The dissertation examines the microfoundations of household saving and consumption decisions and some resulting implications for the macroeconomy. The use of micro panel data reveals that heterogeneity, social beliefs and demographic structure shape household financial behavior.
Chapter I examines the wealth accumulation patterns of U.S. households between 1984 and 1994. Wealth is systematically related to income, point in the life cycle and other demographic characteristics. But apart from those differences, there remains great residual heterogeneity in wealth holdings. Persistent differences in savings and wealth holdings exist across the pre-retired, different races and households in different locations, even after controlling for income and other demographics. Additionally, household wealth accumulation remained constant during the last fifteen years despite the well documented decline in personal savings.
Chapter II focuses on individual heterogeneity in households' decisions to file for bankruptcy. Both the effects of bankruptcy stigma and financial benefit are modeled and empirically estimated within households' bankruptcy decisions. The financial benefit to bankruptcy is highly significant in the households' decision to file. Additionally, households who lived in a bankruptcy district with higher than normal bankruptcies during the last year (the proxy for stigma) were found to be more likely to file for bankruptcy this year, all else equal. This effect is also sizeable and helps to explain the accelerated rise in bankruptcies over the past half-decade.
Chapter III explicitly tests for heterogeneity in discount rates and consumption rules in the population. The standard Permanent Income Hypothesis can be rejected for a group of refinancers who paid a rate premium so as to access home equity. These households, while similar to other refinancers in income and demographics, had their consumption paths respond both to predictable income increases and declines. This result is inconsistent with a theory of exogenous liquidity constraints, but is consistent with rule of thumb or with buffer stock consumption behavior.
My Research