Content area
This dissertation consists of two chapters that explore the relationship between economic structure and the dynamic properties of the macroeconomy.
In the first chapter, I construct a three-sector general equilibrium model consisting of the goods sector, the service sector, and home production with input-output linkages to study how the share of home production affects business cycle volatility. Under a standard CES specification with empirically relevant elasticities of substitution, I show that a decline in the home production share dampens business cycle volatility through two channels: a home production share channel and a market composition channel. The home production share channel reflects that the amplification effect of home production on business cycle volatility depends on the share of home production in the economy. Under empirically relevant parameters, the home production share channel works to dampen business cycle volatility because the amplification effect of home production becomes less significant from the market’s perspective as the home production share decreases. The market composition channel captures the impact of the change in the home production share on business cycle volatility through shifts in the market service share. Intuitively, as the share of home production decreases, the market service share increases, which in turn reduces business cycle volatility since value added in the service sector is less volatile than that in the goods sector. The calibrated model predicts that the reduction of the share of home production accounts for 8-10 percent of the reduction of business cycle volatility from 1950 to 2010 in the U.S. economy. The model also predicts that the difference in home production share accounts for 17-19 percent of the gap in business cycle volatility between developed and developing economies.
In the second chapter, I study the relationship between demographic structure and the macroeconomic impacts of labor income tax shocks theoretically and empirically. A general equilibrium overlapping generations model calibrated to the U.S. economy predicts that from 1976 to 2021, the impulse response of real GDP per capita to labor income tax shocks declined by 7-10 percent, and the labor income tax multiplier decreased by 6-8 percent due to aging. Consequently, to generate a labor income tax multiplier similar to that of the 1970s, the government needs to increase the size of tax shocks for the young by 75 percent. I document that age-specific capital-labor complementarities and Frisch elasticities of labor supply are key determinants of the relationship between demographic structure and the macroeconomic effects of labor income tax shocks. Intuitively, lower capital-labor complementarity makes labor income tax shocks more effective, as the impact of additional hours on output is less constrained by the sluggish response of capital stock. Due to the low capital-labor complementarity of the young, the impact of tax shocks declines as the population ages. In addition, the Frisch elasticity of labor supply of the young is higher than the average Frisch elasticity of the middle-aged and old. I document that the lower capital-labor complementarity of the young, compared to the middle-aged and old, together with this higher Frisch elasticity, leads to a larger impulse response of hours of the young than the average response of the middle-aged and old. These factors reinforce each other and reduce the impact of labor income tax shocks as the population ages. In addition, by exploiting U.S. state-level variations in demographic structure, I empirically find that U.S. states with a higher population share of the young respond more significantly to aggregate labor income tax shocks than states with a lower share of the young, which is consistent with the model’s prediction.