Shelter from the storm: External threat,social identity,and support for the EU

Abstract

Does the return of major warfare override identity-based constraints on European Union (EU) support? While post-functionalist theory suggests exclusive national identities act as a brake on integration, bellicist and social identity accounts argue that external threats generate functional demands, strengthening affective European orientations. Analyzing survey data from 16 countries collected 5 weeks after Russia’s 2022 invasion using a causal forest algorithm, we uncover a heterogeneous “rally” effect. Exclusive nationalists, typically EU skeptics, show the strongest positive association between threat perceptions and European pride, particularly in Central and Eastern Europe—a pattern mirrored in heightened levels of support for a common European army. We conclude that functional security demands can transform the EU from a perceived cultural threat into a necessary geopolitical shelter, even for traditional skeptics.

Keywords

EU support external threat exclusive national identity post-functionalism Russia–Ukraine war

Introduction

Does the return of major warfare to Europe override identity-based constraints on public support for the European Union (EU)? According to post-functionalist theory, the era when elites could advance European integration with little public scrutiny, the so-called “permissive consensus” (Lindberg and Scheingold, 1970), has given way to a “constraining dissensus,” where public opinion, particularly among citizens with strong national identities, acts as a brake on further integration (Hooghe and Marks, 2009). Existing accounts of European integration have long highlighted the tension between the functional requirements of centralization and the resilience of exclusive national identities (Hooghe and Marks, 2009). According to this perspective, while geopolitical crises may generate a functional demand for supranational governance, deeply rooted national identities impose strict limits on the EU’s legitimacy. Consequently, those who identify solely with the nation-state are generally considered resistant to the expansion of EU authority, impeding the formation of a polity (De Vries, 2018).

Russia’s full-scale invasion of Ukraine in 2022 provides a critical historical test for this assumption. The return of war to the continent has sparked a vigorous debate on whether external threats act as a “bellicist” driver of state building, a concept derived from Tilly's (1975) argument that warfare historically drove centralization and has played a decisive role in the emergence of the modern state by creating functional demands for collective defense (Kelemen and McNamara, 2022), or merely encourages intergovernmental coordination without altering fundamental citizen allegiances. Research on the effects of external threats on European attitudes pre-dates the 2022 full-scale invasion. Following Russia’s 2014 aggression in Crimea and Eastern Ukraine, scholars began documenting how threat exposure affects European identity and institutional trust (e.g. Gehring, 2022; Kiratli, 2024). The 2022 full-scale invasion intensified scholarly attention, with recent research documenting a “rally around the flag” effect characterized by immediate solidarity (Steiner et al., 2023) and a “social security logic” (Natili and Visconti, 2023). Yet, scholars caution that this consensus may be fragile (Truchlewski et al., 2023) or limited to loose coordination rather than full sovereignty transfer (Moise et al., 2025; Oana et al., 2025). Importantly, existing studies vary in their research designs and the claims they can support. Some exploit quasi-experimental or pre–post designs to establish causal effects: Gehring (2022) uses a difference-in-differences design, Malet and Hegewald (2025) employ an event-study approach with geocoded data, and Kiratli (2024), Steiner et al. (2023) and Unan and Klüver (2025) leverage a quasi-experimental setting. Others document cross-sectional associations (Natili and Visconti, 2023) or use panel data to track over-time change (Oana et al., 2025; Truchlewski et al., 2023), while Moise et al. (2025) use conjoint experiments to identify preference structures.

Despite this growing body of work, it remains unclear whether the external shock merely invigorated the existing pro-European supporters or also swayed the EU’s traditional skeptics. Post-functionalist theory suggests that exclusive national identity is a resilient trait that should mitigate the rallying effect among nationalists. Conversely, social identity theory proposes that potent external challenges can augment cohesion by diminishing the perceived distance between in-groups and out-groups (Tajfel and Turner, 1979). The full-scale invasion made the EU’s coordinating role even more visible, for example, through sanctions, military aid, and economic stabilization, potentially reframing the union from a distant bureaucracy encroaching on national sovereignty into a necessary “shelter” for collective protection. But did this reframing reach exclusive nationalists?

In this article, we contend that the external threat of the invasion activated a bellicist logic that overrode traditional sovereignty concerns. We argue that the “rally” in European sentiment was driven not by the “choir,” but by the converts. To test our argument, we analyze the Solidarity in Europe (SiE) survey data, fielded across 16 countries just 5 weeks after Russia’s full-scale invasion of Ukraine in 2022—a critical window capturing the peak of the initial rally effect (Truchlewski et al., 2023). We identify the heterogeneity of this effect by employing a double-sample causal forest algorithm. Unlike standard linear regression models that require researchers to pre-specify interaction terms, this machine learning approach allows us to disentangle the structure of treatment heterogeneity organically. Our contribution is distinct from prior work; while the cross-sectional design of our study cannot establish that the invasion caused attitude change, the causal forest approach we take is well-suited for mapping heterogeneity in the threat–pride association, revealing which citizen types show the strongest relationship between threat perceptions and European pride in the immediate post-invasion period. While prior work establishes that external threats can shift attitudes toward the EU on average, we demonstrate who exhibits the strongest association.

The results show that exclusive nationalists, who typically display the lowest levels of support for European integration, report substantially higher levels of European pride in response to the Russian threat. Critically, we demonstrate that this affective shift was not merely symbolic. This pattern extends to support for a common European army, where exclusive nationalists under threat also report higher levels of support. Given that defense represents a “core state power” (Genschel and Jachtenfuchs, 2022) and citizens typically resist such centralization (Wang and Moise, 2023), this finding suggests that the immediate shock of the invasion successfully transformed the EU from a distant bureaucracy into a necessary shelter for the nation-state. These associations are particularly pronounced among respondents from Central and Eastern Europe (CEE), aligning with recent findings that proximity to the aggressor drives support for integration (Malet and Hegewald, 2025; Panchuk, 2024).

External threats, national identity, and European integration: competing theoretical expectations

The existing literature provides inconclusive evidence regarding the nature of the rally-around-the-European flag effect and the heterogeneity of citizens’ responses to the external threat. In what follows, we first outline the post-functionalist constraint on integration before discussing the bellicist alternative and the shelter mechanism. We then consider how geography conditions these effects and finally discuss the implications for policy preferences.

