I am an assistant professor and Donald P. Jacobs scholar at the Kellogg School of Management, Northwestern University. My interests are empirical macroeconomics and household finance.


Kellogg School of Management
Finance Department
Northwestern University
2211 Campus Drive
Evanston, IL 60208

Working Papers

with Anthony DeFusco and Stephanie Johnson
This paper studies how credit markets respond to policy constraints on household leverage. 
Using a research design that exploits a sharp policy-induced discontinuity in the cost of orig
inating certain high-leverage mortgages, we study how the Dodd-Frank “Ability-to-Repay” 
rule affected the price and availability of credit in the U.S. mortgage market. Our estimates 
show that the policy had only moderate effects on prices, increasing interest rates on affected 
loans by roughly 10–15 basis points. The effect on quantities, however, was significantly 
larger; we estimate the policy eliminated 15 percent of the affected market completely and 
reduced leverage for another 20 percent of the remaining borrowers. This reduction in quan
tities is much greater than what would be implied by plausible demand elasticities and sug
gests that lenders responded to the policy primarily by rationing credit. Our results show 
that macroprudential policies that restrict household leverage can have large quantity effects, 
even when the market-priced costs of such regulations are relatively small. Finally, while the 
policy succeeded in reducing leverage, our estimates suggest that this effect would have only 
slightly reduced the aggregate default rate during the housing crisis.

(updated 5/25/2017)
How much did the contraction in the supply of credit to households contribute to the decline in employment during the Great Recession? To answer this question I use variation in shocks to mortgage credit supply across U.S. counties. First, I non-parametrically identify lender-specific shocks, which I then aggregate into a county-level shock. I show this shock can help account for the household credit channel, but direct estimates are biased. I use two related sources of exogenous variation to correct this bias. First, I exploit a county's exposure to a large and previously healthy lender as a natural experiment. The instrumental variable estimates confirm that direct estimates are biased, and that shocks to household credit supply had large effects. A one standard deviation decline reduces employment by 3 percent, the flow of home purchase credit by 7 percent, refinance credit by 20 percent, and home improvement credit by 5 percent. These effects are very persistent, providing cross-sectional evidence that recoveries after financial shocks are slow. Second, I identify additional lenders likely to satisfy the exclusion restriction and recover effectively identical estimates. These results imply the direct effects of the household credit channel account for at least 20 to 30 percent of aggregate employment losses. This shows shocks to the supply of credit to households were significant contributors to the severity of the Great Recession and subsequent slow recovery.

Does Greater Inequality Lead to More Household Borrowing? New Evidence from Household Data
(updated 2/20/2016)
We combine household data on debt during 2001-2012, regional variation in inequality, and a novel imputation procedure to show that low-income households in high-inequality regions accumulated less debt relative to income than similar households in lower-inequality regions. These results are robust across subsamples and within debt categories, with the strongest effects in mortgage debt. To determine if these quantity movements are driven by supply or demand we examine measures of credit cost and access with data on individual mortgage applications. We find that as inequality increases, low-income households are more likely to be denied credit, more likely to be charged a high interest rate, have to travel further to lender branches, and are less likely to have a new branch opened in their neighborhood. We argue that these patterns are consistent with local inequality differentially affecting the supply of credit to low- and high-income households. We propose a simple model to rationalize these findings that hinges on lenders using a household’s relative position in the local income distribution to infer the underlying quality of applicants. Higher inequality allows lenders to more easily discern borrower quality and so channel relatively more credit toward higher-income applicants. These results highlight a novel channel through which changes in inequality can have significant and differential economic effects on household leverage.

Work in Progress

The Housing Crisis and the Rise in Student Loans 
(draft coming soon)
We study if the changes in U.S. house prices over the 2000s affected growth in student loans. Using household-level panel survey data, we find that as home prices fall households
depend less on home equity extraction to finance college enrollment and depend more on student loans. We estimate that for every lost dollar of home equity credit that would have been used
to finance college enrollment, households increase student loan debt by forty to sixty cents. This substitution appears to be driven primarily by households with low levels of liquid assets.
We extend our analysis with credit bureau data to trace longer-run effects of this leverage on students. Our results show that the decline in house prices reduced households’ ability to finance
college enrollment with home equity credit, but that constrained households mostly responded by continuing to enroll in college and relying on student loans. Our estimates suggests the 30%
fall in house prices from the 2006 peak resulted in the average college student borrowing an additional $1,300 in student loans, with some evidence of larger effects on liquidity-constrained
and less-educated households.

Default and Deleveraging in the Business Cycle

Sources of Heterogeneity in Retail Price-Setting Behavior 
with Bulat Gafarov, Daniel Greenwald, and Leonid Ogrel