The essay consists of three chapters. In chapter 1, I find that wages in U.S. regions have been diverging instead of converging from 1975 onward. This coincides with the period of accelerating skill-biased technological change. A decomposition of the divergence rate indicates three channels underlying the divergence: (1) an ever-widening wage gap between college graduates and high school graduates, (2) an increasing within-education group wage differential across regions, and (3) a concentration of skill composition across local labor markets. I then developed an endogenous skill-biased technology adoption model in which firms invest capital more intensively in regions with higher employment share of college graduates, explaining these three channels jointly. Finally I quantitatively assess the model by separately calibrating the regional aggregate production function; the results show that the relative skilled-labor efficiency has been persistently higher in skill-abundant regions, nevertheless the countrywide skill-biased technological change, is the main force making divergence happening.

Chapter 2 studies energy-saving technological change in U.S. manufacturing sector, whose intensive margin and extensive margins are identified. I find that energy and capital are mostly complementary to each other, while labor is substitutive to energy-capital composite. However, a Cobb-Douglas nesting of labor is rejected. Quantitative exercise shows that in the post-crisis period, within in industry energy-saving technological change accounts for the largest proportion of the aggregate sectoral energy efficiency promotion in the long run. In contrast, in the short run, factor adjustment combined with sectoral shift accounts for the largest proportion of energy intensity reduction. Lastly, I provide evidence that structural change has taken place around the oil crisis in 1970s, which is consistent with the existing literature.

In chapter 3, I documented the increasing dispersion of skill composition across different areas in the U.S. Meanwhile, the U.S. Housing Market has experienced a dramatic increase in the housing price, as well as a similarly increase in its dispersion across metropolitan areas. A set of related stylized facts are documented in this paper. First, the real wage goes similarly as real housing prices, but quantitatively different. Second, the rents and housing prices have not been going in the exactly same way, in terms of first two moments. Third, we find that local income inequality is positively correlated to the local housing price level. Based on these observations, we build a model where a dispersed skill-biased technology change can account for all the phenomena at the same time.