The Foundation
Two effects, pulling in opposite directions
The 2008 crisis flooded banks and courts with foreclosure filings, and the congestion stretched the time between a missed payment and a completed foreclosure to years in some states. My job market paper asked what those delays do to a household's decision to stop paying, and found the answer isn't one effect but two, in tension.
Delay creates moral hazard: households realize they can skip payments with limited near-term punishment, which pushes default up. But delay is also insurance: a household with few liquid assets facing a long unemployment spell can survive delinquency and keep the home where, without delay, it couldn't. That effect pushes default down today. The model works out when each force dominates, and evidence from the Survey of Consumer Finances supports the mechanism.
The Tension
Same delay, opposite predictions
Effect One
Moral hazard
Skipping payments carries limited punishment while the foreclosure queue is long. Delay makes default cheaper, so more households do it, including some who could pay.
Effect Two
Insurance
For a liquidity-constrained household in an unemployment spell, delay is a bridge: a real chance to cure the delinquency and keep the home. Anticipating that, fewer households give up today.
The paper: Foreclosure Delay and Default Incentives (2016), job market paper, from the dissertation Institutions, Markets, and Household Behavior (Ph.D., UC Santa Barbara, 2017).
The Second Experiment
The moratoria ran the model at national scale
In 2020, foreclosure delay stopped being a byproduct of congestion and became deliberate policy. Federal and state moratoria and mass forbearance made delay universal, announced, and time-limited, a fundamentally different variation than the court-congestion delays of 2008, and a cleaner one.
That makes the moratoria a second test of the model's two effects. If moral hazard dominates, mass non-payment should have spread well beyond distressed households while the protection was in place. If insurance dominates, the delay should show up as households curing delinquencies and keeping homes once the protection lifted, without the foreclosure wave that congestion-era reasoning would predict. Which pattern the data shows, and for whom, is the empirical question this project is being designed to answer.
Want to talk about this work?
Questions about the model, the moratoria evidence, or where the design is headed are welcome.