The effects of electrification and droughts on rural households in Brazil

In my dissertation, I explore three topics that intersect two fields of economics, environmental and resource economics and development economics. I choose Brazil as a research background. Brazil is a large and diverse country in terms of demographic characteristics and socioeconomic status as well as geographic and climate features, which makes the country an ideal setting for research on socioeconomic outcomes. The keywords for the topics are electrical energy infrastructure and droughts. First, energy infrastructure has reemerged as a powerful driver of economic development in developing countries, but the causal impacts are still controversial. I pay attention to rural household electrification and its impacts on human capital of children. Second, future predictions of global climate change indicate that occurrences and severity of drought phenomena will be increasing. Drought causes adverse socioeconomic shocks to the local economy, in particular to rural communities. My two research projects on drought examine how drought affects vulnerable rural households’ income and fertility outcomes. My dissertation therefore consists of three chapters; the first chapter investigates the effects of rural household electrification on children’s schooling and work in Brazil between 1991 and 2000. The second chapter examines the impacts of droughts on women’s fertility in rural Brazil between 2000 and 2010. In the last chapter, I study the impact of drought on income inequality in Brazil between 1980 and 1991. In Chapter 2, I investigate the development effects of rural household electrification, focusing on school attendance and different types of work of children in Brazil during the period between 1991 and 2000. Using Brazilian Census data and an instrumented electricity measure, I estimate short-run effects of household electricity service on school attendance, market and non market working status, and housework inside the home done by girls. I find that household electricity service raises school attendance and reduces the likelihood that a girl stays home from school to do housework her own home. Household electricity also increases the probability of working for pay for both genders and makes boys less likely to do unpaid work to help family farm or enterprise. The findings suggest implications for behavioral responses of rural households in middle-income countries to household technical changes—people choose to invest in children's human capital, and increase market labor of children for both genders, which may substitute non market labor including housework of girls and boys' unpaid work at family farm or firms. In Chapter 3, I examine the effects of local droughts on rural women’s fertility in Brazil. Using Brazilian Census data and a drought measure that considers temperature and precipitation, the self-calibrated Palmer Drought Severity Index (sc-PDSI), I estimate short-run and medium-run effects of drought episodes on fertility outcomes between 2000 and 2010. I find that drought exposure increases the contemporaneous fertility by 22 percent and raises the number of surviving children in the medium-run by 0.3. The findings suggest implications for behavioral responses of vulnerable rural households in middle-income countries exposed to drought risks from climate change. Drought-induced substitution effect combined with income effect and rising child mortality may induce people to choose higher fertility. In Chapter 4, I research the heterogenous effects of local droughts on rural household incomes in Brazil. Using Brazilian Census data and a drought measure that considers temperature and precipitation, the self-calibrated Palmer Drought Severity Index (sc-PDSI), I estimate short-run and medium-run effects of drought episodes on real incomes of rural households in 1980 and 1991. Using an unconditional quantile regression approach, I find that in the short-run, drought exposures intensify the income inequality within the rural sector in 1980. Meanwhile, in the later decade, negative drought-driven income shocks are larger at upper quantiles than at low quantiles possibly due to increase in market access. In the medium-run, interestingly, the positive impacts of drought are greater at higher quantiles in 1980. By contrast, in 1991, the medium-run impacts of droughts are negative at all the quantiles. The increase in market access may explain the differences in the medium-run impacts between two periods.Ph.D.Includes bibliographical reference