This dissertation studies the impact of technological progress on various aspects of the U.S. labor markets such as the recent decline in the labor's share of income, job and wage polarization, rising income and wealth inequalities and skill accumulation. Chapter 1, "Capital-Task Complementarity and the Decline of the U.S. Labor Share of Income," studies how changes in occupational composition of the labor force contributes to the recent decline of the US labor share of income. Following the job polarization literature and classifying labor by tasks performed, I estimate unitary elasticity between equipment capital and labor performing non-routine tasks. This implies that income loss of labor em- ployed in routine task occupations is the main driver of the decline of the aggregate labor share. For a given path of technological change, decline of the labor share is larger when: (i) the substitutability between equipment capital and routine tasks is stronger, and (ii) equip- ment capital has a larger weight in production. Furthermore, a dynamic general equilibrium model shows that the impact of permanent technology shocks on the labor share gets smaller as the fraction of routine task labor declines. Consistent with this, the model predicts that the labor share should stabilize at around 55% in the long-run even if technological progress continues at its current pace. The model also documents that the fall in relative equipment capital prices alone can explain 72% of the decline of the labor share for the 1967-2013 period. Finally, repeating the analysis by disaggregating labor into educational groups reveals that the theory based on capital-task interactions improves on the capital-skill complementarity theory in explaining the decline of the labor share. Chapter 2, "Impact of Information Technology on the Labor Share: Evidence from the U.S. Sectoral Data," contributes to the debate over the relationship between capital deep- ening and the aggregate labor share from a sectoral perspective. The study focuses on a specific group of equipment capital: information and communication technology (ICT) capital and exploits various sectoral heterogeneities to characterize the conditions under which the surge in ICT capital cause the labor share to fall. First, I document significant capital- task complementarity at each sector. Second, decline in ICT capital prices turns out to have a positive impact on the labor share. However, gains of labor devoted to non-routine task occupations are offset by the losses of labor employed in routine task occupations when a sector has: (i) initially high load of employment in routine task occupations, and (ii) weak absolute complementarity between ICT capital and labor working in occupations associated with non-routine tasks. Since sectors satisfying these two conditions have compromised the majority of the economy, the aggregate labor share has exhibited a downward trend so far, leading to the illusion that information technology has been driving the labor share downwards. However, there are two promising facts concerning the future: in one hand, the share of these sectors in value added is persistently falling and on the other hand, the share of routine task employment continues to fall at every sector. Thus, once the structural shifts and within sectoral adjustments are completed, the decline in the labor share should revert back. Chapter 3, "Job Polarization, Skill Accumulation and Wealth Inequality," is one of the first attempts in literature to incorporate the job polarization idea into an otherwise standard incomplete markets model with heterogeneous agents in two dimensions: skills and idiosyn- cratic productivity shocks. This set up allows us to contribute to the existing literature in two ways: (i) linking job polarization with rising wealth concentration, and (ii) modeling the continuous rise in skill supply in response to technological progress and job polarization accompanying it over time. When calibrated and solved for the years 1981 and 2011, the model shows that the decline in relative computer (ICT) capital prices alone accounts for a significant portion of the increases in employment share and relative wages of high skill occupations, as well as the increase in the supply of labor with a college or above-college degree. Consistent with the routinization hypothesis, the model also shows that advances in computer technology can account for most of the decline of the employment share and rela- tive wage of middle class over the three decades between 1981 and 2011. Furthermore, the model successfully captures the erosion of middle-class wealth, whereas wealth concentration rises substantially at the right tail and slightly at the left tail of the wealth distribution.

Abstract: This dissertation covers two distinct but related topics in Labor Economics: promotion within firms and inequality in education.The first chapter develops a new model that explains why many firms do not favor their own employees for advancement. In equilibrium, some, but not all firms commit to promote internally. These "promotion" firms attract higher quality employees to entry-level jobs, which in turn makes the firm more likely to get skilled workers in upper level jobs. Non-promotion firms benefit by paying lower wages. This divergence in strategies reflects the scarcity of high-quality workers. The model generates several testable predictions regarding the differences in wages and the return to tenure between promotion and non-promotion firms. I confirm these predictions empirically using a matched employer-employee dataset from the UK.Chapter two analyzes how the way we measure achievement affects estimates of the black-white test gap among young children. Although both economists and psychometricians typically treat test scores as interval scales, they are reported using ordinal scales. We use the Early Childhood Longitudinal Study and the Children of the National Longitudinal Survey to examine the effect of order-preserving scale transformations on the evolution of the black-white reading test score gap from kindergarten entry through third grade. Plausible transformations reverse the growth of the gap in the CNLSY and greatly reduce it in the ECLS-K during the early school years. All growth from entry through first grade and a nontrivial proportion from first to third grade probably reflects scaling decisions.To address the measurement problems demonstrated in chapter two, in chapter three we relate test scores to adult outcomes. Using data from the CNLSY, we perform order-preserving scale transformations on reading and math test scores to maximize their ability to predict completed education. We find that the black-white achievement gap grows during the early years of education when measured in terms of test scores' economic value. Classical measurement error is insufficient to explain the growth in the gap.