Keywords

European Social Survey, Right-wing populism (RWP), BMI, Depression (CES-D8), Regional health, Geographies of discontent,Ecological Analysis.

1. Introduction and Theoretical Framework

    While voting is the primary practice of citizenship (Mattila et al., 2013; Pacheco & Fletcher, 2015),cross-national variation in political participation is more convincingly explained by structural and individual-level factors such as residential stability and health. Goerres (2007) argues that higher voter turnout among older citizens is driven less by economic self-interest than by social norms, habituation, and conformity. Beyond age, however, health has emerged as a crucial yet often underexamined determinant of political engagement. Although research has only recently begun to explore how health shapes trust in political institutions (Papageorgiou, Mattila, & Rapeli, 2019), substantial evidence demonstrates that poor health constitutes a significant barrier to participation. Individuals in poor physical condition frequently shift their focus from public political life to personal survival, reducing their engagement in civic activities (Söderlund & Rapeli, 2015).
   Empirical studies consistently show that physical disabilities, limitations in daily activities, and poor self-reported health decrease both psychological involvement in politics and voter turnout (Gidengil & Wass, 2024; Miller & Powell, 2016; Schur & Adya, 2013). Although isolated exceptions exist, the prevailing consensus is that poor health diminishes voter registration and electoral participation (Brown et al., 2020; Couture & Breux, 2017; Denny & Doyle, 2007; Mattila et al., 2013; Pacheco & Fletcher, 2015). Mental health conditions further reinforce this pattern. Depression, for example, undermines both internal political efficacy (belief in one’s own competence) and external political efficacy (belief in government responsiveness), often through mechanisms such as low self-worth and cognitive distortions (Joormann, 2009; LeMoult & Gotlib, 2019).
   Over time, reduced external efficacy may intensify into a “negative bias” toward political institutions, further suppressing electoral participation (Bernardi et al., 2022). While depression is typically treated as a cause of disengagement, evidence suggests a reciprocal dynamic in which diminished political efficacy may also exacerbate depressive symptoms (Bernardi et al., 2022). Ultimately, health disparities contribute to a representation gap: for individuals in poor physical or mental health, the perceived costs of voting frequently outweigh the anticipated benefits (Brown et al., 2020; Denny & Doyle, 2007).
   At the same time, the European political landscape has undergone substantial transformation over the past decade, marked by the rapid rise of right-wing populist (RWP) parties. Because poor health is associated with lower political trust (Papageorgiou, Mattila, & Rapeli, 2019), it is important to consider whether health disparities may also shape support for anti-establishment political actors. Political engagement generally requires resources such as time, financial stability, and civic skills; consequently, individuals with higher socioeconomic status and education levels tend to participate more actively in conventional politics (Kirbiš et al., 2024). When health constraints intersect with socioeconomic disadvantage, barriers to mainstream engagement may intensify.
   The rise of right-wing populism is often interpreted as a symptom of institutional failure. If poor health limits participation in conventional political channels, it may also erode trust in moderate institutions, thereby increasing receptiveness to anti-establishment alternatives. Although support for the extreme right has traditionally been associated with younger, moderately educated men and contextual factors such as unemployment (Arzheimer, 2016), demographic and regional health disparities may add an additional explanatory layer. Healthier populations have been found to favor conservative economic policies aligned with protecting material interests (Smith & Dorling, 1996). Expanding this perspective, Laverty and Hopkinson (2025) propose a “biological” dimension to political instability, arguing that deteriorating population health (particularly rising obesity and chronic disease) can serve as a proxy for regional neglect. In communities where healthcare and social systems are perceived as failing, trust in mainstream institutions may erode, enhancing the appeal of anti-establishment “outsider” parties as a reaction to upstream social determinants of health.
    While initial research identified a strong association between long-term health conditions and support for Reform UK in England, this study broadens the scope to 23 European countries. Its objective is to examine whether contemporary “geographies of discontent” are underpinned not only by economic and cultural grievances but also by biological and psychological markers of systemic neglect.