The post-functionalist constraint

Under the post-functionalist perspective on public support for EU integration, the era of “permissive consensus” (Lindberg and Scheingold, 1970)—which allowed elites to advance integration with minimal public scrutiny—has shifted toward a “constraining dissensus” (Hooghe and Marks, 2009), where public opinion presents as a significant potential brake on further integration (Schulte-Cloos, 2018). This notion is also echoed in earlier work distinguishing between utilitarian and identity-based opposition to integration (Carey, 2002), emphasizing that identity concerns have become increasingly salient as integration has deepened into areas of core state sovereignty (Hooghe and Marks, 2018). While citizens without a strong national identity are generally supportive of further integration, citizens with an exclusive national identity tend to impede the building of an EU polity (De Vries, 2018; Van Klingeren et al., 2013). Exclusive national identity here refers to a categorical self-classification regarding group membership and who one identifies with (Turner et al., 1987)—specifically, identifying solely with one’s nation without simultaneously recognizing oneself as European (Hooghe and Marks, 2004). Research suggests that European identification is typically associated with higher education, transnational experiences, and cosmopolitan orientations (Fligstein, 2009; König, 2024; Kuhn, 2015), while exclusive national identity is more prevalent among those with lower socio-economic status and limited cross-border mobility (Fligstein et al., 2012). Moreover, different forms of national attachment carry different political consequences. While civic attachments tend to support democratic engagement, ethnic nationalism is associated with worse citizenship outcomes (Dražanová and Roberts, 2024). Importantly, strong national and European identities are not mutually exclusive, as they relate to distinct yet interconnected communities that exist in a nested relationship and tend to become salient in different social contexts (Díez Medrano and Gutiérrez, 2001; Risse, 2010).

Critically, citizens’ national versus European identities are often relatively stable (Duchesne and Frognier, 2008). Unlike economic policy preferences, which may shift with market conditions, exclusive national identity is rooted in deep-seated cultural and symbolic attachments that are generally resistant to short-term shocks (Hooghe and Marks, 2009). For citizens harboring an exclusive national identity, European integration is framed as a zero-sum game, where they perceive any transfer of authority to the EU as a direct erosion of national sovereignty (Kriesi and Schulte-Cloos, 2020; McLaren, 2002). Consequently, during previous crises such as the Eurozone or migration crises, citizens with exclusive national identities have consistently driven the opposition to centralization (Börzel and Risse, 2018).

Thus, under standard post-functionalist assumptions, the outbreak of war should not fundamentally alter the opposition of citizens with exclusive national identities to further EU integration. While exclusive nationalists might support national defense or intergovernmental cooperation, their general skepticism of the EU might be sticky and persist. This assumption is echoed in recent research suggesting that nationalist voters remain less likely to support centralized EU foreign policy than their cosmopolitan counterparts (Wang and Moise, 2023). Under this view, any aggregate rally in EU support following the invasion would primarily reflect intensified commitment among those with existing European identities, rather than conversion among skeptics.

The bellicist alternative and the shelter mechanism

Yet, the return of major warfare may also override these identity constraints. Contrasting the post-functionalist view, the “bellicist” tradition of state formation suggests that existential external threats create functional pressures for centralization that can supersede domestic political constraints (Kelemen and McNamara, 2022; Tilly, 1990). While the term originates in literature on war-making and state formation, in our context it captures a defensive logic: citizens may rally around supranational institutions not to project power externally, but to seek collective protection against an external aggressor. This macro-level process finds support in a potential psychological realignment of the citizenry under threat, as articulated by Social Identity Theory (SIT) (Tajfel and Turner, 1979).

According to SIT, external threats may augment in-group cohesion by reducing the perceived distance between the members of the threatened group (Tajfel and Turner, 1979), resulting in a recategorization of group boundaries in line with the meta-contrast principle (Turner et al., 1987). In this view, the extreme hostility of the new out-group (Russia) might minimize the perceived distance between the subordinate in-group (the nation) and the superordinate group (the EU). For exclusive nationalists, the Russian invasion might render previous cultural conflicts with the EU negligible compared to the existential threat posed by Russia (Neumann, 1999). Self-categorization theory (Turner et al., 1987) provides additional mechanisms for understanding when superordinate identities become salient. According to this theory, which social category becomes salient depends on how ready the perceiver is to use that category and how well it fits the social context (Oakes et al., 1991). A central concept is comparative fit, referring to the principle that a categorization becomes salient when perceived intergroup differences exceed intragroup differences (Oakes, 1987). The Russian invasion heightened comparative fit by making the distinction between “Europe” and “Russia” more salient than intra-European divisions. Previously contentious issues like migration policy or fiscal transfers may fade in comparison to an existential military threat. In Central and Eastern Europe, historical memories of Soviet domination may further heighten this effect (Gehring, 2022). These mechanisms help explain both why exclusive nationalists may show the strongest threat–pride association since they have the most room for recategorization given their low baseline European attachment and why effects should concentrate in CEE, where the threat is most proximate and historically resonant. Second, the external threat might also activate a “shelter” logic (Malet and Hegewald, 2025; Thorhallsson, 2011), creating a perception of “common fate,” where the survival of the nation is seen as inextricably linked to the survival of the union (Nicoli et al., 2024). Rather than viewing the EU as a threat to national sovereignty, the invasion forces even exclusive nationalists to re-evaluate the EU as a necessary instrument for preserving national sovereignty (Hobolt and Tilley, 2014). Consequently, we argue that the Russian threat should augment European pride most strongly among exclusive nationalists, for whom the EU has shifted from a perceived cultural threat to a functional necessity for national survival. The invasion made the EU’s role in coordinating collective responses highly visible, demonstrating its capacity for decisive security action through sanctions and military aid. Political elites publicly reframed the EU as essential for national security, providing cues that may have shifted citizen attitudes (Sojka et al., 2025). Moreover, the war triggered not only military insecurity but also economic vulnerability through energy disruptions and inflation, activating what Natili and Visconti (2023) term a “social security logic” alongside the geopolitical one. For exclusive nationalists in particular, these overlapping threats may have rendered previous sovereignty concerns secondary to the immediate need for collective protection. Affective realignment (SIT) and instrumental realignment (shelter logic) may operate jointly, leading to the following hypothesis.

H1 (Bellicist/Social Identity): Perceiving Russia as a threat is positively associated with European pride across all citizens, but the association is strongest among exclusive nationalists.

It is important to clarify that European pride, our main dependent variable, is conceptually distinct from European identity. While exclusive national identity refers to categorical group membership (who one identifies with), European pride captures an affective orientation toward the EU (how one feels about membership). These constructs are empirically correlated but not identical: individuals can identify exclusively with their nation while expressing varying degrees of pride in EU membership for instrumental reasons. The bellicist argument holds that an external threat may elevate European pride among exclusive nationalists without requiring a fundamental shift in their underlying identity.

Equally important is understanding how threat perceptions themselves are formed. Threat perceptions are not simply objective responses to external events but are filtered through prior predispositions in authoritarianism and general trait in anxiety (Feldman and Stenner, 1997; Stenner, 2005; Suthammanont et al., 2010). In the context of the Russian invasion, several factors shape who perceives Russia as threatening. Geographic proximity plays a key role (Lewis and Topal, 2023), with citizens in frontline states reporting higher threat levels (Malet and Hegewald, 2025; Moshagen and Hilbig, 2022). Historical experience also matters; Gehring (2022) finds that post-invasion effects were stronger among individuals with direct exposure to Soviet rule. This raises the possibility that threat perceptions are endogenous to prior political orientations, and we address this concern in our limitations section in the discussion.

