Essays on Consumption
Date
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
This dissertation contains two essays that use structural models to study economic agents' consumption and savings behavior. In the first essay, I study how student debt affects consumption and wealth accumulation over the life cycle through its interactions with homeownership and constraints in the mortgage market. Using the Survey of Consumer Finances, I document a negative correlation between student debt and the probability of homeownership over the life cycle, which is more negative for college dropouts than for graduates. To analyze the implications of this relationship over the life cycle, I build and calibrate a model with graduates and dropouts, student debt, and endogenous housing tenure and mortgage refinancing. Consumption is smoothed by adjusting liquid savings and, for homeowners, home equity. Payments on student debt tighten constraints in the mortgage market, which causes indebted consumers to delay homeownership, preventing them from building home equity. Implementing an income-based payment plan, as opposed to a fixed-payment plan, on student debt increases welfare by 1.24% and -0.75% for dropouts and graduates, respectively. In the second essay, which is joint work with Steve Craig, Annie Hsu, Bent Sorensen, Vasundhara Tanwar, and Priyam Verma, we fit a consumer model to U.S. state governments. U.S. state governments are subject to balanced budget rules, but most have “rainy day funds,” which are dedicated savings for use in economic downturns. However, the balances are small relative to fluctuations in revenue. Using data on state finances from the Census Bureau, we find that state governments maintain other cash accounts, which are much larger than the rainy day funds, and absorb shocks through pro-cyclical balances. We ask if state governments save rationally in the sense that a buffer-stock consumer model can describe the spending and savings of a representative state. We estimate the parameters using Indirect Inference and find that state governments are risk-averse and impatient, with the estimates being similar to those found for consumers.