1.1. Research Hypotheses
Based on the theoretical framework established by Laverty and Hopkinson (2025), this study tests the following three hypotheses across the European regional context:
—H1 (The Physical Health Hypothesis): Regional obesity prevalence (Mean BMI) is positively associated with right-wing populist (RWP) vote shares.
—H2 (The Mental Health Hypothesis): Higher levels of regional mental distress (Mean CES-D8 scores) predict higher levels of RWP support.
—H3 (Robustness Hypothesis): The association between regional health indicators and right-wing populist support remains statistically significant after controlling for key regional socio-economic and demographic characteristics.
2. Methods
2.1. Research Design and Data Sources

   The empirical analysis is structured as a cross-national ecological study, replicating and enhancing the framework created by Laverty and Hopkinson (2025). This replication redirects its focus to 23 European countries. The main data source is the European Social Survey (ESS), particularly combining Round 10 (2020) and Round 11 (2022). This ecological perspective is essential to examine the “geographies of discontent” hypothesis, facilitating the comparison of health results and political actions across different local settings.

2.2. Variable Operationalization and recoding

   To ensure comparability with the original study, variables were extracted and processed as follows:
Right-Wing Populism (Dependent Variable): Regional support for RWP parties (rwpop) was calculated by identifying respondents who voted for parties classified as right-wing populist in the PopuList or CHES datasets (e.g., AfD in Germany, VOX in Spain, Chega in Portugal, and the PVV in the Netherlands). This individual-level voting data was aggregated into a regional percentage share.Physical and Subjective Health: these were generated: regional means (mean_bmi) and the proportion of the population meeting the clinical definition of obesity (BMI > 30). Subjective Health: This was recoded to identify the proportion of residents reporting “good health” (pct_good_health) .
   Mental Health (CES-D8): Psychological distress was measured using the 8-item Center for Epidemiologic Studies Depression Scale. Healthcare Satisfaction (mean_health_sat): A regional mean (0–10 scale) representing the perceived quality of state health provision.Institutional Trust (mean_trust): An index of trust in parliament, politicians, and the legal system, serving as a control for broader political alienation. In addition to the main health indicators, to this replication, a range of control variables to consider other possible reasons for populist backing was added, such as Demographic factors encompassing the gender distribution by region (pct_male) and age demographics. To reflect the “social” aspect of the Laverty thesis, indicators of social frequency regional Social Integration Index (mean_social) was added. Additionally, the model addresses Economic Precarity using both objective and subjective indicators, notably the proportion of individuals facing economic pressure (pct_econ_strain) and the percentage of people unemployed for more than 3 months (pct_longterm_unemp). The mean Happiness and the % of native people born in their country was also analyzed. To account for life-cycle effects and regional demographic shifts, the models include regional-level controls for Mean Age (mean_age) and Education levels (mean_edu ). Incorporating these controls guarantees that any detected association between BMI or depression and right-wing voting is not simply an outcome of wider regional discontent or economic struggle.

2.3. Analytical Strategy

   The analysis proceeds in two stages. First, bivariate visualizations (scatterplots and dual-axis bar charts) and Pearson correlations. Second, the study employs regression models (Table 2). The models utilize a “step-wise” technique to test if health variables retain their predictive power when economic variables and social variables (Trust) are introduced. All models are estimated using OLS regression with region-level observations (N = 99).This allows for a clear comparison of effect sizes across different European contexts and directly mirrors the multivariate approach used in the Laverty and Hopkinson study.

2.4. Software and Package Documentation

   Every aspect of data handling, visualization, and statistical modeling was performed with R (version 4.x). The analysis mainly utilized Base R functions for data manipulation and aggregation (e.g., subset, aggregate). Bivariate visualizations were created using R’s built-in graphic tools (plot, barplot), while the modelsummary package was employed to generate regression tables suitable for publication. The original ESS datasets were imported using the foreign package.

3.Results


Table 1: Sample Description
Dimension Values Count Statistics
Gender 1 8256 48.1%
2 8896 51.9%
Age group 18-29 2067 12.1%
30-59 8215 47.9%
60+ 6870 40.1%
Education (years) 17152 13.1 ± 4
BMI 17152 25.8 ± 4

**Gender 1= Male


The following Scatter plots presents regional-level bivariate relationships (N = 99) between right-wing populist (RWP) vote share and the main explanatory variables. The final analytical sample consists of 99 regions across 23 European countries. Because these associations may be influenced by confounding demographic, socioeconomic, and institutional factors, multivariate controls are required to isolate the true effects.