Geographic heterogeneity and the dual shelter

We further expect the tension between the post-functionalist constraint and the bellicist logic to be conditioned by geography. The “shelter” mechanism should be most potent where the threat is most immediate—countries in CEE. Recent analyses confirm that regions geographically closer to the Russian border exhibit significantly higher increases in support for common defense policies (Malet and Hegewald, 2025).

For CEE countries, the Russian invasion represents a direct existential threat, reactivating historical memories of domination and what Mälksoo (2016) terms “liminal” identities—caught between Western integration and Eastern threat—making the EU’s protective function particularly salient in this region. This mechanism aligns with research showing that current events can activate dormant collective memories, shaping contemporary attitudes. Ochsner and Roesel (2024) demonstrate that political campaigns in Austria activated memories of Ottoman sieges from centuries earlier, while Fouka and Voth (2023) show that the Greek debt crisis reactivated memories of German wartime atrocities. In the EU context, Gehring (2022) finds that post-invasion solidarity effects were stronger among individuals with direct exposure to Soviet rule, suggesting that historical experience with Russian domination conditions contemporary threat responses. Critically, however, this threat is not solely military in nature. As Natili and Visconti (2023) argue, the war triggered a “social security logic” alongside the military one. The energy crisis and inflation spikes threatened the economic viability of households, creating a demand for EU-level “reinsurance” against economic volatility (Oana et al., 2025).

Hence, in CEE countries, the EU provides a dual shelter: it serves as both a geopolitical buffer against Russian aggression and an economic buffer against the costs of the war. This dual exposure to geopolitical and economic threats creates a distinctive puzzle for exclusive nationalists in the region, confronting them with a situation in which the material and security benefits of EU membership become difficult to dismiss, even through the lens of exclusive nationalism. Unlike in Western Europe, where the EU might still be viewed primarily through a technocratic lens, in CEE countries, the proximity to the conflict might highlight the functional benefits of membership (Panchuk, 2024).

H2 (Geographic Heterogeneity): The effect of threat perceptions on European pride among exclusive nationalists is particularly pronounced in Central and Eastern Europe.

Consequences for policy preferences: support for a European army

A potential limitation of analyzing affective orientations is that reported “European Pride” may only reflect expressive responding without simultaneously translating into substantive policy preferences. To address this concern and to corroborate the meaningfulness of the rally effect, we also examine preferences in the domain of defense, a “core state power” traditionally reserved for the nation-state (Genschel and Jachtenfuchs, 2022; Wang and Moise, 2023) with low public support for centralization (Krotz and Maher, 2017).

While citizens might generally support increased European cooperation, they nevertheless often remain hesitant to support full centralization, preferring a “reinsurance” model where the EU supports but does not replace national capacity (Moise et al., 2025; Oana et al., 2025). Yet, if the bellicist logic holds for exclusive nationalists, the existential nature of the threat should incentivize not only a stronger affective attachment to the EU (European pride), but also a preference for stronger integration beyond mere coordination in the form of support for a common EU army (Irondelle et al., 2015; Schoen, 2008). If even exclusive nationalists—usually the staunchest defenders of national sovereignty—shifted their support toward a common EU army in response to the external threat posed by the Russian invasion of Ukraine, this would give particularly strong support for the bellicist/social identity hypothesis.

H3 (Policy Preferences): Higher European pride among exclusive nationalists is accompanied by greater support for “hard” defense integration, specifically the creation of a common European army.

Research design and results

Data and research design

To get empirical leverage on these theoretical propositions, we draw on the 2022 data from the SiE project (Hemerijck et al., 2022). This dataset includes cross-sectional survey data from more than 10,000 respondents living in 16 democracies that are currently part of the EU.¹ The fieldwork was conducted between 1 and 25 April 2022, approximately 5 weeks after the Russian invasion of Ukraine. This timing allows us to exploit a critical window where the immediate shock of the invasion was salient, yet distinct from long-term habituation effects (Johansson et al., 2021), and allows for broad cross-national coverage.

Our main outcome variable of interest is a measure of affective orientation toward the EU (Boomgaarden et al., 2011; Lubbers, 2008). We measure the extent to which respondents are proud to be part of the EU using a scale of one to four, ranging from “not at all proud” to “very proud.”² For ease of interpretation, we standardize this variable to a mean of zero and a standard deviation of one. To measure the independent variable of interest, we rely on a binary indicator variable that records whether respondents consider Russian power and influence a threat to their country. Table 1 provides a compact summary of the coding and scaling of all key variables used in the analyses.

Table 1.

Summary of variables, coding, and scales.

Variable	Description	Scale/coding
Outcome variables
European Pride	Extent to which respondent is proud to be part of the EU.	1 $-$ 4 scale ( $z$ -standardized: mean 0, SD 1); Higher values indicate greater pride
European Army	Support for the creation of a common European army.	1 $-$ 4 scale ( $z$ -standardized: mean 0, SD 1); Higher values indicate greater support
National Pride (Placebo)	Extent to which respondent is proud to be part of their nation.	1 $-$ 4 scale ( $z$ -standardized: mean 0, SD 1); Higher values indicate greater pride
Key independent variables
Russian Threat	Respondent considers Russia the biggest security threat to their country.	Binary (1 $=$ yes, 0 $=$ no)
Exclusive National Identity	Identifies only with nation, not simultaneously or alternatively with Europe.	Binary (1 $=$ yes, 0 $=$ no)
Control variables
Age	Respondent’s age category.	Categorical (18 $-$ 24, 25 $-$ 34, 35 $-$ 44, 45 $-$ 54, 55+; Ref: 45 $-$ 54)
Female	Respondent’s gender is female.	Binary (1 $=$ yes, 0 $=$ no)
Tertiary Education	Respondent obtained tertiary education.	Binary (1 $=$ yes, 0 $=$ no)
Subjective Economic Status	Perceived economic position vis-à-vis others.	Continuous (1 $-$ 5); Higher values indicate better status
Government Approval	Approval of government’s record to date.	Binary (1 $=$ yes, 0 $=$ no)
Democracy Importance	Importance of living in a country governed democratically.	Continuous (0 $-$ 10); Higher values indicate greater importance
Religious	Identifies with a religion.	Binary (1 $=$ yes, 0 $=$ no)
Rural	Lives in a village, settlement, or isolated dwelling.	Binary (1 $=$ yes, 0 $=$ no)
Party Family Choice	Vote choice by party family (Conservative, Left, Green, Liberal, Right-Wing, Special Issue, Other, Abstention).	Categorical (Ref: Conservative)

We begin by discussing the correlates of a sense of European pride, with a particular focus on the baseline effect of Russian threat perception, which is present, on average, among all respondents. Subsequently, we apply a data-driven approach to disentangle the heterogeneity of this effect. To do so, we employ a double-sample causal forest algorithm. Unlike standard regression approaches, which require researchers to pre-specify interaction terms, this machine learning approach allows us to exploit the data to uncover complex, non-linear patterns of heterogeneity. This helps us to isolate the specific response of exclusive nationalist citizens—those who identify only with their nation state and do not express any European identification—without imposing strong functional form assumptions.

