Abstract
In this dissertation, we study the interaction of consumer bankruptcy and labor-market policies such as Unemployment Insurance (UI) and minimum wages (MW). Chapters 1 and 2 focus on UI. Chapter 1 uses cross-states differences in the maximum amount of UI available and county-level bankruptcy rates to estimate how UI affects consumer bankruptcy. In Chapter 2, I quantitatively evaluate how UI affects unsecured credit markets and how the welfare implications of UI depend on consumer bankruptcy. In Chapter 3, we use cross-state differences in the minimum wage (MW) and county-level consumer bankruptcy rates to estimate the effect of MW on consumer bankruptcy.
In chapter 1, I use cross-states differences in unemployment insurance (UI) and county-level consumer bankruptcy rates from 1991-2007 to estimate the effect of UI on consumer bankruptcy by exploiting policy discontinuities at the state borders. I find that Chapter 7 bankruptcy rates are significantly lower in counties belonging to states with higher UI compared to its neighboring county in the lower UI state. A 10% increase in the maximum amount of UI available decreases the bankruptcy rate by around 1.9%. The effect of UI on bankruptcy, as studied in chapter 2, is ambiguous. The result in this chapter is informative about the direction of this effect for the US data and will serve as a testable implication of the model developed in the next chapter.
In chapter 2, I quantitatively evaluate how UI affects unsecured credit markets and how the welfare implications of UI depend on consumer bankruptcy. Theoretically, higher UI benefits can reduce default risk since they imply higher income during a situation of low-income. However, they can also reduce precautionary savings, encourage borrowing and unemployment, and require more taxes, which would increase default risk. I construct a general equilibrium model of unsecured consumer credit and unemployment. The model accounts for the cross-state negative relationship between bankruptcy rates and the maximum amount of UI available. I use the model to study changes in the UI replacement rate. For low levels of replacement rate, the model predicts that the first effect dominates, and more UI benefits reduce default risk and increase ex ante welfare. As UI increases, default risk increases, and welfare falls. Bankruptcy is a barrier for the UI to increase welfare. If bankruptcy is not available, increasing the replacement rate above the current 50% to 60% would increase welfare by 1.3% in terms of lifetime consumption; with a bankruptcy option, it reduces welfare by 3.6%.
In chapter 3, we use cross-state differences in MW and county-level consumer bankruptcy rates from 1991-2017 to estimate the effect of MW on consumer bankruptcy by exploiting policy discontinuities at the state borders. We find that Chapter 7 bankruptcy rates are significantly lower in counties belonging to states with higher MW compared to its neighboring county in the lower MW state (a 10% increase in MW decreases bankruptcy rate by around 4.4%). However, for Chapter 13, we find no statistically significant relationship. Also, the data suggest that before the 2005 Bankruptcy Reform, the effect of MW on reducing bankruptcy was almost as twice as large than for the overall period.