Table 2: Step-wise Regression Models on a aggregate level with Right-Wing Populism as dependent variabel
Model 1: Baseline Model 2: Psychological/Well-being Model 3: Socioeconomic Model 4: Full/Institutional
+ p < 0.1, * p < 0.05, ** p < 0.01, *** p < 0.001
Mean BMI 0.066** 0.065** 0.079*** 0.055*
Mean Age -0.008 -0.008 0.002 -0.004
% Male 1.241*** 1.046*** 0.894** 0.673*
Proportion Depressive (CES-D8) -0.498* -0.486* -0.553*
% Subjective good Health 0.156 0.146 -0.152
Mean Happiness 0.085 -0.023 -0.004
Satisfaction with Health -0.022 -0.051* -0.038
Satisfaction with Democracy 0.112*** 0.065*
% Long-term Unemployed -0.380+ -0.582**
Mean Education (Years) -0.026
Institutional Trust 0.085*
% Native Born -0.974*
Social Integration Index -0.034
Constant -1.765*** -2.217* -1.938* 0.797
Num.Obs. 99 99 99 99
R2 0.381 0.446 0.533 0.626
R2 Adj. 0.362 0.403 0.486 0.568



3.2. Data analysis
   The dependent variable, RWP Vote Share, exhibits substantial regional variation, ranging from 0% to 100%. Regarding the primary health determinant, the regional Mean BMI is 25.8, placing the average European region in the ‘overweight’ category (BMI>25) according to WHO standards. Strong positive correlation appear for % Male (r=0.56), institutional trust (r=0.44), and mean happiness (r=0.36). The BMI–RWP relationship remains statistically significant after introducing a broad range of controls as shown in Table 2.
   In Model 1 (Baseline), the demographic specification supports the hypothesis regarding physical health. Average BMI exhibits a positive and statistically significant association with RWP vote share (β = 0.066, p < 0.01). Gender remained one of the strongest predictors overall, with male-dominated regions showing significantly higher RWP support. Model 2 incorporates indicators of psychological and subjective well-being. The coefficient for Mean BMI stays positive and significant (β = 0.065, p < 0.01), indicating stability once mental health controls are considered. Contrary to theoretical expectations, the percentage of respondents exhibiting depression (CES-D8) is inversely related to RWP vote share (β = −0.498, p < 0.05). Measures of subjective health, well-being, and satisfaction with health do not attain statistical significance. Even though the coefficient for the male share decreases (β = 1.046, p < 0.001), it still holds significant strength.
   Model 3 includes socioeconomic controls. The average BMI shows a small rise in magnitude (β = 0.079, p < 0.001), indicating that the link is not entirely accounted for by economic factors. Symptoms of depression continue to show a negative association with RWP vote share (β = −0.486, p < 0.05). Contentment with democracy is positively associated with RWP vote share (β = 0.112, p < 0.001). Extended unemployment exhibits slight significance (β = −0.380, p < 0.10), while health satisfaction emerges as negatively significant (β = −0.051, p < 0.05). The male proportion remains positive but diminished (β = 0.894, p < 0.01).
   The fully specified Model 4 additionally includes institutional and compositional factors. Mean BMI continues to be positive and statistically significant (β = 0.055, p < 0.05), verifying the consistency of the physical health relationship. Depressive symptoms still show a negative association (β = −0.553, p < 0.05).
   Prolonged unemployment becomes more strongly negative and statistically significant (β = −0.525, p < 0.01), whereas institutional confidence (β = 0.085, p < 0.05) and contentment with democracy (β = 0.065, p < 0.05) show a positive association with RWP vote share. Education and social integration lack statistical significance. The attenuation of some effects in Model 4 may reflect multicollinearity among predictors (e.g., education and institutional trust).
   In all models, the explanatory power grows significantly (R² increasing from 0.381 in Model 1 to 0.626 in Model 4), suggesting that the broader specifications account for more regional variation in populist support.

