Essays in weather impacts on economy

• Autor(a) principal: Palialol, Bruno Toni
• Data de Publicação: 2020
• Tipo de documento: Tese
• Idioma: eng
• Título da fonte: Biblioteca Digital de Teses e Dissertações da USP
• Texto Completo: https://www.teses.usp.br/teses/disponiveis/12/12138/tde-14012021-204527/
• topic: Avaliação de impacto; Dados em painel; Impact evaluation; Panel data; Variações climáticas; Weather variations

This thesis examines three topics relating weather and economic outcomes. The first chapter, entitled \"Do temperature shocks affect non-agriculture wages in Brazil? Evidence from individual-level panel data\" asks whether there is a causal link between temperature and labor productivity in Brazil\'s formal labor markets. To that end, we use monthly individual-level panel data spanning January 2015 to December 2016. Our results show that a 1oC shock increases wages where climate temperatures are colder, but reduces wages where climate temperatures are hotter. We calculate that wages fall 0.42% on average, an income loss equivalent to 0.06% of GDP annually. Assuming future temperatures rise uniformly by 2oC, and that no adaptation occurs, we expect annual income losses five times larger. The heterogeneous effects we find also suggest that weather vulnerability may deepen existing income inequalities. The second chapter, entitled \"Does weather influence COVID-19 transmission?\" uses a global cross-region panel of daily data on 416 regions and 93 days, with multiple fixed effects, and explore the exogenous variations of weather variables to find that temperature reduces COVID-19 transmission up to 8.98%. These effects are stronger in cold, arid and dry climates, diminishing propagation by 12.86%, 12.08% and 10.08%, respectively. Still, it is not clear whether climatic conditions can alone stop or slow the outbreak, but they should be included in models to predict future cases of the disease. The third chapter, entitled \"Weather variations and impacts on mortality in Brazil\" uses an annual panel of 2,198 municipalities from 1980 to 2012 and interacts weather shocks with climate normals to estimate non-linear effects of climate on health. We find that positive precipitation shocks (+1mm) in a rainy season, increase mortality rates of children between 0 and 4 years old by 0.02%. Results suggest cities with low levels of education and infrastructure are more affected by temperature and precipitation variations. Our findings will help improving economic models by shedding light on key determinants of the weather-health relation.