Essays on housing wealth, credit constraints and intertemporal consumption

Abstract

This dissertation considers how housing wealth affects household behavior. Three essays focus on the importance of credit constraints and their role in explaining the spillovers from housing markets on the real economy. In particular, they examine the extent to which house values affect non-housing consumption demand by serving as collateral for households to borrow against to smooth their spending.

Chapter 2 of the dissertation, entitled ``Cross-state Variation in the Response of Consumption to Aggregate Housing Market Shocks in the US: The Role of Collateralized Constraints,' estimates the effects of aggregate US housing market shocks on state-wide consumer spending and home prices. Most studies rely on aggregate data or impose inappropriate restrictions on the behavior of state-level variables. I estimate a latent factor vector autoregression (FAVAR) model and identify housing demand and supply shocks using a signs-restriction bayesian approach. I find that the effects of aggregate housing market shocks are heterogeneous across states and that modifying the assumption about the driving forces behind the run-up in real house prices largely affect the results. The heterogeneity of the effects is partly explained by different levels of mortgage market development across states, which is consistent with an important role for collateral or liquidity constraints as an explanation for the observed correlation between house prices and non-housing expenditures.

Chapter 3, entitled ``Home Equity Borrowing and Households' Spending: Case Study of a Credit Reform in Texas,' investigates whether household spending is affected by an increase in the availability of credit provided by a 1997 Texas legislative change that relaxed severe constraints on the ability of homeowners to use home equity as collateral for consumption loans. Using difference-in-differences methods, I find that the legislative change had sizeable and significant effects on overall spending in Texas, and that the magnitude of the effects is correlated with both the level of income and the amount of equity released by the amendment. The results are consistent with claims that credit constraints are important.

Chapter 4, entitled ``Credit Constraints and Intertemporal Consumption: Theory and Evidence from a Credit Reform in Texas,' investigates the extent to which a general equilibrium model with financial frictions can explain the aggregate spending effects of the 1997 Texas credit reform. I develop a simple two-agent dynamic general equilibrium model in which home values affect the debt capacity and consumption possibilities of a subset of households. I derive an aggregate consumption Euler equation and estimate its structural parameters. I find that the fraction of total consumption accruing to constrained households has decreased after the passage of the reform, and that this smaller fraction is more than offset by an increase in the loan-to-value ratio. The estimates suggest that the credit reform has induced a strong feedback from collateral values to consumption dynamics through the effect that it has generated on borrowing.