Date of Award

5-2018

Document Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

Department

Economics

Committee Member

Dr. Kakin Kevin Tsui, Committee Chair

Committee Member

Dr. William Dougan

Committee Member

Dr. Chungsang Tom Lam

Committee Member

Dr. Aspen Gorry

Abstract

There are three chapters in this dissertation. In the first two chapters, I focus on the deceleration of college wage premium growth at national level. In the first chapter, I use a decomposition and counterfactual simulation approach to evaluate three explanations of flattening college wage premium during 2000-2010 and 2000-2015, i.e., demand reversal, polarization and supply change. The simulation results suggest: i. supply change is the most powerful explanation; ii. polarization has expla-natory power but not robust; and iii. demand reversal has little eÿcacy in explaining the fact. By a close look of three explanations, I argue that the “failure” of demand reversal story and the “unstable” impact of polarization story indeed suggest some kinds of weakness of the analytical framework. And there are still some unexplained facts related to the supply change story. A better framework should depend on the endogenous supply of high-skilled workers and heterogenous quality of high-skilled workers. Due to the weakness of the analytical framework in the first chapter, in the second chapter, I use a di˙erent framework to study the deceleration of college wage premium growth. The novel feature of the framework is that the supply of high-skilled workers endogenously depends on the progress of technologies and other shocks. I show the change in progress basis of skill-biased technologies is the primary reason that leads to the deceleration relative to the change in progress speed of skill-biased technologies and the progress of automation. This result suggests the deceleration of college wage premium growth is an inevitable outcome of skill-biased technological change while both demand reversal and polarization have only moderate explanatory power. About the direct mechanism of the deceleration, I find that the faster decline in the mean quality of high-skilled workers followed by a greater o˙set of skill-biased technological change’s positive impact on college wage premium growth due to increase in the progress basis of skill-biased technologies is the primary direct mechanism. Unlike the first two chapters, the third chapter looks into sub-national level labor market inequality change. This chapter studies a spatial pattern and a possible channel of local labor market inequality change. That is, large cities have the greater losses in the declining industries and greater gains in the growing industries. When the declining and growing industries are low-skilled and high-skilled intensive respectively, and the growing industries are skill intensive, the losses and gains of industries lead to a greater increase in the local labor market inequality of those first larger cities. The empirical results show that, one standard deviation change in initial city size accounts for 71.3% to 80.5% of one standard deviation change in industrial composition where the industrial composition change refers the losses in manufacturings (NAICS 31-33, low-skilled worker intensive) and gains in professional services (NAICS 52, 54, and 62, high-skilled worker intensive) here. One standard deviation change in the industrial composition change accounts for 79.5% to 89.8% of one standard deviation change in the local labor market inequality. And overall, the one standard deviation change in initial city size accounts for 62.1% to 65.1% of one standard deviation change in the local labor market via the industrial composition change channel. These empirical results verify the spatial pattern and the channel. And the paper compliments many related discussions in various ways.

Share

COinS