Socioeconomic status and social capital as predictors of happiness: evidence and gender differences

The study investigates how socioeconomic status (SES) and social capital (community solidarity, generalised trust, and locus of control) predict happiness among South African adults and whether these relationships differ by gender. Prior research links lower SES to poorer subjective health and life satisfaction and suggests social capital facilitates wellbeing, but findings vary by context and measurement. Given South Africa’s complex history, multiculturalism, and persistent inequalities, components of social capital such as trust, solidarity, and empowerment (locus of control) may play distinct roles in wellbeing. The study aims to: (1) examine SES as a predictor of happiness, (2) assess associations between social capital components and happiness, and (3) test gender as a moderator of these relationships within the South African context.

Research shows SES relates to subjective wellbeing and health, with lower SES associated with poorer outcomes (Hoffmann et al., 2018; Proctor et al., 2017; Wang & Geng, 2019). Social capital, encompassing social interactions and participation, has been linked to health promotion and wellbeing (Ehsan et al., 2019; Wind & Villalonga-Olives, 2019). Studies in high-income settings find SES may enable autonomy and control, promoting happiness (Fournier, 2020), and SES indicators are associated with happiness with gender interactions present (Maharlouei et al., 2020). European evidence indicates social capital predicts happiness, with SES and demographics (age, gender) influencing the social capital-happiness link (Rodríguez-Pose & von Berlepsch, 2014). However, the formation and meaning of social capital differ in developing countries, and measures vary across studies, yielding inconsistent findings. In South Africa, trust, solidarity, and empowerment are salient for wellbeing amid historical segregation and inequality; trust has been tied to mental wellbeing among older adults (Chipps & Jarvis, 2016), solidarity supports mutual aid (Laitinen & Pessi, 2014), and locus of control is crucial for health, especially among disadvantaged groups (Kesavayuth et al., 2020). These contextual and cultural factors motivate examining how these social capital facets and SES relate to happiness in South Africa and by gender.

Design and sampling: Cross-sectional online survey conducted across South Africa’s nine provinces from January to September 2021 using LimeSurvey (English language). Initial N=1062; 13 cases with extensive missing data removed. Given the focus on gender, participants identifying as “other” or preferring not to disclose gender (n=49) were excluded from inferential analyses. Final analytic sample N=1000. Recruitment used snowball sampling via social media (Facebook, Twitter, WhatsApp), project webpage, personal and professional networks. Participants received a 10 Rand mobile airtime voucher incentive. Sample characteristics: Mean age 26.36 years (SD=7.20; range 18–45); 62.5% female. Racial identity: 78.6% Black African, 11.1% Coloured/mixed, 7.8% White, 2.5% Indian. Educational attainment ranged from none (2.0%) to doctorate/postdoctoral (3.3%); income predominantly in the “poor” category (68.1%). Provinces represented all nine provinces, with Gauteng (36.8%) and North West (24.8%) largest. Measures: - Happiness: Single-item 5-point Likert scale from “very unhappy” to “completely happy,” validated for subjective life evaluation. - Social capital: • Community solidarity: Two items from the World Bank Social Capital Scale assessing likelihood of community cooperation and willingness to help; 5-point Likert, summed 2–10; higher scores indicate stronger solidarity; Cronbach’s alpha=0.33 (noted limitation due to two-item scale). • Generalised trust: Single dichotomous item (“most people can be trusted”=1 vs “can’t be too careful”=0) from SRC study, established as a valid trust measure. • Locus of control: Single item from World Bank Integrated Questionnaire (“Do you feel able to make important decisions that will affect the course of your life?”) rated 1 (totally unable) to 5 (totally able). - Socioeconomic status (SES): Composite index aggregating household income category (per South African thresholds: poor to emerging affluent), highest educational attainment (CASMIN-based categories), and occupational ranking relative to qualification (4 options). Total SES range 3–18; higher scores indicate better SES; categorised as very low (3–6), low (7–9), moderate (10–12), high (13–15), very high (16–18). Aggregation followed Lampert et al. (2013). Data analysis: Conducted in SPSS 27. Checked normality via skewness and kurtosis (all within acceptable bounds: skew<2, kurtosis<7). Outliers screened via z-scores and Mahalanobis D2 (no significant concerns). Descriptive statistics computed; bivariate correlations among SES components, social capital variables, and happiness with Cohen’s guidelines (r=0.10 small, 0.30 medium, 0.50 large). Hierarchical regression examined predictors of happiness: Model 1 included SES, community solidarity, generalised trust, locus of control, and gender (dummy coded 1=male, 0=female). Model 2 added interaction terms (SES*gender, community solidarity*gender, locus of control*gender, trust*gender) to test gender moderation. Multicollinearity assessed via Tolerance and VIF (all acceptable; VIF<5 and tolerance >0.4 thresholds). Model fit assessed by F-statistics; moderation by significant ΔR2 when adding interactions.