Results

We begin by assessing the baseline relationship between threat perceptions and European pride. Figure 1 displays the estimated effect of Russian threat perceptions on European pride and on support for a common European army, derived from fully-specified linear regression models with region-fixed effects.³ Formally, we estimate:

Y_{i r} = α + β_{1} {Threat}_{i r} + β_{2} {ExclID}_{i r} + γ X_{i r} + δ_{r} + ϵ_{i r}

(1)

where

Y_{i r}

represents the outcome (European pride or support for a European army) for individual

i

in region

r

;

{Threat}_{i r}

and

{ExclID}_{i r}

indicate perceived Russian threat and exclusive national identity;

δ_{r}

are region-fixed effects; and

X_{i r}

is a vector of individual controls.

Figure 1.

Average effects of Russian threat perceptions and exclusive national identity on European pride and support for a European army.

All models include indicator variables measuring respondents’ voting behavior in the most recent national election, and indicator variables for respondents’ age (categorical). To account for the impact of government approval on both threat perceptions and the outcome variables of interest, we also control for respondents’ satisfaction with the government’s track record.⁴ We present the coefficient estimates together with bootstrap distributions and confidence intervals that include the respective estimates in 95% of a bootstrap distribution. To obtain the bootstrap distribution, we perform 1000 resamples from our data, estimating our fixed-effects regression model on each resample and utilizing the percentile method to construct confidence intervals based on the empirical distribution of estimates.⁵

Consistent with the post-functional view on public support for European integration, we find a strong negative effect of respondents’ exclusive national identity. Individuals who identify solely with their country, without simultaneously (or even alternatively) recognizing themselves as part of the EU, are 0.73 standard deviations less inclined to feel a sense of pride in being part of the EU. This effect is relatively more pronounced in the CEE countries (0.78 standard deviations) than in the Western European countries (0.69 standard deviations). A similar picture emerges when examining support for a common European army, with exclusive national identity exerting a negative effect across all regions.

Critically, the results also reveal a robust positive effect of Russian threat perceptions on European pride, even after accounting for the powerful constraint of exclusive national identity. Respondents who perceive Russia as a major threat to their country are statistically significantly more likely to express pride in being part of the EU. The magnitude of this effect is also substantively large: respondents who identify Russia as a threat report European pride levels 0.23 standard deviations higher than those who do not. Consistent with the geographic heterogeneity hypothesis (H2), this effect is particularly pronounced in CEE (0.32 standard deviations). When it comes to respondents’ support for a common European army, we observe a similar pattern, though the impact of perceiving Russia as a threat to one’s country is even larger. Respondents who consider Russia a threat are 0.28 standard deviations more likely to support a common European army (all countries), with respondents in CEE countries even being 0.35 standard deviations more likely.

This initial finding is in line with recent research by Gehring (2022) and Malet and Hegewald (2025). However, the average effect masks the critical question of who is rallying around the European flag in response to the Russian threat. Specifically, who expresses the highest levels of affective attachment to the EU or the greatest support for deeper integration in the defense domain. Gehring’s (2022) causal design provides evidence that individuals across the whole distribution of prior EU identification shifted toward stronger European identity, with effects varying by threat intensity and proximity to Russia. Our contribution is complementary: while prior causal designs establish that shifts occurred, our approach maps heterogeneity in the threat–pride association, revealing which citizen types show the strongest relationship in the immediate post-invasion period. As shown in Figure 1, exclusive national identity exerts a strong negative baseline effect both on European pride and on support for a common European army. To understand whether the threat mitigates these specific constraints, we next move beyond average effects.

Before employing non-parametric machine learning techniques, we first examine the heterogeneity of this effect using a standard linear interaction specification. Table 2 reports the results of a fully specified Ordinary Least Squares model interacting Russian threat perceptions with exclusive national identity.⁶ The results confirm that the “rally” effect is not uniform but is substantively larger for exclusive nationalists. The interaction term between Russian threat perception and an exclusive national identity is positive and statistically significant, with a particularly strong effect in CEE countries when predicting European pride. This provides initial parametric evidence that the external threat of the full-scale Russian invasion of Ukraine had the most pronounced impact on those with an exclusive national identity in increasing their sense of European pride and their support for a European army.

Table 2.

Linear interaction of Russian threat perceptions and exclusive national identity.

	Outcome: European pride			Outcome: support for EU army
	All respondents	Western Europe	Central and Eastern Europe	All respondents	Western Europe	Central and Eastern Europe
Russian threat perception (0/1)	0.17 $^{* * *}$	0.12 $^{* * *}$	0.24 $^{* * *}$	0.22 $^{* * *}$	0.18 $^{* * *}$	0.31 $^{* * *}$
	[0.13; 0.20]	[0.07; 0.17]	[0.17; 0.31]	[0.17; 0.26]	[0.13; 0.23]	[0.23; 0.39]
Exclusive national identity (0/1)	${- 0.80}^{* * *}$	${- 0.75}^{* * *}$	${- 0.89}^{* * *}$	${- 0.42}^{* * *}$	${- 0.41}^{* * *}$	${- 0.42}^{* * *}$
	$[- 0.85; - 0.75]$	$[- 0.81; - 0.69]$	$[- 0.98; - 0.80]$	$[- 0.47; - 0.37]$	$[- 0.48; - 0.35]$	$[- 0.53; - 0.31]$
Excl. Nat. identity $\times$ Russian threat	${0.18}^{* * *}$	${0.15}^{* * *}$	${0.28}^{* * *}$	${0.16}^{* * *}$	${0.17}^{* * *}$	${0.14}^{*}$
	[0.11; 0.25]	[0.06; 0.23]	[0.15; 0.42]	[0.08; 0.24]	[0.08; 0.26]	$[- 0.02; 0.30]$
$R^{2}$	0.306	0.317	0.303	0.156	0.166	0.146
Num. Obs.	12,754	9149	3605	12,108	8749	3359

Note: Region-fixed effects linear regression including sample weights predicting European pride and support for a European army (mean 0, SD=1). Russian threat is coded 1 if the respondent considers Russia the biggest security threat to their country, 0 otherwise. Exclusive national identity is coded 1 if the respondent identifies only with their nation and not European, 0 otherwise (reference: non-exclusive identity). All models include survey weights and control for: age (categorical; reference: 45 $-$ 54 years), gender (reference: male), tertiary education, urban/rural residency, religiosity, subjective economic status, government approval, democracy importance, and party family choice.The 95% percentile confidence intervals in brackets derived from 1000 bootstrap resamples stratified by country. Unit of observation: individual respondent.

