Economic inequalities and discontent in European cities

The paper addresses the urban well-being paradox: despite cities offering jobs, amenities, and social opportunities, residents in denser, larger urban areas report lower subjective well-being. Competing explanations include: (1) compensating differentials where negative externalities outweigh benefits; (2) migrants’ optimism or willingness to accept lower well-being in exchange for urban advantages; (3) compositional effects wherein individual characteristics and their interaction with place features drive lower well-being in cities; and (4) the often-overlooked role of interpersonal income inequality, which tends to scale with city size and can fuel discontent. The study jointly examines these mechanisms, testing whether inequalities and discontent scale with city size and whether individual disadvantage amplifies the negative effects. Using city-level measures (rather than regional), it explores overall life discontent and multiple domains of city life across 83 European cities, aiming to disentangle individual versus contextual contributions to discontent and the role of inequality in large cities.

Prior work documents lower subjective well-being in more urbanized settings (the urban malaise/urban well-being paradox), despite theoretical expectations of urban benefits (agglomeration economies, amenities). Explanations include compensating wage differentials offset by congestion, pollution, crime, and high living costs; migrant optimism and unmet aspirations leading to disappointment; and compositional effects—cities’ social heterogeneity and segregation increase the likelihood that many residents, particularly non-elite groups, do not fully benefit from urban advantages. A growing literature highlights that interpersonal inequalities and their scaling with city size are high in large cities, driven by concentration of high-skilled, high-income elites alongside a prevalence of low-skill, precarious jobs. Inequalities—both interpersonal and spatial—are tied to political and individual discontent. However, few studies jointly assess these explanations or examine heterogeneity across groups and life domains; this paper fills that gap by focusing on inequality scaling, compositional-contextual interactions, and domain-specific dissatisfaction.

Data: The study uses the 2019 European Commission “Perception survey on the quality of life in European cities,” covering 52,500 respondents in 83 cities across 35 countries (79 cities and 4 Greater cities). The survey focuses exclusively on urban residents and includes detailed perceptions of life satisfaction and multiple city life domains (transport, education, environment, trust/safety). Limitations of the data include absence of rural/peripheral areas and its cross-sectional nature.

Variables: Dependent variables are reverse-coded measures of discontent (1=very satisfied/very low discontent to 4=not at all satisfied/very high discontent): (i) discontent with life in general, (ii) discontent with life in the city, and (iii) discontent with specific urban domains (job opportunities, safety, trust, public transport, healthcare, cultural facilities, schools/education, air quality, green/environment, etc.).

Key city-level independent variables: (a) intra-city (interpersonal) inequality measured by the Gini index of after-tax disposable income (available at NUTS2 level), (b) inter-city disparities measured by average annual GDP per capita growth over 2011–2018 (NUTS3), (c) level of GDP per capita PPP in 2018 (NUTS3), and (d) city size rank (1st: >1M; 2nd: 500k–1M; 3rd: 250k–500k; 4th: <250k residents). Continuous city-level variables are mean-standardized.

Individual controls: age (and squared), gender, household composition, occupation, education, difficulty paying bills (low-income proxy), and a dummy for previously living in another city (to mitigate sorting concerns).

Econometric approach: Multilevel linear models (random intercepts) with three levels—individual (level 1), city (level 2), and country (level 3)—estimate associations between individual/contextual factors and discontent. This structure accounts for clustering by city and country. Robustness checks use multilevel ordered logit models, yielding consistent results. The main hypothesis tests whether the association between intra-city inequality (Gini) and discontent is stronger for economically disadvantaged individuals; this is implemented via an interaction between Gini and the “difficulty paying bills” dummy.