4.Discussion
   In general, Mean BMI serves as a stable and statistically significant predictor of RWP voting through all model specifications, offering consistent backing for the physical health hypothesis. Conversely, the prevalence of regional depression shows a negative association with populist electoral support, challenging the mental health theory. Although the percentage of male participants continues to be a significant factor, its influence diminishes when structural and institutional controls are implemented. Combined, the results indicate that regional physical health and demographic makeup are more reliably linked to RWP support than measures of psychological distress.While Laverty and Hopkinson (2025) identify a positive association between long-term illness and Reform UK support in England, the present cross-national regional analysis confirms the broader physical health mechanism but does not replicate the mental distress channel. In contrast to their mobilization interpretation, depressive prevalence is negatively associated with RWP support in the European context. This suggests that the health–populism relationship may be context-dependent.
   The findings provide support for H1. Regional mean BMI emerged as a robust predictor of RWP support across all specifications, largely independent of economic or institutional factors. Even after controlling for unemployment and education, the link remained, suggesting that the association between BMI and RWP support is not fully explained by economic controls.This aligns with the somatic marker hypothesis, where physical heaviness may correlate with a preference for political stability or tradition, and also confirms the theory presented by Gidengil & Wass (2024).
   Results for H2 run counter to theoretical expectations and therefore do not support the hypothesis. because data suggests that psychological distress in Europe may lead to political withdrawal rather than radicalization. Unlike anger, which may mobilize voters toward populist causes, depressive symptoms are more commonly associated with reduced political engagement. Rather than fueling populist anger, regional depression was negatively associated with RWP voting. This confirms the “withdrawal” hypothesis suggested by Bernardi et al. (2022): populations struggling with high levels of depressive symptoms may lack the ‘civic energy’ required for political mobilization, even for populist causes.
   Results for H3 receive partial support. While structural variables such as the share of males and native-born residents remain positively associated with RWP vote share in the fully specified model, the unemployment coefficient is negative, contradicting conventional economic grievance expectations. This suggests that structural and demographic composition matter for regional populist support, but not always in the direction predicted by traditional economic insecurity frameworks. Instead, the findings indicate that RWP voting in this sample may be more closely linked to demographic composition and identity-related factors than to economic deprivation alone.
   Unlike the mobilization effect seen in Laverty and Hopkinson’s (2025) study of physical illness, mental distress in this European context appears to lead to apathy and abstention rather than radical voting.Together with the negative unemployment coefficient, this suggests that RWP support in this sample may be driven more by cultural or identity-based concerns in relatively stable regions rather than by economic deprivation alone.

5.Conclusion
   This analysis of 99 European regions provides differentiated support for health-based explanations of RWP voting. Regional mean BMI shows a consistent and statistically significant positive association with RWP vote share across all model specifications, even after controlling for socioeconomic, demographic, and institutional factors. In contrast, depressive symptoms are negatively and significantly associated with RWP support in all models in which they are included, offering no support for the mental health mobilization hypothesis. Beyond health, the share of males, structural unemployment, and native-born residents emerge as significant predictors in the fully specified model.    Overall, physical health disparities appear to be stronger correlates of RWP support than psychological distress or subjective economic hardship. While BMI functions as a stable structural predictor, depressive prevalence is associated with lower, rather than higher, populist support—consistent with a “mobilization versus withdrawal” distinction. These findings suggest that European “geographies of discontent” are heterogeneous: some regions may mobilize around physical grievances, whereas others characterized by psychological distress may experience political disengagement instead of radicalization. In contrast to Laverty’s findings for England, which link chronic illness to Reform UK support, this cross-national replication confirms the broader physical health mechanism but does not reproduce the mental distress pathway.


6.Limitations
   These findings are primarily constrained by the ecological fallacy and the use of cross-sectional, post-pandemic data. Unlike the original UK study’s clinical NHS records, this replication relies on self-reported ESS metrics. Using post-pandemic ESS data offers a vital snapshot, but it remains cross-sectional. We cannot definitively state that declining health leads to populism; it is equally plausible that the socioeconomic conditions fostering populism (e.g., regional decay, lack of infrastructure) simultaneously contribute to poor health outcomes.While ESS weights aim to mitigate bias, this method risks introducing cultural reporting variations. Future research should distinguish between specific health domains and also utilize qualitative interviews to better understand voting norms.