$^{*} p < 0.1$ , $^{* *} p < 0.05$ , $^{* * *} p < 0.01$ .

To further uncover the heterogeneity without imposing functional form assumptions, we next rely on machine learning techniques. Specifically, we use a causal forest algorithm to determine which groups of respondents may exhibit a stronger (or weaker) relationship between perceiving Russia as a threat to their country’s security and expressing a sense of pride in being part of the EU. The method of causal forests is a modified version of the random forest algorithm proposed by Breiman (2001), tailored to solve the challenge of estimating heterogeneous effects.⁷ The causal forest approach has gained prominence in political science for uncovering treatment effect heterogeneity without strong parametric assumptions (Grimmer et al., 2017; Imai and Ratkovic, 2013). Unlike traditional interaction models that require ex-ante specification of moderators, causal forests use recursive partitioning to identify subgroups with differential treatment effects, learning the structure of treatment effects from the data itself (Athey and Imbens, 2016). Thus, the algorithm allows for the flexible modeling of high-dimensional interactions by constructing numerous regression trees and averaging their predictions and serves as a robust, non-parametric validation of our linear estimates. While this data-driven approach offers significant advantages in uncovering non-linear patterns of heterogeneity without pre-specifying interaction terms, the causal forest algorithm relies on the assumption of unconfoundedness conditional on the observed covariates. This is critical as the algorithm in itself does not resolve endogeneity concerns arising from unobserved confounders, such as shared latent response styles that might simultaneously influence threat perceptions and European pride. Consequently, all causal forest models control for a wide range of potential confounders, including political orientation and government satisfaction (see Table 1).

In our implementation, we specify the model inputs as follows: the outcome variable ( $Y$ ) is the respondent’s level of European pride (or support for a European army in the subsequent analysis); the treatment variable ( $W$ ) is the perceived Russian threat; and the feature space ( $X$ ) used to grow the forest consists of the full set of controls used also for linear estimation (see Table 1). The online Supplemental Appendix discusses the results of two omnibus tests to establish the presence of a heterogeneous effect in our data (Athey et al., 2019). We rely on 5000 double-sample causal trees (Athey and Imbens, 2016; Wager and Athey, 2018), each grown on a random subset $J$ of the data by recursively splitting the feature space and maximizing the variance with respect to $\hat{τ}$ ( $X_{i}$ ), that is, the effect of Russian threat perceptions on European pride, in each leaf. The second subset $I$ of the data, in turn, is used to estimate $\hat{τ}$ ( $X_{i}$ ) among all $I$ observations falling within the same leaf of a tree grown based exclusively on the disjunct $J$ subset of the data. As for each tree, an observation is either only used for the splitting or for the estimation, double-sample causal trees satisfy the condition of “honesty” (Athey et al., 2019). The chosen subsets $J$ and $I$ are randomly resampled for each given tree, using a 50% split for structure learning and 50% for effect estimation.

Figure 2 visualizes the results of the algorithm: the small, jittered points represent the estimated conditional average treatment effect (CATE) for each respondent in the sample ( $\hat{τ} (X_{i})$ ), effectively showing the specific impact of the Russian threat to a person with that profile. The large, darker markers indicate the simple mean of these individual predictions for the subgroups of interest. Finally, to formally test whether the visible differences are statistically significant, we report the best linear projection of the CATEs onto the subgroup indicators (text annotation), providing inference regarding the magnitude and statistical significance of the heterogeneity.

Figure 2.

Heterogeneous effects of Russian threat perceptions by exclusive national identity.

The left panel of Figure 2 shows the estimated heterogeneous effect of Russian threat perceptions on European pride, stratified by respondents’ exclusive national identity. The right panel of Figure 2 presents the same analysis, but focuses on support for a common European army as the outcome variable. The effect of Russian threat perceptions on European pride is, on average, 0.128 standard deviations larger for exclusive nationalists compared to respondents who at least partially consider themselves European. Consistent with H1, it appears that individuals who exclusively identify with their nation are particularly influenced by the Russian threat, reporting higher levels of pride in being part of the EU. The strength of this relationship is particularly pronounced among respondents in CEE (0.383 standard deviations), consistent with heightened threat salience in CEE. In the face of Russian aggression, those with an exclusive national identity—who typically report the lowest levels of support for European integration (Marks and Hooghe, 2003)—show the strongest association between threat perceptions and European pride.

Having established that the Russian threat has reinforced a positive affective orientation toward the EU, especially among those who otherwise feel far from being European, we next examine whether this shift translates into substantive policy preferences. A potential limitation of analyzing identity shifts alone is the risk that reported feelings of European pride might reflect expressive responding and only pertain to a symbolic alignment with the EU, without a genuine commitment to integration. To corroborate the meaningfulness of the rally effect, we examine preferences in the domain of defense, a “core state power” traditionally reserved for the nation-state (Genschel and Jachtenfuchs, 2022). The right panel of Figure 2 displays the estimated effect of Russian threat perceptions on support for the creation of a common European army, conditional on exclusive national identity. The results provide support for H3: among exclusive nationalists who perceive Russia as a security threat, we observe significantly higher levels of support for defense centralization (0.106 standard deviations). Again, this effect is somewhat more pronounced in CEE countries (0.175 standard deviations) than in Western European countries (0.112), where the difference between those harboring an exclusive national identity and those at least partially identifying as European is only significant at the 10% level.

This finding is noteworthy for a demographic that typically prioritizes national sovereignty. It implies that the “bellicist” pressure of the war may have superseded traditional sovereignty concerns in this context. For these citizens, the EU seems to have been cognitively reframed from a challenge to national autonomy into a potential geopolitical shield. This shift from a preference for loose coordination to a demand for “hard” integration lends support to H3, suggesting that the rally among exclusive nationalists is driven also by a functional imperative for security rather than only a symbolic sentiment (Unan and Klüver, 2025).⁸

Yet, a key question remains: Is the observed positive association in European pride among exclusive nationalists truly indicative of a specific orientation toward the supranational level, or does it merely reflect a generalized “rally-around-the-flag” effect, where individuals seek solace in any collective identity in the face of external threats? To address this concern, we conduct a placebo test examining respondents’ sense of national pride. If the Russian invasion merely triggered a generalized higher need for group belonging in the face of the Russian threat, we would not only expect uniformly higher levels of both national and European pride, but also that exclusive nationalists would exhibit particularly high levels of national pride when perceiving Russia as a threat. When looking at any potential heterogeneity of the effect of Russian threat perceptions on national pride by exclusive national identity, however, we find no such pattern. The left panel of Figure 3 shows the average effect of Russian threat perceptions on national pride, while controlling for respondents’ national identity. While we do observe an overall positive effect of Russian threat perceptions on national pride, this effect is much smaller in size than the effect on European pride, and insignificant in the CEE countries. More importantly, however, we also find no significant heterogeneity of this effect by exclusive national identity. The right panel of Figure 3 shows the estimated coefficients of Russian threat perceptions on national pride, stratified by exclusive national identity. Unlike for European pride, we find no significant difference in the effect of Russian threat perceptions on national pride between exclusive nationalists and respondents who at least partially identify as European. This finding seems particularly relevant in suggesting that the “rally” is not simply a retreat into traditional national communities in the face of danger. Instead, it corroborates the premise that the EU has assumed a distinct function as a security provider that the nation-state alone cannot fulfill. The existential threat of the full-scale Russian invasion of Ukraine in 2022 appears to have heightened the salience of the superordinate European identity relative to the subordinate national identity—an effect that was particularly felt among exclusive nationalists.

