Quantifying the socioeconomic implications of climate change: Three essays

Abstract

Efforts to quantify the future socioeconomic effects of climate change are important for present-day decision-making. Generating these predictions requires nuanced modeling approaches that accurately capture the complexity of the socio-environmental system whilst being grounded in real-world data. This thesis evaluates the effect of climate change on three different outcomes by combining large socioeconomic and climatic datasets, with modeling tools developed by different disciplines.

In the first study I look at how climatic changes may alter HIV prevalence through behavioral responses to poor agricultural years. Statistical relationships are found between warmer periods and behaviors that increase risk of HIV. Using results from approximately 400,000 individuals tested for HIV, I find that the probability of testing positive increases after multiple years of warming. A mechanistic model, parameterized with this statistical data, is used to investigate the implications of this association for long-run changes to the climate.

In the second study I investigate the effect of climate on varicella, a directly-transmitted infection. Laboratory studies for similar diseases have shown a relationship between humidity and transmission. With co-authors, I develop a novel methodology for testing the effect of climatic changes on the transmission rate for varicella that combines statistical tools with a mechanistic model to capture known dynamics of this disease. I find a significant relationship between drier days and increases in varicella transmission. The mechanistic model is used to simulate the effect of climate change on varicella incidence in Mexico.

In the final study I look at how temperature affects labor supply using an employment survey from six urban areas in Brazil. Labor supply changes are hypothesized to be one of the causal mechanisms driving the well-studied linkage between the climate and economic growth. I find an association between hot weeks and decreased work time, particularly in high risk sectors such as agriculture and mining. Using this estimated association, I project the total lost working time as climate change drives up temperatures in this region. I find that the coupled effect of urban heat islands and climate change may result in large losses in work-hours, unless cooling technologies are implemented.