References


Arzheimer, K. (2016). Electoral sociology – who votes for the Extreme Right and why – and when? In R. Eatwell & M. Goodwin (Eds.), The post-war radical right in Britain (pp. 295–307). Routledge.    https://doi.org/10.4324/9781315514574-25

Bernardi, L., Mattila, M., Papageorgiou, A., & Rapeli, L. (2022). Down but not yet out: Depression, political efficacy, and voting. Political Psychology, 44(2), 233–251. https://doi.org/10.1111/pops.12837 .

Brown, C. L., Raza, D., & Pinto, A. D. (2020). Voting, health and interventions in healthcare settings: A scoping review. Public Health Reviews, 41(1), Article 16. https://doi.org/10.1186/s40985-020-00133-6 .

Couture, J., & Breux, S. (2017). The differentiated effects of health on political participation. European Journal of Public Health, 27(4), 599–604. https://doi.org/10.1093/eurpub/ckw245 .

Denny, K., & Doyle, O. (2007). “… Take up thy bed, and vote.” Measuring the relationship between voting behaviour and indicators of health. European Journal of Public Health, 17(4), 400–401. https://doi.org/10.1093/eurpub/ckm002.

Gidengil, E., & Wass, H. (2024). Health and psychological engagement with politics. In The Oxford handbook of health and politics. Oxford University Press.

Goerres, A. (2007). Why are older people more likely to vote? The impact of ageing on electoral turnout in Europe. The British Journal of Politics and International Relations, 9(1), 90–121. https://doi.org/10.1111/j.1467-856X.2007.00245.x.

Joormann, J. (2009). Cognitive predictors of depressive symptoms and depression: The role of negative cognitive biases. In I. H. Gotlib & C. L. Hammen (Eds.), Handbook of depression (pp. 285–304). The Guilford Press .

Kirbiš, A., Mattila, M., & Rapeli, L. (2024). Physical health conditions and political participation in Europe: The moderating effects of age. Comparative European Politics, 22(2), 212–242. https://doi.org/10.1057/s41295-023-00347-3.
Laverty, A. A., & Hopkinson, N. S. (2025). The politics of breathlessness: Health and voting patterns in the 2024 UK general election. The Lancet Regional Health – Europe. https://doi.org/10.1016/j.lanepe.2024.101037.

LeMoult, J., & Gotlib, I. H. (2019). Depression: A cognitive perspective. Clinical Psychology Review, 69, 51–66. https://doi.org/10.1016/j.cpr.2018.06.008.

Mattila, M., Denny, K., Wass, H., & Nettelfield, T. (2013). Healthy voting: The effect of self-reported health on turnout in 30 countries. Electoral Studies, 32(4), 886–891. https://doi.org/10.1016/j.electstud.2013.07.010 .

Miller, P. S., & Powell, L. W. (2016). Overcoming voting obstacles: The participation of newly registered voters with disabilities. American Politics Research, 44(1), 28–59. https://doi.org/10.1177/1532673X15586618 .

Pacheco, J., & Fletcher, D. (2015). Incorporating health into studies of political behavior: Evidence for turnout and partisanship. Political Research Quarterly, 68(1), 104–116. https://doi.org/10.1177/1065912914563548 .

Papageorgiou, A., Mattila, M., & Rapeli, L. (2019). The health gap in political trust: Evidence from 27 European countries. European Journal of Political Research, 58(3), 856–878. https://doi.org/10.1111/1475-6765.12311 .

Schur, L., & Adya, M. (2013). Sidelined or mainstreamed? Political participation and attitudes of people with disabilities in the United States. Social Science Quarterly, 94(3), 811–839. https://doi.org/10.1111/ssqu.12028 .

Smith, G. D., & Dorling, D. (1996). “I’m all right, John”: Voting patterns and mortality in England and Wales, 1981–92. BMJ, 313(7072), 1573–1577. https://doi.org/10.1136/bmj.313.7072.1573 .

Söderlund, P., & Rapeli, L. (2015). Poor health and voter turnout: The moderating effect of national wealth. European Journal of Public Health, 25(4), 624–628. https://doi.org/10.1093/eurpub/ckv012 .