Figure 3.

Placebo test: average and heterogeneous effects of Russian threat perceptions on national pride.

Conclusion and discussion

The return of major state-on-state warfare to the European continent has served as a critical stress test for the EU’s political cohesion, bringing the role of public support for the EU in times of crisis into sharp focus. In this article, we assess whether the external threat posed by Russia’s 2022 invasion of Ukraine merely strengthened the European commitment of the existing pro-European base or whether it possessed the “bellicist” force (Eilstrup-Sangiovanni, 2022) to sway the EU’s traditional skeptics (De Vries, 2018; Hooghe and Marks, 2009). Drawing on data collected just 5 weeks after the initial shock of the full-scale Russian invasion of Ukraine in February 2022, our results not only show that perceiving Russia as a threat is associated with significantly higher European pride among EU citizens, particularly in CEE countries closest to the conflict. Critically, leveraging a double-sample causal forest algorithm, our findings also highlight that this association is strongest among exclusive nationalists—individuals who tend to view the EU as a threat to national sovereignty and identity. Finally, we show that Russian threat perceptions are also associated with greater support for concrete policy change, specifically for establishing a common European army, suggesting that the affective patterns were coupled with tangible preferences for deeper integration.

Our findings align with and extend several strands of existing research. First, they corroborate recent work documenting immediate solidarity effects following the invasion (Gehring, 2022; Nicoli et al., 2024; Steiner et al., 2023) and the geographic conditioning of threat perceptions (Malet and Hegewald, 2025; Panchuk, 2024). We advance this work by demonstrating significant heterogeneity in the threat–pride association across the population: rather than simply amplifying existing pro-European sentiment among citizens identifying as European, the external threat appears to have triggered a cognitive reframing among exclusive nationalists, leading them to view the EU as a protector rather than a challenger of national sovereignty. Second, our results speak to the debate on whether external threats drive genuine integration or merely intergovernmental coordination (Freudlsperger and Schimmelfennig, 2022; Moise et al., 2025; Oana et al., 2025). The fact that exclusive nationalists report higher support for a common European army, which implies a transfer of core state powers, suggests that the bellicist logic can potentially override the constraining dissensus (Hooghe and Marks, 2009). Third, in highlighting that the effect is particularly pronounced in CEE countries, our results align with the “dual shelter” mechanism: for CEE citizens, and particularly among exclusive nationalists, the EU became a geopolitical buffer and an economic reinsurance mechanism (Natili and Visconti, 2023) simultaneously, amplifying the shelter effect beyond what either logic would produce independently.

While our findings provide important insights into the dynamics of public support for the EU in times of existential threat, they also come with certain limitations that open avenues for future research. First, drawing on cross-sectional data captured during the immediate shock of the invasion, our analysis allows us to exploit the peak of the rally effect by capturing attitudes shortly after the invasion, yet it prevents us from tracking the durability of these shifts. Recent work suggests that maintaining crisis-driven support requires a sustained demonstration of EU effectiveness (Truchlewski et al., 2023; Zeitlin et al., 2019). Future research using longitudinal data should therefore assess whether the cognitive reframing we document persists or represents a fleeting emotional response. Second, despite controlling for key confounders and relying on a causal forest algorithm well-suited for identifying causal effects even in the presence of confoundedness (Dandl et al., 2024), observational data cannot fully rule out endogeneity. While our theoretical framework posits that perceiving Russia as a threat leads individuals to embrace the EU as a protective shelter, the reverse pathway is also plausible; individuals with stronger European attachments may perceive Russia as more threatening because they see it as endangering a political community they value. However, we believe this alternative explanation is less compelling for our core finding regarding exclusive nationalists. By definition, these individuals lack strong prior affective attachments to the EU that would heighten their sensitivity to threats against the European project. If reverse causality were driving our results, we would expect the threat–pride association to be strongest among those with pre-existing European identities. Yet, we observe the opposite pattern. The fact that exclusive nationalists, who are typically indifferent or hostile to the EU, show the strongest association between threat perceptions and European pride is more consistent with a bellicist logic whereby acute security threats generate functional demands that override prior identity-based constraints. Hence, future studies should rely on experimental designs manipulating threat or identity salience, as well as panel data tracking how these attitudes evolve, to provide stronger causal leverage. We note that, Gehring (2022), using a causal difference-in-differences design around the 2014 Crimea annexation, finds that individuals across the entire distribution of prior EU identification shifted toward stronger European identity, including those with initially weaker EU attachment. This causal evidence is consistent with our finding that exclusive nationalists show a strong threat–pride association. However, our cross-sectional design cannot establish that exclusive nationalists shifted more than other groups; we can only document that in the immediate post-invasion period, the threat–pride association is strongest among this group. Future research relying on panel data should assess whether the 2022 invasion produced heterogeneous changes across identity groups, or whether our findings reflect pre-existing differences in how threat perceptions relate to European pride. In addition, future studies should test whether the same heterogeneous effects can also be found in media discourse, applying text-as-data methods to newspaper articles before and after the Russian full-scale invasion of Ukraine and assessing any potential divergence in framing between previously anti-EU and pro-EU outlets. Finally, the shelter logic we identify may be specific to the existential nature of military conflict, leaving unaddressed the potential of other external crises in overriding identity concerns.

Shedding light on the heterogeneity in support for the EU in the face of an external threat, our study reveals that the functional demand for security can supersede the cultural resistance of nationalism in specific contexts, particularly where threats are acute and proximate, suggesting that the EU’s legitimacy can be bolstered by the very citizens who once sought to constrain it. Thus, the return of war to Europe has not only redrawn the geopolitical map but may have also reconfigured the psychological boundaries of the European polity.

