UC Berkeley

A substantial body of empirical evidence has found a connection between education and health. An emerging area of research interest is links between characteristics of school settings and health. An inductive case can be made: characteristics of the school setting, including school inputs, student body demographics, and school culture, are associated with educational attainment, which is, in turn, a key social determinant of health. This dissertation assesses the links between characteristics of the school setting and health outcomes directly.

The existing small evidence base assessing the association between class size reduction and student health outcomes is inconclusive. I conducted a quasi-experimental analysis to evaluate the impact of North Carolina’s elementary class size reduction policy on student sedentary behavior, using an instrumental variables approach. I observed no association between class size and screen time (recreational television and/or electronic device use), after accounting for grade size and school size, year fixed effects, and clustering at the school and district level. These findings suggest that, in state-wide policy implementation settings, there do not appear to be any immediate spillover benefits of class size reduction policies on student health.

Little is known about the extent to which characteristics of the school setting are associated with health outcomes in adulthood, especially in recent decades. I analyzed data from a recent, nationally representative American cohort, the National Longitudinal Survey of Youth 1979 cohort, to explore this question. After adjusting for confounders, high school socioeconomic composition, but not racial/ethnic composition or dropout prevalence, was weakly associated with both obesity and poor self-rated health at age 40. However, after adding adult educational attainment to the model, only the weak association between high school socioeconomic composition and obesity remained statistically significant. Future research should explore possible mechanisms, and also if findings are similar for elementary and middle school composition. These results suggest that policies that seek to break the link between socioeconomic composition and negative outcomes, like Title I, remain important but may have few spillover effects onto health.

Researchers consider positive school climate to be multidimensional and essential for a supportive school setting. I systematically review the relationship between school climate and students’ mental health in the K-12 grades. I identified 40 studies that examined the association between school climate and psychological functioning and met the eligibility criteria. The majority of studies used internalizing and/or externalizing symptoms as the outcome(s) of interest; measures of school climate were more heterogeneous. U.S. middle and high school populations were most frequently studied. Ninety percent of studies found an association between at least one measure of school climate and at least one domain of psychological outcomes. However, there was also room for potential single source bias, since most articles used student reports of both school climate and mental health, and potential residual confounding, since results sometimes differed depending on if the study was cross-sectional or longitudinal in nature.

In conclusion, it appears that some characteristics of the school setting are associated with some health outcomes, although it depends on the specific school setting constructs and health outcomes and the time point at which the health outcomes are measured.