• Scaling of inequalities: Larger (top-rank) cities exhibit significantly higher interpersonal inequality. Mean Gini by rank: 1st=0.330 (0.034), 2nd=0.304 (0.036), 3rd=0.290 (0.034), 4th=0.284 (0.046); ANOVA F-test=5.34 (p<0.01). Per capita GDP is higher in larger cities (1st rank mean 44.894 vs 4th rank 29.697, PPPs), but average GDP growth (2011–2018) does not significantly differ across ranks (F=0.31), implying that large-city wealth and inequality are not driven by superior growth.
• Scaling of discontent and role of inequality: Living in top-rank cities is associated with higher life discontent (e.g., 1st-rank city coefficient 0.057**, model [a]). Higher GDP per capita slightly mitigates discontent (−0.031*), while GDP growth does not. Crucially, once the Gini is included, the city-rank and GDP-per-capita effects lose significance, and the Gini becomes strongly positive and significant (≈0.041**, models [b],[c]). This indicates intra-city inequality is a primary driver of discontent, beyond city size and growth.
• Individual determinants: Economic hardship and lower socioeconomic status consistently raise discontent: difficulty paying bills (~0.255–0.260***), low education (~0.079***), unemployed (~0.124–0.133***), and manual workers (~0.068–0.077***). Managers/professionals report lower discontent (≈−0.072 to −0.096***). Age has a concave relationship with discontent (positive linear, negative squared). The effect of previously living in another city is small and becomes non-significant once inequality is controlled.
• Domain-specific dissatisfaction (Table 3): Large cities increase discontent in many domains (life in the city, safety, trust, schools, air quality), but not in job opportunities, public transport, or cultural facilities—consistent with large cities as labor-market matching hubs and consumption/amenity centers. Higher Gini correlates with more discontent in trust (0.060***), healthcare (0.134***), cultural facilities (0.109***), schools (0.095***), and life in the city (0.050*). Low education and difficulty paying bills increase discontent across nearly all domains.
• Inequality–disadvantage interaction (Table 4): The interaction between Gini and difficulty paying bills is positive and significant across outcomes (e.g., life in general: 0.024**; life in the city: 0.027**; job opportunities: 0.045***), showing that inequality’s negative effects are amplified for financially constrained individuals.
• Variance decomposition (Table 2): Most variance is at the individual level (level-1 variance ≈0.516). Intraclass correlations: city-level ICC≈0.009; country-level ICC≈0.044–0.056, indicating modest clustering by place, with individual characteristics and intra-city inequality explaining much of the variation.

Findings indicate that large cities are intrinsically more unequal and that these inequalities, rather than city size per se or recent growth, are the key drivers of individual discontent. Urban ranking appears to proxy inequality: when inequality is directly controlled, the city-size effect vanishes. The heterogeneity of domain-specific results shows that urban agglomeration advantages (jobs, transport, culture) coexist with heightened dissatisfaction in other areas (safety, trust, schools, environment), especially where inequality is high. Disadvantaged individuals experience compounded negative effects, consistent with compositional and contextual factors interacting to produce the urban well-being paradox. These results underscore that the benefits of large cities accrue disproportionately to elites, while many residents—particularly those with low income or education—report higher dissatisfaction, aligning with broader narratives linking socioeconomic disparities to political discontent.

The paper contributes by jointly testing scaling effects of inequality and discontent and by demonstrating that intra-city interpersonal inequality is the dominant predictor of individual discontent in European cities. It shows that large cities’ advantages are not broadly shared and that domain-specific dissatisfaction cumulates with inequality, disproportionately affecting disadvantaged groups. Policy implications include prioritizing equitable access to public services (transport, healthcare, education, cultural amenities, green/environmental quality) and implementing redistributive and labor policies to curb rising interpersonal and occupational inequalities, thereby mitigating risks that socioeconomic discontent translates into political instability. Future research should investigate the spatial structure of intra-urban inequalities (e.g., center–periphery divides) and their relationship with discontent, leveraging finer-grained spatial data.

• Data cover only urban areas, preventing comparison with rural/peripheral settings.
• Cross-sectional survey limits causal inference and tracking dynamics over time.
• Key contextual measures (GDP, Gini) are unavailable at true city level; proxies at NUTS3 (GDP) and NUTS2 (Gini) are used.
• Potential sorting bias is addressed via a migration-history control, but residual endogeneity cannot be entirely ruled out.
• Parsimonious set of contextual indicators due to data availability constraints.