Supplemental Material

sj-zip-1-eup-10.1177_14651165261450461 - Supplemental material for Shelter from the storm: External threat, social identity, and support for the EU

Supplemental material, sj-zip-1-eup-10.1177_14651165261450461 for Shelter from the storm: External threat, social identity, and support for the EU by Julia Schulte-Cloos and Lenka Dražanová in European Union Politics

Supplemental Material

sj-zip-2-eup-10.1177_14651165261450461 - Supplemental material for Shelter from the storm: External threat, social identity, and support for the EU

Supplemental material, sj-zip-2-eup-10.1177_14651165261450461 for Shelter from the storm: External threat, social identity, and support for the EU by Julia Schulte-Cloos and Lenka Dražanová in European Union Politics

Supplemental Material

sj-pdf-3-eup-10.1177_14651165261450461 - Supplemental material for Shelter from the storm: External threat, social identity, and support for the EU

Supplemental material, sj-pdf-3-eup-10.1177_14651165261450461 for Shelter from the storm: External threat, social identity, and support for the EU by Julia Schulte-Cloos and Lenka Dražanová in European Union Politics

Footnotes

Acknowledgements

We would like to thank Liesbet Hooghe, Hanspeter Kriesi, Gary Marks, and Guido Tintori for their invaluable feedback on earlier drafts of this manuscript. We are also grateful for the helpful comments received from participants at the Amsterdam Centre for European Studies (ACES) European Politics and Society Seminar Series, the University of Frankfurt Comparative Politics Speaker Series, and the European University Institute’s Solidarity in Europe Data Conference. Finally, we thank three anonymous reviewers for their constructive comments during the review process.

ORCID iDs

Julia Schulte-Cloos

Lenka Dražanová

Author contributions

JS-C: conceptualization, data curation, formal analysis, methodology, project administration, validation, visualization, writing—original draft, and writing—review and editing. LD: conceptualization, methodology, project administration, and writing—review and editing.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the HORIZON EUROPE European Research Council [grant no. 885026]; Horizon 2020 Framework Programme [grant no. 754388].

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Data availability statement

The data, code, and replication materials required to reproduce all findings in this article are available on the Harvard Dataverse at: https://doi.org/10.7910/DVN/B1LG0G (Schulte-Cloos, 2026).

Supplemental material

Supplemental material for this article is available online.

Notes

References

Athey

Imbens

(2016) Recursive partitioning for heterogeneous causal effects. Proceedings of the National Academy of Sciences 113(27): 7353–7360.

Athey

Tibshirani

Wager

(2019) Generalized random forests. The Annals of Statistics 47(2): 1148–1178.

Boomgaarden

Schuck

Elenbaas

, et al. (2011) Mapping EU attitudes: Conceptual and empirical dimensions of Euroscepticism and EU support. European Union Politics 12(2): 241–266.

Börzel

Risse

(2018) From the Euro to the Schengen crises: European integration theories, politicization, and identity politics. Journal of European Public Policy 25(1): 83–108.

Breiman

(2001) Random forests. Machine Learning 45: 5–32.

Carey

(2002) Undivided loyalties: Is national identity an obstacle to European integration? European Union Politics 3(4): 387–413.

Dandl

Bender

Hothorn

(2024) Heterogeneous treatment effect estimation for observational data using model-based forests. Statistical Methods in Medical Research 33(3): 392–413.

De Vries

(2018) Euroscepticism and the Future of European Integration. Oxford: Oxford University Press.

Díez Medrano

Gutiérrez

(2001) Nested identities: National and European identity in Spain. Ethnic and Racial Studies 24(5): 753–778.

10.

Dražanová

Roberts

(2024) National attachments and good citizenship: A double-edged sword. Political Studies 72(2): 759–782.

11.

Duchesne

Frognier

A-P

(2008) National and European identifications: A dual relationship. Comparative European Politics 6(2): 143–168.

12.

Eilstrup-Sangiovanni

(2022) The anachronism of bellicist state-building. Journal of European Public Policy 29(12): 1901–1915.

13.

Feldman

Stenner

(1997) Perceived threat and authoritarianism. Political Psychology 18(4): 741–770.

14.

Fligstein

(2009) Euroclash: The EU, European Identity, and the Future of Europe. Oxford: Oxford University Press.

15.

Fligstein

Polyakova

Sandholtz

(2012) European integration, nationalism and European identity. JCMS: Journal of Common Market Studies 50(s1): 106–122.

16.

Fouka

Voth

H-J

(2023) Collective remembrance and private choice: German–Greek conflict and behavior in times of crisis. American Political Science Review 117(3): 851–870.

17.

Freudlsperger

Schimmelfennig

(2022) Transboundary crises and political development: Why war is not necessary for European state-building. Journal of European Public Policy 29(12): 1871–1884.

18.

Gehring

(2022) Can external threats foster a European Union identity? Evidence from Russia’s invasion of Ukraine. The Economic Journal 132(644): 1489–1516.

19.

Genschel

Jachtenfuchs

(2022) Bellicist integration? The war in Ukraine, the European Union and core state powers. Journal of European Public Policy 29(11): 1716–1735.

20.

Grimmer

Messing

Westwood

(2017) Estimating heterogeneous treatment effects and the effects of heterogeneous treatments with ensemble methods. Political Analysis 25(4): 413–434.

21.

Hemerijck

Genschel

Cicchi

, et al. (2022) SiE survey dataset on solidarity in Europe (2022). Data file Version 1.0.0, https://doi.org/10.7802/2506.

22.

Hobolt

Tilley

(2014) Blaming Europe?: Responsibility without Accountability in the European Union. Oxford: Oxford University Press.

23.

Hooghe

Marks

(2004) Does identity or economic rationality drive public opinion on European integration? PS: Political Science & Politics 37(3): 415–420.

24.

Hooghe

Marks

(2009) A postfunctionalist theory of European integration: From permissive consensus to constraining dissensus. British Journal of Political Science 39(1): 1–23.

25.

Hooghe

Marks

(2018) Cleavage theory meets Europe’s crises: Lipset, Rokkan, and the transnational cleavage. Journal of European Public Policy 25(1): 109–135.

26.

Imai

Ratkovic

(2013) Estimating treatment effect heterogeneity in randomized program evaluation. The Annals of Applied Statistics 7(1): 443–470.

27.

Irondelle

Mérand

Foucault

(2015) Public support for European defence: Does strategic culture matter?. European Journal of Political Research 54(2): 363–383.

28.

Johansson

Hopmann

Shehata

(2021) When the rally-around-the-flag effect disappears, or: When the COVID-19 pandemic becomes “normalized”. Journal of Elections, Public Opinion and Parties 31(1): 321–334.

29.

Kelemen

McNamara

(2022) State-building and the European Union: Markets, war, and Europe’s uneven political development. Comparative Political Studies 55(6): 963–991.

30.

Kiratli

(2024) Russian aggression and Europeans’ attitudes toward the EU—Evidence from the 2014 annexation of Crimea. European Journal of Political Research 63(4): 1699–1711.

31.

König

(2024) Forms of national and European identity: A research note reviewing literature of cross-national studies. Nationalities Papers 52(4): 707–734.

32.

Kriesi

Schulte-Cloos

(2020) Support for radical parties in Western Europe: Structural conflicts and political dynamics. Electoral Studies 65: 102–138.

