Social capital, human capital and ethnic occupational niches: an analysis of ethnic and gender inequalities in the Spanish labour market

The study addresses why immigrants—particularly Moroccans and Ecuadorians in Spain—are overrepresented in precarious, low-wage segments of the labour market and how gender intersects with ethnicity to shape these outcomes. It evaluates three explanatory perspectives: human capital theory (differences in education, training and transferability of skills), labour market stratification/segmentation (institutionalized divisions producing a primary vs. secondary market, with immigrants concentrated in ethnic occupational niches), and social networks/social capital (how relational resources affect job access and status). The research question asks which types of social capital (ties to native-born Spaniards, co-ethnics in Spain, or transnational co-ethnics) are most useful for immigrants’ occupational attainment in Catalonia, once human capital and sectoral segmentation are considered, and whether returns differ by gender. This is important given Spain’s ethno-stratified and gender-segregated labour market, policy barriers to credential recognition, and the Great Recession’s impact on immigrant employment.

Three strands of literature frame the study. (1) Human capital theory posits that education and training investments determine labour market rewards; for migrants, transferability (language, credentials, job-related skills) evolves over time, potentially producing U-shaped occupational trajectories (Becker; Chiswick). Critiques argue this view individualizes outcomes and can blame those who fare poorly. (2) Labour market stratification/segmentation emphasizes demand-side structures and institutionalized practices that sort workers into primary vs. secondary segments. In Spain, policy and economic dynamics channel immigrants—especially those from the Global South—into labour-intensive ethnic occupational niches (construction, agriculture, hotels and catering, retail, domestic work), often precarious and stigmatized. This concentration is sharper for immigrant women, who face horizontal and vertical gendered discrimination and are steered into undervalued care-related jobs. (3) Social networks and social capital research shows contacts provide information and influence over hiring, wages and mobility (Granovetter; Lin). Debates concern which ties are most valuable: bridging ties to natives may span structural holes and improve outcomes; co-ethnic networks and ethnic enclaves can either enable mobility through networked opportunities or constrain workers within niches; transnational ties may aid high-status entrepreneurs but their value for low-skilled migrants is unclear. Intersectionality perspectives further suggest that the form, access to, and returns from social capital vary by gender and ethnicity, potentially reproducing inequalities as immigrant men and women have different network structures and labour market embeddings.

Design and context: Data come from a broader project on social cohesion in Catalonia, Spain. Fieldwork took place in 2009–2010, after a decade of rapid immigration growth and amid the Great Recession. Unemployment was ~14.8% overall and ~30% among immigrants. Sample: A personal network survey was administered using non-probability quota sampling stratified by city size, age, gender, and origin. In total, 446 interviews were conducted, including 153 immigrants (77 Moroccans, 76 Ecuadorians). Valid occupational status data were available for 137 cases. Recruitment involved institutions/organizations and diverse community venues; snowballing was minimized. Measures: For each respondent (ego), a fixed list of 30 named alters contacted within the last two years was elicited to capture both strong and weak ties. Detailed information was collected on alters and exchanges of employment-related information or assistance. Dependent variable: Occupational status (binary), distinguishing medium-high status (business owners, professionals, qualified workers, workers with management responsibilities) from low status (unqualified/low-skilled); most respondents were wage-earners (98.7%). Independent variables: Social capital measured as the number of alters who exchanged employment-related support, categorized by (a) autochthonous contacts (born in Spain), (b) co-ethnics in the country of origin (transnational ties), and (c) co-ethnics in Spain. For regression, each category was dichotomized at the median. Human capital measured by education: low (basic/compulsory), medium (post-compulsory secondary/professional), high (university). Sectoral segmentation captured by whether current/last job lies within an ethnic occupational niche (construction, agriculture, hotels and catering, retail, domestic work) versus other sectors. Controls: country of birth (Ecuador/Morocco), length of residence in Spain (years), gender. Not available: language competence and legal status. Analysis: Logistic regression models assessed associations between social capital, human capital, sector and occupational status, incrementally adding variables and then a full model. Interaction models tested whether returns to social capital, education, and niche sector differ by gender. Given small N, significance thresholds were set at 0.05 (**) and 0.10 (*). Descriptives: 40.9% had medium-high occupational status; 70.8% worked in ethnic niches; education: 43.8% up to primary, 38.0% secondary, 18.2% higher; 55.5% male; mean years in Spain 7.92 (SD 5.03). Mean number of labour-support alters: 14.07 (SD 7.85); autochthonous 11.14 (SD 12.08; 46.7% above median); in country of origin 10.46 (SD 12.34; 49.6% above median); co-ethnics in Spain 19.22 (SD 15.49; 46.7% above median). A contact-level crosstab showed higher likelihood of employment-related support exchanges with autochthonous alters (54.1%) versus transnational alters (34.6%).

