Date of Degree


Document Type


Degree Name





Thom Thurston

Subject Categories



distributional decomposition; firm size; household debt; income inequality; VECM; wealth inequality


This dissertation consists of three essays on income and wealth inequality. The essays examine various aspects of this complex feature of the economic system.

The first essay shows that the distribution of firm sizes in an economy is an important determinant of wage distribution. I use data from the U.S. Current Population Survey and the ExecuComp between 1992 and 2012 to construct a new dataset and estimate wage distribution and various measures of wage inequality. I decompose differences in wage inequality across firm sizes and over time by using semi-parametric methods. In 1992, wages were distributed more unequally in small than in large firms. A decomposition shows that this was solely due to inequality among workers with the same observed characteristics, i.e. residual inequality. Inequality due to the distribution of observed characteristics and returns to those characteristics was higher in large firms at that time. By 2012, inequality in small firms grew further, but not as fast as in large firms. Over the same period, employment share of large firms increased but this had little effect on changes in overall wage distribution.

The second essay proposes a general equilibrium model suitable for studying wealth distribution. One of the challenges in modeling wealth distribution is reproducing the high concentration of wealth observed in the U.S. I build upon the benchmark model developed by Krusell and Smith (1998) by introducing firm heterogeneity and managerial class. I develop a model, propose a numerical method for solving it, simulate it, and compare the results with the data and the benchmark model. The simulated distribution fits the data well in the upper tail of the distribution.

The third essay explores the role that changes in income inequality may play in households balance sheets. This is particularly important considering the high levels of household debt that was a key contributing factor in the financial crisis of 2008. I use a family of error-correction models to estimate long-run relationship between income inequality and household debt. Tests based on Westerlund (2007), Pedroni (1999, 2004) and Johansen (1988) are used on a panel of 14 developed countries. No evidence of cointegration between the time series for income inequality and household debt is found.

Included in

Economics Commons