Essays on Structural Transformation and Inequality

Chapter 1: In this chapter, I replace the Cobb–Douglas production function in Acemoglu and Guerrieri (2008) with a two-sector model using the CES production technology to analyze how the U.S. capital deepening process must cause unbalanced growth. The U.S. economy is classified into labor- and capital-intensive sectors. My model shows that due to the difference in the two sectors\u27 elasticities of substitution between capital and labor, capital deepening leads to a growing concentration of labor in the labor-intensive sector, but a continuous reallocation of capital away from this sector. This causes sector outputs to increase at different rates. Combined with exogenous technological progress, this unbalanced growth affects the path of factor costs and output prices. Furthermore, the model shows that even as the real interest rate remains relatively constant and wage rate rises, the labor income share in GDP diminishes over time, and the capital income share increases. The ease with which labor can be replaced with capital in the capital-intensive sector and consumers\u27 preferences for how to combine the two sectors\u27 outputs in their final consumption are key parameters influencing the rate of decline of the labor income share. Chapter 2: Land is a crucial production factor in the agricultural sector. An important recent literature argues that a large part of the very large measured cross-country differences in productivity could be accounted for by land misallocation within developing countries and, in particular, that many countries have too few large farms (e.g. Adamopoulos and Restuccia, 2014). These studies measure misallocation against a USA benchmark established by the assumption that the unobserved distribution of farming skills in the population can be inferred from the observed size distribution of farms in the USA, measured at a national level. This paper uses the US Census of Agriculture data to decompose the US farm size distribution by region and land type (e.g. cropland and pastureland). I find important differences in the distribution of farm size across regions and land types, suggesting that any inference about the distribution of unobserved skills should take these factors into account, and that misallocation studies are producing biased results. I also explore whether the top tail of the farm size follows the power law distribution. Chapter 3: This paper studies the relationship between the expansions of manufacturing and service sectors and wage inequalities among employees in US metropolitan areas. Sector expansion is defined as an increase of the number of employees working in the sector. I use the American Community Surveys (ACS) data to conduct my analyses. The metrics that I use to measure wage inequality are wage variance and the 90-50 percentile and 50-10 percentile gaps. In my analyses, I keep the distribution of observed skills across sector size and the price and quantity of skills unchanged since 1989. The basic assumption is that an individuals\u27 wage is the sum of his/her skill group\u27s mean wage and some random residual, where the mean differs across skill groups and locations. I investigate the variance and the 90-50 percentile and 50-10 percentile gaps of mean wage and residual. My finding is that sector expansion is associated with rising wage inequality. However, in service sector, wage inequality is more sensitive to observed skills, and in manufacturing sector wage inequality is more sensitive to unobserved skills