Working Papers

Labor Market Power Across Cities
Revise and resubmit at Journal of the European Economic Association.

Latest Version Abstract

Workers in larger cities are paid higher wages. The city-size wage premium may reflect the productivity gains from agglomeration or sorting of more productive workers in densely populated areas. However, local labor markets in large cities have more firms and are expected to be more competitive, which could also generate part of the urban earnings premium. I quantify the importance of this channel with rich administrative data for Spain using a spatial equilibrium model to guide the empirical strategy. To address the identification challenge posed by labor market power and wages moving endogenously with unobserved local productivity shocks, I first control for firms' revenues per worker and for time trends that are heterogeneous across local labor markets. I then develop a new instrumental variable that leverages quasi-experimental variation in monopsony power stemming from changes over time in the size of local public firms. I conclude that 20–30% of the city-size wage premium can be attributed to differences in labor market power across locations.

Labor Market Power and Spatial Policies with Sophie Mathes and John Morehouse
New draft!

Latest Version Abstract

Spatial models typically assume competitive labor markets. We develop a spatial equilibrium model with labor market power and calibrate it to four decades of U.S. data. Wage markdowns have roughly halved since 1980, driven by growth in the number of local establishments and rising worker mobility across industries. Relaxing housing supply constraints and subsidizing migration both draw workers from the rest of the country to the targeted cities, so that establishments enter these cities and exit the rest of the country. Because markdowns respond more to establishment entry and exit where they are high, a model without labor market power overstates the welfare gain of these policies when they target large productive cities, where markdowns are already low, and understates it when they target concentrated labor markets. The migration subsidy also makes labor supply more elastic, which narrows markdowns and partly offsets the effect of establishment exit.

Segregation and Sorting of U.S. Households: Who Marries Whom and Where? with Davide Alonzo and Nezih Guner

Latest Version Abstract

We use a spatial heterogeneous agent model of family formation to study the role of geography in shaping trends in marriage and inequality in the US. The United States has undergone dramatic shifts in household and family structures since the 1980s. Married female labor force participation increased significantly, marriage has declined, and positive assortative mating has risen. Along with a rising skill premium and a declining gender gap, income inequality among households has also widened. We document that these changes had an important geographic dimension: the decline in marriage was much more significant in smaller cities, and marital sorting declined in smaller cities but increased in larger ones. We interpret these facts within a model where households decide where to live, whether or not to get married, and with whom. The framework allows us to quantify the importance of each channel in shaping trends in marriage and inequality.

Should Governments Subsidize Homeownership? A Quantitative Analysis of Spatial Housing Policies

Latest Version Abstract

Should governments promote homeownership? Although such policies are widespread, their welfare implications are not straightforward. While subsidies can overcome financial frictions and, by increasing homeownership, insure against rent volatility, they can also reduce internal migration to productive locations. To address this question, I build a dynamic spatial equilibrium model with coresidence, homeownership, internal migration, and savings decisions. Homeownership provides utility and insurance against aggregate rental price risk but reduces migration due to the transaction costs associated with selling the property. Migration decisions, in turn, affect homeownership. In particular, non-migrant workers can coreside with their parents, which allows them to save and buy a house earlier than migrants. I develop a new strategy to solve dynamic spatial models with aggregate uncertainty, which models agents' expectations about local endogenous prices and wages using lower-rank factors. The model is estimated for Spain and validated using quasi-experimental evidence from recent place-based homeownership subsidies. I find that mortgage interest deduction policies reduce wealth inequality and are welfare-increasing, despite reducing migration to productive locations. Targeting high-wage locations leads to lower welfare gains and does not increase homeownership due to higher house prices.