Essays on Human Capital and Social Network Effects

The first chapter of this dissertation looks at the relationship between parental investment in daughters' human capital and marital transfers in India. I evaluate the existence, nature, and magnitude of a causal relationship between a daughter's education level and the price her family pays for a groom in India using nationally representative data and a two stage least squares instrumental variable estimation methodology. I construct a policy instrument using an Indian national school-latrine-construction initiative to estimate the effect of an exogenous increase in a woman's years of education on the groom price that her parents pay. My OLS estimates show that a woman with an additional year of education pays on average an extra Rs. 14,314 (USD 298) for her groom price, while my 2SLS results indicate that a woman with an additional year of education pays on average an extra Rs. 22,283 (USD 464) for her groom price, which is 6 percent of the average groom price in my sample. However, when I account for their groom's education, my estimates indicate that an increase in a woman's education results in her marrying a groom with an additional 0.5 years of education. These results suggest that while a woman's own education has a negative effect on her groom price, she pays an extra amount of groom price for each additional year of her groom's education with the total effect of a woman with more education paying more groom price on average. The second chapter looks at whether farmers' irrigation decisions in the Southeastern United States are affected by social or peer behavior. This paper looks at whether the adoption of irrigation by a county is influenced by that of its' neighboring counties. I estimate a peer-effects model to investigate the effect that neighboring counties have on each other's likelihood to irrigate using three-year panel data from the US Department of Agriculture censuses of 2002, 2007, and 2012 for the 439 counties in Alabama, Florida, Georgia, North Carolina, and South Carolina. Two-stage least squares instrumental variable fixed effects estimations suggest that after controlling for farm operator and farm characteristics, the extent of irrigation among neighboring counties positively and significantly affects the percentage of farmers who irrigate in a county. This suggests that learning from others may be one of the mechanisms through which farmers in the Southeastern US make their irrigation decisions. The results also suggest that larger farms (in area) and farms with operators who are primarily farmers, are more likely to irrigate. My results also suggest having peers of the same race category may have an effect on irrigation when the category is a very small group as a percentage of all farm operators in a county on average.