voters. The net party-level effect depends on whether gains among the target voters exceed losses elsewhere.

Hjorth and Larsen (2022) study the Danish Social Democrats after the 2019 election, when the party’s immigration position was plausibly ambiguous to voters. Their survey experiment told some respondents that the party had taken an accommodative immigration position and others that it had taken an adversarial one. The study finds that accommodation can attract anti-immigration and Danish People’s Party voters. Losses among pro-immigration voters were partly contained because many of those voters remained inside the left bloc. This is the clearest case in which accommodation looks electorally useful, but it also shows why institutional context matters.

Kollberg (2026) separates positional accommodation from populist rhetoric in a pre-registered factorial survey experiment. The central result is that position matters more than populist style. Accommodation increases propensity to vote among right-leaning voters but repels left and centrist voters. Again, the spatial mechanism is present, but so is the offsetting-loss mechanism.

An important recent non-experimental case is the preprint by Turnbull-Dugarte et al. (2025). They use the timing of British Election Study fieldwork around Starmer’s May 2025 speech. Some respondents were interviewed just before the speech and others just after it, so the episode works like an unexpected-event treatment. That makes the design unusually close to the question here: what happens after a real social-democratic accommodation episode? It is still not a pure test of policy delivery, since the treatment combines rhetoric, policy announcement, and media attention. The paper finds that exposure made Labour look more anti-immigration and right-leaning, reduced Labour support, and did not reduce Reform UK support.

Overall, the strongest causal designs suggest that accommodation can move some anti-immigration or radical-right voters, but these gains are often offset by losses among centrist and pro-immigration voters. Experiments also hold constant many parts of real campaigns: media attention, elite conflict, credibility, and radical-right responses. That makes the evidence useful but incomplete, and how much you trust it comes down to how much weight you put on a clean causal estimate versus the realism of an actual campaign.

Interlude: What Researchers Should Do Next

Before concluding, I would like to add a few thoughts on what ambitious grad students and other folks who care about this issue can do next. The main evidence gap is pretty clear at this point. We need more designs in which mainstream accommodation is the shock and both mainstream and radical-right electoral outcomes are observed in real elections. The key challenge is to separate accommodation from the political trends that usually precede it, such as rising immigration salience, radical-right growth, or shifts in public opinion.

Cross-national research will remain valuable, but even the best versions face the same problem. Parties accommodate when the political environment is already changing, and that endogeneity is difficult to solve with aggregate data alone.

One promising direction is subnational or candidate-level research. Close elections between candidates with clear programmatic differences on immigration, asylum, policing, or integration could provide useful evidence, especially when the office has visible policy authority. MPs, mayors, and local party branches also sometimes take positions that differ from their national party. These cases create more variation than national party-level studies, although national party brands may still drown out local differences.

Another promising direction is unexpected-event designs. Survey fieldwork sometimes overlaps with high-profile party pivots or policy shocks. These designs capture real-world accommodation without asking voters to imagine a hypothetical campaign. But they depend on the case that happens to occur. They cannot tell us what the effect would have been if the pivot had been more credible, or if issue salience had been held constant.

What Accommodation Can and Cannot Do for You

The spatial voting argument remains the strongest case for accommodation. If mainstream parties are more liberal on immigration than many voters (which they often are!), a more restrictive position can reduce the distance between the mainstream party and those voters. The evidence reviewed above suggests that this works. Accommodation can make mainstream parties more attractive to some anti-immigration voters. But a spatial gain among some voters is not the same as a positive net gain after factoring in the “side effects” of accommodation.

Recent arguments that “mimicking” the radical right is electorally self-defeating capture an important concern. Accommodation can make immigration more salient, activate an issue on which the radical right is often viewed as more credible, alienate centrist and pro-immigration supporters, appear opportunistic when it conflicts with a party’s prior record, and weaken the boundary between mainstream and radical-right positions.

But the broader conclusion that accommodation cannot work is too strong. The evidence points to a more conditional claim: accommodation can win some voters – but the shift has to be credible, voters must be persuadable, and losses elsewhere must be contained. The Danish case is the one most often associated with these favorable conditions.

The skeptical view is that these conditions are demanding and may often not hold. That skepticism is important, but it does not make the strategy irrational from the perspective of parties that choose it. Mainstream parties often accommodate from a position of weakness, after immigration is already salient and after some voters have already begun to defect. Mainstream parties may therefore choose accommodation even when they know it is risky. If defeat already looks likely, accommodation is less a path back to power than a bet to slow the bleeding. And that is the deeper bind: by the time accommodation looks tempting, a mainstream party is rarely choosing between winning and losing. It is choosing between two ways of losing, and betting that accommodation is the slower one.

1. The disagreement reflects both measurement and design: studies differ in what they count as accommodation, whether they separate position from salience, and whether they focus on mainstream-party support or radical-right support.