Essays on Economic Development and Urbanization Open Access
Downloadable ContentDownload PDF
This dissertation consists of three separate essays that empirically investigate issues in economic development and urbanization. Chapter 1 investigates the role played by cultural and political institutions in determining gender norms. Chapter 2 explores the role played by early-life rainfall in determine individuals' human capital and socioeconomic outcomes later in life. Chapter 3 uses novel data to document new stylized facts about the extent of city areas and how these are related to city size.In Chapter 1, I explore how cultural values regarding gender norms interact with political institutions to determine human capital outcomes. I investigate this question by exploiting a unique quasi-experimental setting that arises in the history of Pakistan: the nationwide strengthening of orthodox religious groups during Zia-ul-Haq's Islamization program, beginning in 1977. I document evidence using a triple-difference analysis, exploiting cohort-level variation in exposure to the new regime and regional variation in the preexisting strength of Islamic parties. Regions where Islamic parties had been strong beforehand experience a sharp decline in primary schooling for girls relative to boys during the Islamization regime. Evidence on changes in sex ratios and child mortality is also consistent with these results being due to a reduction in household investment in female children. My results suggest a mechanism of penetration of norms promulgated by Islamic organizations in regions with weaker institutional environments. The institutional shock signals a decrease in the relative payoffs of human capital investments in girls and households adjust child investment behavior accordingly.In Chapter 2, I examine the role of rainfall shocks at the time of birth in determining adult outcomes for individuals born in rural Pakistan. My identification strategy exploits regional and time variation in annual rainfall. I find that individuals exposed to higher rainfall during the time of birth are more likely to have attended school, have better health outcomes as measured by height, and are more likely to live in wealthier households. That these results are not found to hold true for individuals born in urban areas suggests that rainfall at birth affects adult outcomes through its effects on rural households' agricultural incomes. I also show that my results are not driven by extreme rainfall events, and are due to the productivity shocks induced by variation in rainfall closer to the mean. Heterogeneity analysis reveals important gender differences in how early-life effects on human capital attainment translate into adult socioeconomic status, and an important role of market access in determining the salience of rainfall-induced productivity shocks.In Chapter 3, coauthored with Harris Selod (World Bank) and Brian Blankespoor (World Bank), we construct a new dataset that provides us with measures of population, areas and their growth between 1990 and 2015 for over 4000 cities in more than 100 countries. Using this dataset we examine stylized facts about the joint distribution of city areas and populations. We find cross-section partial correlation between city populations and city areas in line with previous estimates in the literature, and exploit the time dimension of our dataset to document the partial correlation between their growth rates as well. The large geographic coverage of our dataset allows us to check the robustness of these relationships to a series of fixed effects specifications and robustness checks. We find evidence for the existence of Zipf's law for city areas that closely follows the well documented Zipf's law for populations. Lastly, we use our data to examine the role of changes in built-up density at the city core and city edge in determining the spatial expansion of cities.