33.

Krotz

Maher

(2017) Europe in an age of transition. Global Affairs 3(3): 193–210.

34.

Kuhn

(2015) Experiencing European Integration: Transnational Lives and European Identity. Oxford: Oxford University Press.

35.

Lewis

Topal

(2023) Proximate exposure to conflict and the spatiotemporal correlates of social trust. Political Psychology 44(3): 667–687.

36.

Lindberg

Scheingold

(1970) Europe’s Would-Be Polity: Patterns of Change in the European Community. Englewood Cliffs, NJ: Prentice-Hall.

37.

Lubbers

(2008) Regarding the Dutch “Nee” to the European constitution: A test of the identity, utilitarian and political approaches to voting “no”. European Union Politics 9(1): 59–86.

38.

Malet

Hegewald

(2025) The changing geography of support for European integration in the shadow of the Ukraine war. Journal of European Public Policy 1–25.

39.

Mälksoo

(2016) Liminality and contested Europeanness: Conflicting memory politics in the Baltic space. In: Berg E and Ehin P (eds), Identity and Foreign Policy: Baltic–Russian Relations and European Integration. Aldershot: Routledge, 65–83.

40.

Marks

Hooghe

(2003) National identity and support for European integration. WZB Discussion Paper.

41.

McLaren

(2002) Public support for the European Union: Cost/benefit analysis or perceived cultural threat? The Journal of Politics 64(2): 551–566.

42.

Moise

Oana

I-E

Truchlewski

, et al. (2025) Two functionalist logics of European Union polity formation under external threat: Evidence from a conjoint experiment. European Union Politics 26(2): 369–392.

43.

Moshagen

Hilbig

(2022) Citizens’ psychological reactions following the Russian invasion of the Ukraine: A cross-national study. PsyArXiv preprint: 10.31234/osf.io/teh8y.

44.

Natili

Visconti

(2023) A different logic of polity building? The Russian invasion of Ukraine and EU citizens’ demand for social security. Journal of European Public Policy 32(1): 1–25.

45.

Neumann

(1999) Uses of the Other: “The East” in European Identity Formation. Minneapolis, MN: University of Minnesota Press.

46.

Nicoli

Kuhn

Burgoon

(2024) Closer during crises? European identity during the COVID-19 pandemic and the Russian invasion of Ukraine. Journal of European Public Policy 31(8): 2234–2256.

47.

Oakes

(1987) The salience of social categories. In: Turner

Hogg

Oakes

, et al. (eds), Rediscovering the Social Group: A Self-Categorization Theory. Oxford: Basil Blackwell, 117–142.

48.

Oakes

Turner

Haslam

(1991) Perceiving people as group members: The role of fit in the salience of social categorizations. British Journal of Social Psychology 30(2): 125–144.

49.

Oana

I-E

Moise

Truchlewski

(2025) Demand for EU polity building in the shadow of the Russian threat. Comparative European Politics 23: 1–23.

50.

Ochsner

Roesel

(2024) Activated history: The case of the Turkish sieges of Vienna. American Economic Journal: Applied Economics 16(3): 76–112.

51.

Panchuk

(2024) The impact of the Russian invasion of Ukraine on public support for EU enlargement. Journal of European Public Policy 31(10): 3128–3150.

52.

Risse

(2010) A Community of Europeans? Transnational Identities and Public Spheres. Ithaca, NY: Cornell University Press.

53.

Schoen

(2008) Identity, instrumental self-interest and institutional evaluations: Explaining public opinion on common European policies in foreign affairs and defence. European Union Politics 9(1): 5–29.

54.

Schulte-Cloos

(2018) Do European Parliament elections foster challenger parties’ success on the national level? European Union Politics 19(3): 408–426.

55.

Schulte-Cloos

(2022) Political potentials, deep-seated nativism and the success of the German AfD. Frontiers in Political Science 3.

56.

Schulte-Cloos

(2026) Replication Data for: Shelter from the Storm: External Threat, Social Identity, and Support for the EU. Harvard Dataverse. DOI: 10.7910/DVN/B1LG0G.

57.

Schulte-Cloos

Leininger

(2022) Electoral participation, political disaffection, and the rise of the populist radical right. Party Politics 28(3): 431–443.

58.

Sojka

Terraza

Caravaca Crespo

, et al. (2025) Russian invasion as a European issue: Vertical Europeanisation of national political debates and the war in Ukraine. European Union Politics 26(2): 310–343.

59.

Steiner

Berlinschi

Farvaque

, et al. (2023) Rallying around the EU flag: Russia’s invasion of Ukraine and attitudes toward European integration. Journal of Common Market Studies 61(2): 283–301.

60.

Stenner

(2005) The Authoritarian Dynamic. Cambridge: Cambridge University Press.

61.

Suthammanont

Peterson

DAM

Owens

, et al. (2010) Taking threat seriously: Prejudice, principle, and attitudes toward racial policies. Political Behavior 32: 231–253.

62.

Tajfel

Turner

(1979) An integrative theory of intergroup conflict. In: Austin WG and Worchel S (eds), The Social Psychology of Intergroup Relations. Brooks/Cole, 33–37.

63.

Thorhallsson

(2011) Domestic buffer versus external shelter: Viability of small states in the new globalised economy. European Political Science 10(3): 324–336.

64.

Tilly

(1975) The Formation of National States in Western Europe. Princeton, NJ: Princeton University Press.

65.

Tilly

(1990) Coercion, Capital, and European States, AD 990–1990. Cambridge, MA: Blackwell.

66.

Truchlewski

Oana

I-E

Moise

(2023) A missing link? Maintaining support for the European polity after the Russian invasion of Ukraine. Journal of European Public Policy 30(8): 1662–1678.

67.

Turner

Hogg

Oakes

, et al. (1987) Rediscovering the Social Group: A Self-Categorization Theory. Oxford: Basil Blackwell.

68.

Unan

Klüver

(2025) Europeans’ attitudes toward the EU following Russia’s invasion of Ukraine. Political Science Research and Methods 13(4): 1025–1030.

69.

Van Klingeren

Boomgaarden

De Vreese

(2013) Going soft or staying soft: Have identity factors become more important than economic rationale when explaining Euroscepticism? Journal of European Integration 35(6): 689–704.

70.

Wager

Athey

(2018) Estimation and inference of heterogeneous treatment effects using random forests. Journal of the American Statistical Association 113(523): 1228–1242.

71.

Wang

Moise

(2023) A unified autonomous Europe? Public opinion of the EU’s foreign and security policy. Journal of European Public Policy 30(8): 1679–1698.

72.

Zeitlin

Nicoli

Laffan

(2019) Introduction: The European Union beyond the polycrisis? Integration and politicization in an age of shifting cleavages. Journal of European Public Policy 26(7): 963–976.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.11 MB

0.18 MB

22.47 MB

0.00 MB