This item is under embargo and not available online per the author's request. For access information, please visit

Date of Award

Spring 5-15-2018

Author's School

Graduate School of Arts and Sciences

Author's Department


Degree Name

Doctor of Philosophy (PhD)

Degree Type



Chapter 1: The labor share of national income in the United States has declined since the 1980s and especially after 2000. My paper focuses on the role played by technological change in this process. In particular, firms that adopt new technologies achieve a low labor share, and grow and take a larger market share over time. An example is online retailers such as Amazon, empowered by information technology, that have a lower labor share than traditional retailers, and have continually expanded over the last 20 years. This realloca- tion process drives down the aggregate labor share. I first document three facts: (i) across sectors, there is a negative correlation between change in concentration and change in the labor share; (ii) large firms usually exhibit a smaller labor share; (iii) in sectors where the labor share declines, the decline is especially strong among large firms. Specifically, gains in labor productivity are not associated with comparable increases in wages. Then I provide a rationale for these facts by assuming that capital and labor are complemen- tary inputs and technological progress is labor saving and embodied in the capital stock. Under these assumptions, my model predicts a negative correlation between firm size and labor share. Further, the adoption of new technologies diminishes the labor shares in large firms and increases their market share. As a consequence, the aggregate labor share declines. This technological channel is consistent with the evolution of labor productivity across sectors during the last 30 years.

Chapter 2: This paper constructs a quantitative history of the homicide rate in Qing China and in- vestigates its social and economic drivers. Estimates based on historical archives indicate that this annual rate ranged between 0.35 and 1.47 per 100,000 inhabitants during the 1661-1898 period, a low level unmatched by Western Europe until the late 19th century. China?s homicide rate rose steadily from 1661 to 1821 but declined gradually thereafter until the turn of the century. Although extreme, homicide represents a random sampling of the entire distribution of interpersonal violence; hence the homicide rate serves as a proxy for overall violence, and its rise implies a decline in personal security. We use national and cross-provincial panel data to show that population density, state capacity, local self-governance, interregional grain market integration, and grain price level (which captures crop failure and other survival distress) are all statistically significant drivers of the homicide rate in 18th- and 19th-century China.

Chapter 3: The measured labor wedge, defined as the difference between marginal product of la- bor and marginal rate of substitution, is relatively stable from 1940s to 70s, and declines secularly from 1980s onwards. This paper aims to investigate the effect of a particular labor market institution, labor union, on labor wedge. Labor unions command a wage premium, which invites job application queues and job rationing in the unionized sector. This waiting values of unionized jobs creates a wedge between wages and households’ willing to work. We provide sectoral evidence that supports a union-wedge connection in the manufacturing sector. A quantitative model which features two labor market, one competitive and the other unionized, is developed to estimate the effect of union power on labor wedge. Based on our quantitative results, approximately 20% of the decline in labor wedge from 1970s to 2000s is accounted for by the decrease in union densities.


English (en)

Chair and Committee

Michele Boldrin

Committee Members

Rodolfo Manuelli, Yongseok Shin, Ping Wang, Yi Wen,


Permanent URL:

Available for download on Wednesday, April 17, 2019

Included in

Economics Commons