Three Essays on U.S. State Public Service Expenditures and Population Health

Abstract

Across multiple population health measures, the United States (U.S.) underperforms relative to peer, high-income nations. Compared to other high-income nations, the U.S. had the lowest life expectancy at birth, the highest death rates for avoidable or treatable conditions, and the highest rate of multiple chronic conditions in 2020. While much research has analyzed the relationships between health expenditures, healthcare service consumption, and health outcomes in the U.S., less has focused on determining the influence of social and environmental services on health outcomes. Furthermore, less research has focused on identifying the subgroup-specific health effects of these services across individual-level social characteristics such as socioeconomic status (SES), race, and gender. Social determinants of health theories, including fundamental cause theory (FCT) and eco-social theory, posit that social factors (e.g., SES, gender, and race) comprise fundamental causes of disease, due to social structures that distribute power and other health-relevant resources differentially on the basis of these characteristics. Similarly, eco-social theories of health emphasize the importance of environmental context in addition to social context for health outcomes.Building on social determinants of health and eco-social theoretical frameworks, the present analysis studies the associations between U.S. states’ investments in healthcare, social, and environmental services and population health outcomes, considering variation in these associations across marginalized subpopulations stratified by intersections of race, gender, and class. Chapter 2 analyzes the associations between the ratio of states’ social to healthcare spending and infant mortality rates (IMR) and heart disease mortality rates (HDMR); Chapter 3 considers the differential impacts of state-level investments in parks and recreation services on obesity prevalence across SES; and Chapter 4 the association between states’ investments in social services and self-rated health across subpopulations stratified by race and gender. Chapter 2 finds that spending on social services is associated with reductions in IMR and HDMR with greater time-lags (five-year time-lags), as hypothesized; healthcare spending is associated with reductions in IMR across all time-lags, however, for the HDMR outcome, healthcare expenditures are only negatively associated with HDMR in the unadjusted models. These results aligned with our hypothesis that social services, which are more likely to comprise “proactive” interventions addressing upstream socioeconomic inequalities, would demonstrate the largest effects on health outcomes when outcomes are lagged behind expenditures. In contrast, we hypothesized that healthcare expenditures are more likely to comprise “reactive” interventions, reflecting spending that responds to existing disease burdens; as such, we expected the magnitude of the association between healthcare expenditures and outcomes to be largest in the contemporaneous model and to decrease with increasing time-lags. This hypothesis was confirmed for HDMR, but not for IMR. Chapter 3 finds that the health benefits of states’ investments in parks and recreation services accrue disproportionately to low-income relative to high-income individuals, with low-income individuals experiencing the greatest reductions in obesity prevalence, followed by middle-income individuals, and lastly high-income individuals. This finding aligned with our hypothesis that individuals with lower baseline levels of resources would experience larger benefits from incremental increases in parks and recreation service provision, in accordance with previous evidence demonstrating diminishing marginal returns to socioeconomic resources as SES rises. Chapter 4 finds that, across all race-gender intersections of Hispanic or Black and white men and women, greater social spending was associated with the largest benefits to self-rated health for Black or Hispanic men, followed by white women, Black or Hispanic women, and lastly white men. In adjusted models, a divide across race arose: social expenditures were associated with superior self-rated health for white men and women, and worse self-rated health for Hispanic or Black men and women.These results illuminate the multifaceted and complex nature of the relationship between states’ investments in healthcare, social, and environmental services and population health outcomes across individual-level social characteristics such as race, gender, and class. Consistent with previous evidence that pro-welfare state-level spending environments benefit population health outcomes (Fenelon and Witko 2021; Riley et al. 2021; Montez et al. 2022), we find that increased social expenditures improve both IMR and HDMR outcomes, though, as supported by our FCT-motivated hypothesis, this negative association only arises with time-lags. As suggested by studies of intersectionality and the cumulative impacts of multiple discriminatory social institutions on population health outcomes, the present analysis reveals that the association between state-level spending on environmental and social services varies across SES, race, and gender. Social service expenditures advantaged the self-rated health of whites, while adversely affecting the self-rated health of Blacks and Hispanics, and environmental services disproportionately benefited low-income relative to high-income individuals. In sum, while certain services disproportionately benefited individuals who experience more marginalization (e.g., environmental services disproportionately benefited low-income individuals), other services disadvantaged more marginalized individuals (e.g., social services delivered health benefits to white individuals, and health disadvantages to Black or Hispanic individuals). Policymakers aiming to allocate public service expenditures in order to optimize population health outcomes should thus bear in mind the sensitivity of the associations between expenditure categories and health outcomes to subpopulation-level differences. The present findings are consistent with previous intersectional analyses of health disparities, highlighting the complex nature with which social and environmental factors combine to influence health outcomes. Future research should continue to interrogate the health impacts of public expenditures on social, healthcare, and environmental services across gender, race, and class, as increasing efforts are directed towards optimizing healthcare expenditure efficiency and addressing the social and environmental determinants of health.