Non-Material Determinants of Well-being: a Human-Centered Inquiry Across the Americas

This thesis is composed of three empirical studies, united by the premise that non-material dimensions – such as social capital, subjective happiness, and the environment – play a crucial role in influencing well-being, public health, and social cohesion. By focusing on different contexts (Brazil, the United States, and Latin America and the Caribbean) and distinct outcomes (mental health, substance use, and institutional trust), the three studies jointly investigate how social and ecological factors interact with individual vulnerabilities and societal stability. Together, they contribute to a broader, human-centered understanding of well-being that goes beyond material conditions and economic indicators, emphasizing the importance of social, psychological, and environmental determinants in shaping both individual and collective outcomes. The first paper, entitled “The therapeutic power of social capital: a quantitative study on happiness and mental health in peripheral Brazil”, investigates the protective and therapeutic role of social capital in shaping well-being among individuals with psychiatric disorders, mild psychological conditions, and a control group without diagnoses. Fieldwork was carried out in a peripheral neighborhood in Fortaleza, Ceará, producing 504 structured interviews on socioeconomic indicators, social capital, mental health status, and quality of life. Using nonparametric balancing techniques (PSM and npCBGPS) combined with OLS and Ordered Probit models, the analysis shows a positive association between social capital and subjective happiness across all groups, with bootstrapped confidence intervals confirming the robustness of the results. Nonlinearity tests indicate that individuals with moderate to severe conditions benefit most, although marginal gains diminish beyond a certain threshold. Overall, the study emphasizes the therapeutic role of social capital in preserving subjective well-being – especially among individuals with mild mental health vulnerabilities – and calls for targeted public health interventions in socially marginalized neighborhoods of Fortaleza. The second paper, entitled “Can subjective happiness protect against crack/cocaine use and injection drug use? An instrumental variable study of the U.S. population”, examines the causal relationship between happiness and severe forms of substance use. Addressing the problem of reverse causality, the paper specifically seeks to shed light on the directionality of the effect of happiness in its relationship with crack/cocaine use and drug injection. Drawing on nationally representative data from the U.S. General Social Survey and employing an instrumental variable approach (2SRI), results show that higher levels of happiness significantly reduce the likelihood of crack/cocaine use and drug injection, even after controlling for demographics, socioeconomic conditions, and adverse adolescent events. Exploratory analyses further indicate that individuals raised in large towns or suburbs at age 16 are more likely to use crack/cocaine, whereas living with both parents reduces risk; higher family income at age 16, however, is positively associated with substance use. These findings clarify the direction of the relationship, implying that happiness serves as a buffering factor against substance use and subsequent addiction. Policy implications stress the importance of prioritizing well-being, social capital, and collective resilience over punitive and prohibitionist approaches and individual responsibility frameworks. The third paper, entitled “Does forest cover loss undermine institutional trust? A cross-country analysis from Latin America and the Caribbean (2008–2023)”, studies the impact of forest degradation on citizens’ trust in public institutions. A composite measure of institutional trust was constructed from items assessing confidence in the president, congress, political parties, elections, municipality, police, and the army. Combining AmericasBarometer survey data with geospatial tree cover loss metrics from Global Forest Watch produced a macro-panel dataset of nearly 200,000 observations from LAC countries (2008–2023) with high spatial and temporal granularity. To address potential reverse causality, lagged fixed effects models were estimated using two indicators: general tree cover loss (≥10% initial canopy cover) and specific tree cover loss (≥50% and ≥70%) in critical biomes. Both measures show a significant negative association with institutional trust. Disaggregated analyses reveal particularly marked effects for trust in the army, congress, elections, and political parties. Moreover, short-term within-country and long-term structural forest loss across countries are linked with declining institutional trust. Forest cover loss constitutes an ecological crisis and a critical threat to the legitimacy of public institutions in Latin America and the Caribbean.