Working Papers

Are Retail Banking Markets Still Local? Quasi-Experimental Evidence from Bank Mergers (2020)

The rise of nonbank and online lending has created a national market for many retail financial products. What does this mean for bank regulations and theories of competition that assume banks compete locally? This paper studies the effects of changes in local bank competition coming from antitrust rules which discontinuously shift bank mergers’ competitive impact. Antitrust intervention within a region increases the number of local banks, decreasing concentration. Relative to markets without intervention, in markets with intervention 6-month deposit rates rise by 0.11 percentage points and mortgage origination increases by 16%, both consistent with rising competition. Markets without intervention experience a relative shift towards nonbank mortgage lending. These findings suggest that local competition remains important for household financial products but that nonbank lenders can partially substitute for declines in bank competition.

Why is the Rent So Darn High?, with Greg Howard (2020)

Because of black changing location demand. In a spatial equilibrium framework, we show that three-quarters of the CPI rent increase in the United States from 2000 to 2018 is due to increased demand to live in ex ante housing-supply-inelastic cities. Moving one person to a less elastic city raises the average rent because the positive effect on rents in the inelastic city outweighs the negative effect in the elastic one. In these years, the quantitative importance of this location demand channel is greater if people are mobile in response to rent changes. Empirically, we show that people have high long-run mobility by estimating that income changes have similar effects on rents across cities regardless of housing supply elasticity. Supporting this location demand channel, the cross-sectional pattern of location demand implied by our model matches patterns of labor-market and amenity changes.

Regional Divergence and House Prices, with Greg Howard (2020)

This paper develops a model of the U.S. housing market that explains much of the time series of rents and house prices since World War II. House prices depend on expectations of future rents. We show that rents are tied to regional income inequality, and therefore, house prices are determined by how much faster incomes are growing in richer regions. This theory also matches many cross-sectional facts, including regional variation in rents and prices, differing house price sensitivities to national trends, patterns of inter-state migration, and surveys of income expectations. An industry shift-share instrument provides causal evidence for our channel. The model implies that while interest rates have an ambiguous effect on house price levels, low rates increase house price volatility.

When and Why Does Debt Overhang Matter? Evidence from the Retail Apocalypse with Ricardo Correa and Martin Sicilian (2019).

We study the causes and consequences of debt overhang using plausibly exogenous variation in the leverage of retail properties. We use a difference-in-differences strategy that instruments for leverage by exploiting variation in housing prices that occurs after commercial mortgages are originated. A 10% decrease in property prices, which corresponds to a 7% increase in leverage for the median loan, is associated with 0.4% lower average net operating income and 1.4 percentage points lower occupancy. Decomposing the negative effects of leverage, we show that higher occupancy rates come entirely from levered properties’ poor responses to negative shocks. We provide further evidence on the effects of debt overhang by studying landlords’ responses to the unexpected closure of anchor tenants following national chain store closings. Leverage magnifies the negative effects of these shocks substantially.

The Cyclicality of CEO Turnover, with Heidi Packard (2019)

New draft coming soon

CEO turnover is highly pro-cyclical. This paper aims to explain why. We begin by showing that the cyclicality is driven almost entirely by executives of retirement age. We further provide evidence that executives time their retirement to maximize the value of their pensions. Since CEO pay is pro-cyclical and pensions are based on pay in the final years of tenure, executives have the incentive to retire when the economy is doing well. Cyclicality is particular strong in firms with strong corporate governance, which suggests that retirement cyclicality is a tool firms use to constrain CEO behavior.

Works in Progress

Liquidity and Moving Costs with Alex Bartik and Jens Kvaerner

Using comprehensive administrative data from Norway, we study 1) How financial liquidity determines which households move when new opportunities arise in the oil sector 2) How unexpected income shocks coming from oil price movements are invested.

Understanding High-Growth Entrepreneurs with Victor Lyonnet and Maxime Bonelli

Using administrative data from the universe of French firms matched to a large administrative survey of entrepreneurs, we investigate the determinants of high-growth entrepreneurship in France.

Resting Papers

Housing Demand, Regional House Prices and Consumption (2017)

This paper provides a new explanation for regional variation in the 2000-2006 housing and consumption boom. Cities with relative increases in housing demand resulting from differences in industrial composition had greater house price increases from 2000-2006 and greater declines from 2007-2012. Consistent with theory, price effects are stronger in inelastic cities. City-level differences in housing demand are also correlated with housing supply elasticity. Controlling for demand, I estimate a durables consumption-house price elasticity of 0.08 from 2000-2006, 40% smaller than previous estimates. Post-2006, I estimate an elasticity of 0.31 and find that housing prices rather than demand explain consumption changes.

Interpreting Instrumented Difference-in-Differences, econometrics note with Sally Hudson and Peter Hull (2017)

Scaling a difference-in-differences effect on an outcome by a difference-in-differences effect on a mediating treatment variable is a longstanding and increasingly common practice in applied microeconomics. We provide a framework for such instrumented difference-in-differences (DD-IV) estimation when the effect of treatment is heterogeneous across a panel of individuals, and propose intuitive assumptions under which DD-IV with panel data identifies a local average treatment effect.