Aperçu

Résumé (français)

Veuillez noter que les résumés n'apparaissent que dans la langue de la publication et peuvent ne pas avoir de traduction.

Résumé (anglais)

This thesis comprises of three chapters that study the association of economic conditions with the health and health behaviours of adults and the health outcomes of infants. The first chapter examines the impact of economic conditions on modifiable health behaviours, weight-related health (proxied by BMI), self-assessed mental health, life stress and general health of Canadians using data from the first cycle of the Canadian Community Health Survey (CCHS) to the year 2013. More specifically, using variation in unemployment rates at the Census Metropolitan Area (CMA) level, we estimate two-way fixed effect models to investigate how local labour market changes affect these health outcomes. Results suggest that most estimates are small and not statistically significant but that important exceptions exist which imply that recessions pose threats to the health status of the population and need to be appropriately considered by health planners and practitioners alike. In the second chapter we use the Canadian Vital Statistics Birth and Death database over the 1976 to 2011 period to investigate how economic conditions are associated with infant mortality, neonatal mortality (death within the first 28 days of life), postneonatal mortality (death from 28 days onward to less than 365 days), the crude birth rate, birth weight, the percentage of babies born weighing less than 2500 gram, and the percentage of babies born weighing less than 1500 grams. A model with province and year fixed effects, clustered at the provincial level, is estimated. This model differences out time invariant unobserved heterogeneity over the period studied and provides unbiased coefficient estimates. For example, some geographic areas may experience both poor health and high unemployment but a causal relationship does not exist. If individuals within provinces have similar unobservable characteristics errors may be correlated – clustering addresses this to provide robust standard errors. The results of this study suggest that worse economic times negatively impact the health of infants. The third chapter re-examines the relationship between infant health outcomes and economic conditions studied in chapter 2 undertaking a sensitivity analysis related to the choice of economic proxy, model specification and the time frame of analysis. This study confirms that worse economic times are not good for infant health, however, we find that using the employment rate there is a robust relationship between good economic times and infant health.