Essays on Population, Environment and Development

Ecological factors and the policy environment are central constraints on population well-being. This dissertation emphasizes the role of shocks to help understand the nature of such constraints, and explores the relationship between population, environment, and development in greater detail than is typically possible.

Chapter 1 opens by contributing to a growing body of evidence around the impacts of old-age pensions on the well-being of pension recipients and their families. I draw from the unique disbursement structure of a popular, widely utilized benefits program in rural Brazil, and data from two nationally representative surveys conducted in 2013 by the Brazilian Institute of Geography and Statistics. I first employ regression discontinuity design to measure the direct effect of the program’s age threshold on pension take-up. Second, I compare differences in reports of health and well-being among age-eligible and age-ineligible adults in rural areas to the same differences among populations that generally do not qualify for the benefit (i.e., urban populations). This difference-in-differences shows robust evidence of a beneficial pension effect, though along somewhat different dimensions by gender. I then show evidence of two credible mechanisms for improved health and well-being: first, improved food security within households that have eligible pension recipients; second, the cohabitation of younger family members, potentially providing support to aging family members. Taken together, this chapter demonstrates that the rural benefits program in Brazil leads to tangible health benefits for its recipients, through channels that are likely to complement rather than crowd out other public services.

Chapter 2 moves on to explore how a massive natural disaster affected smoking behavior, a common coping mechanism. External stressors are commonly hypothesized to play a role in the adoption of certain health behaviors, but understanding the role of exposure is frequently hampered by research designs and data that are inadequate for tracing causality. I use this study to evaluate the relationship between unanticipated exposure to a natural disaster and smoking behaviors using longitudinal data collected from families in Aceh, Indonesia before and after the 2004 Indian Ocean tsunami. Exposure to the tsunami is measured as a community indicator of physical proximity to damage, and as individual indicators of personal experiences at the time of the tsunami. My analysis indicates that the effect of exposure on smoking uptake varies considerably by age, and that most forms of exposure increase smoking volume. These effects appear to be temporary, but even in the context of Indonesia’s extraordinarily high smoking rates an impact is perceptible even ten years after the tsunami.

Chapter 3 delves further into the effects of the tsunami, exploring the distribution of resources after the broad destruction of infrastructure and subsequent, fast-paced reconstruction. I evaluate multiple aspects of water access for roughly 6,000 families through 2014. Logit regression analysis show increasing disparities in access to basic amounts of water, and multinomial logit regression analysis indicates that shifts are driven by a massive increase in the market for privately distributed bottled water. This study disentangles key distributional processes to show how reconstruction influenced a central social determinant of health among an already vulnerable population.

The chapters to follow aim to relate the well-being of individuals to the influences that arise from interconnected policy choices and ecological factors. The first chapter emphasizes a policy shock, the second an ecological shock, and the second seeks to identify a combined effect of the two. This original research is intended to help illuminate the role that institutions might play in improving population well-being.