- Social capital composition matters: Having above-median supportive ties with native-born Spaniards is positively and significantly associated with higher occupational status across models (β ≈ 0.94–1.03, p ≤ 0.05). In contrast, supportive ties with co-ethnics in Spain and with co-ethnics in the country of origin show negative but non-significant associations. Transnational ties are mobilized less for job-related help and do not predict higher status. - Ethnic occupational niches constrain mobility: Working within the immigrant niche sectors is strongly and negatively associated with attaining medium-high occupational status (β ≈ −1.15 to −1.44, p ≤ 0.05). - Human capital has limited returns under segmentation: High education shows a positive association in models without sector controls (β ≈ 0.94, p ≤ 0.10) but loses significance once sector is included, indicating that sectoral placement can nullify educational advantages for immigrants. - Gender inequality is pronounced: Being a woman is strongly associated with lower occupational status (β ≈ −1.15, p ≤ 0.05) net of other factors. Ethnic group (Moroccan vs. Ecuadorian) and length of residence are not statistically significant in the full model. - Gendered interactions: The benefit of autochthonous supportive ties persists overall, and the interaction with gender is not statistically significant; however, segmented percentages suggest a stronger practical effect for men (68.6% with such ties in qualified jobs vs. 34.1% without) than for women (34.5% vs. 25.0%). The interaction between working in an ethnic niche and gender is strongly negative (β ≈ −3.10, p ≤ 0.05), indicating women are particularly trapped in low-status positions within niche sectors. - Network-level support patterns: Autochthonous alters are significantly more likely to exchange employment-related support (54.1%) than transnational alters (34.6%), consistent with the positive effect of bridging ties on status attainment. Overall, the findings highlight that bridging social capital to natives facilitates occupational advancement, while ethnic niche employment and gendered segmentation restrict it; education alone is insufficient when structural constraints prevail.

The results challenge purely meritocratic accounts of immigrant labour market outcomes. Even controlling for education and sector, the composition of personal networks significantly shapes occupational status: bridging ties to native-born Spaniards provide valuable information and influence that translate into higher-status jobs, aligning with theories of structural holes and the strength of position. Conversely, co-ethnic and transnational ties do not yield similar benefits for these groups in Spain’s secondary labour market, suggesting that such networks can reinforce concentration in niches or lack the leverage needed for mobility at lower occupational levels. The strong negative effects of ethnic niche employment and of being a woman confirm the importance of labour market segmentation and intersectionality: immigrant women face compounded ethnic and gender barriers, with niche sectors functioning as a sticky floor that limits upward mobility. Education’s reduced returns once sector is considered underscores how institutional barriers (e.g., credential non-recognition) and demand-side structures mediate human capital, diminishing its payoff. These findings integrate network processes with structural segmentation, showing that both meso-level relational resources and macro-level labour demand shape immigrant inequalities.

The study demonstrates that social capital, human capital, and labour market segmentation jointly structure immigrant labour market outcomes in Catalonia. Bridging social capital—supportive ties to native-born Spaniards—consistently predicts higher occupational status, whereas transnational and co-ethnic supportive ties do not. Employment within ethnic occupational niches substantially constrains status attainment, and women are disproportionately penalized within these niches, evidencing a gendered sticky floor. Education alone cannot overcome sectoral and institutional constraints. Future research should: (1) use longitudinal designs and mixed methods to address causality and capture network evolution across life domains; (2) incorporate distinctions between host- vs. origin-acquired human capital and assess language proficiency; (3) include legal status, given its role in structuring rights and vulnerability; and (4) further explore gendered and ethnic intersectional mechanisms that condition the mobilization and returns of social capital.

- Causality: Cross-sectional data limit causal inference; reverse or reciprocal causation between occupational status and network composition is possible. - Sample size and design: Small N typical of personal network surveys reduces statistical power and generalizability; quota non-probability sampling limits population inference. - Measurement gaps: No data on language skills or legal status, both likely to affect job access and the value/returns of social capital; potential misrecognition of foreign credentials not directly measured. - Timing/context: Fieldwork during the Great Recession may influence employment patterns and network effects, potentially limiting temporal generalizability.