Files
Abstract
This dissertation covers three separate topics in the fields of health and labor economics. The first chapter estimates the causal effect of retirement and spousal retirement on health and health-related behaviors by exploiting variation in statutory retirement policies in a fuzzy regression discontinuity framework. Using data from the first three waves of the China Health and Retirement Longitudinal Study, I find that own retirement has adverse effects on subjective health status and mental health for both genders, but its effects on health behaviors are mixed. I also find that women produce beneficial retirement spillovers on husbands’ health, but do not experience any spillover effects themselves. Lastly, I provide estimates of the retirement effects under a counterfactual policy of raising the official retirement ages.
In the second chapter, I use Texas’s constitutional amendment in 1997 that expanded the scope of home equity loans as a source of exogenous variation to estimate the effects of relaxing credit constraints on small businesses. Using standard panel data methods and restricted-use microdata from the U.S. Census Bureau, I find that the Texas amendment increased the use of home equity finance by small businesses, increased new business and job creation and reduced firm exit and job loss. The effects are larger and significant for businesses with fewer than ten employees. In the third chapter, I examine how leisure-based physical activity affects earnings and wages. Using data from the American Time Use Survey and an instrumental variable strategy, I find that in the short run, physical activity does not enhance labor productivity; instead, it appears work time and physical activity are substitutes. A one-hour increase in average weekly physical activities decreases earnings by 1--2% within a location, but has no effect on wages. However, I find that in the long run, physical activity does enhance labor productivity. A one-hour increase in average weekly physical activities increases average earnings and wages by 6--7% for a location.
In the second chapter, I use Texas’s constitutional amendment in 1997 that expanded the scope of home equity loans as a source of exogenous variation to estimate the effects of relaxing credit constraints on small businesses. Using standard panel data methods and restricted-use microdata from the U.S. Census Bureau, I find that the Texas amendment increased the use of home equity finance by small businesses, increased new business and job creation and reduced firm exit and job loss. The effects are larger and significant for businesses with fewer than ten employees. In the third chapter, I examine how leisure-based physical activity affects earnings and wages. Using data from the American Time Use Survey and an instrumental variable strategy, I find that in the short run, physical activity does not enhance labor productivity; instead, it appears work time and physical activity are substitutes. A one-hour increase in average weekly physical activities decreases earnings by 1--2% within a location, but has no effect on wages. However, I find that in the long run, physical activity does enhance labor productivity. A one-hour increase in average weekly physical activities increases average earnings and wages by 6--7% for a location.