- Correlations (N=1000): • Happiness correlated positively with SES (r=0.13, p<0.01), community solidarity (r=0.35, p<0.01), locus of control (r=0.39, p<0.01), and negatively with generalised trust (r=-0.08, p<0.05). Age correlated negatively with happiness (r=-0.12, p<0.01). Gender (1=male) correlated negatively with happiness (r=-0.12, p<0.01), indicating higher happiness among females. - Hierarchical regression predicting happiness: • Model 1 (SES, social capital variables, gender): R2=0.25, F(5,994)=67.58, p<0.001. Significant predictors: SES (β=0.09, SE=0.03, p<0.01), community solidarity (β=0.27, SE=0.03, p<0.001), locus of control (β=0.33, SE=0.03, p<0.001; unique variance sr2≈0.09), generalised trust (β=-0.08, SE=0.03, p<0.01), gender (β=-0.09, SE=0.06, p<0.01; females happier). • Model 2 (adds interactions): R2=0.26, ΔR2=0.01, ΔF=3.79, p<0.05, indicating overall moderation by gender. Significant interactions: community solidarity*gender (β=0.08, SE=0.06, p<0.05), generalised trust*gender (β=0.08, SE=0.06, p<0.05). Locus of control*gender not significant (β=-0.03, p=0.40). The text also interprets SES*gender as indicating a stronger SES–happiness association for women. - Interpretation of moderation: Association between SES and happiness is stronger for women; associations between community solidarity and happiness, and between generalised trust and happiness (negative), are stronger for men; locus of control’s association with happiness does not differ by gender. - Overall, SES and social capital indices explain approximately 25% of the variance in happiness, with locus of control showing the largest unique contribution.

Findings support that higher SES (income, education, occupational status) is associated with greater happiness, consistent with literature linking SES to improved resources, social engagement, and health. In South Africa, where inequality and poverty are pronounced, SES emerges as a key determinant of life outcomes and happiness. Community solidarity predicts higher happiness, likely reflecting the psychological benefits of belonging, mutual aid, and social support, which may serve as coping mechanisms in a historically segregated, multi-group society. Locus of control shows the strongest direct association with happiness, underscoring the role of perceived control and empowerment in motivating optimism, proactive behavior, and better handling of life challenges—particularly salient in contexts of uncertainty and socioeconomic hardship. Unexpectedly, generalised trust is negatively associated with happiness, which may reflect South Africa’s sociohistorical context of intergroup distrust and public distrust amid widening inequalities. Gender moderates some associations: SES relates more strongly to happiness for women—possibly due to greater socioeconomic barriers and unemployment faced by women—while community solidarity and the (negative) trust-happiness link are stronger for men, potentially tied to gender role socialization regarding support seeking and trust dynamics. Locus of control’s effect does not differ by gender, indicating its broad importance for wellbeing across genders in this context. These results address the research objectives by demonstrating significant roles for SES and specific social capital components in predicting happiness and by identifying gender-specific patterns relevant for targeted interventions.

The study provides evidence that SES and social capital—particularly community solidarity and locus of control—are significant predictors of happiness among South African adults, explaining about a quarter of its variance. Gender differences emerge: the SES–happiness link is stronger for women, while community solidarity and trust (negatively) relate more strongly to happiness for men; locus of control is equally important across genders. These insights suggest that policies and programs that improve SES (education, income, employment), promote social participation and cooperation, and empower individuals to exert control over life decisions are likely to enhance happiness. Gender-tailored strategies that account for differing social and economic realities may be especially effective. Future research should further unpack cultural and group-specific dynamics and employ longitudinal designs to clarify causality.

- Cross-sectional design prevents causal inference and limits generalizability. - Social capital is a contested, multifaceted construct; the operationalization here may omit relevant components affecting happiness. - Non-probability snowball sampling and English-only survey may bias the sample and exclude individuals with limited English proficiency, reducing representativeness. - Analyses treat South Africans as a single group, potentially obscuring racial, cultural, and regional differences. Longitudinal and more inclusive studies addressing group diversity and broader social capital dimensions are needed.