Research
Research
Work from Home and Spatial Misallocation
I study whether the recent shift towards work from home (WFH) in the US alleviates or exacerbates the impact of existing policy distortions—progressive taxation and housing supply restrictions—on the aggregate level of output and inequality. These policies divert labour towards less productive locations, hurting aggregate productivity. In principle, WFH can alleviate this misallocation by allowing workers to supply their labour to high-productivity cities while living in more affordable areas and paying lower income taxes. In turn, firms in high-productivity cities can substitute in-person labour with cheaper remote workers. However, WFH may also exacerbate the misallocation, as remote work remains subject to relatively high tax rates and is complementary to in-person labour. Using a spatial equilibrium model calibrated with US Census data, I find that the shift to WFH does not change the distortionary effect of housing regulations, but it increases the distortionary effect of progressive taxation. In particular, adopting a flat tax scheme yields an additional 0.5pp output increase in the economy with the WFH shift, relative to the economy in 2019. However, this greater productivity gain comes at the cost of additional inequality between college- and non-college-educated workers, compared to the 2019 economy, as lower-skilled workers have less access to remote work. Thus, WFH intensifies the spatial equity-productivity trade-off, underscoring the need for more nuanced tax policy.
I study the relationship between housing price bubbles and spatial labour mobility with a novel spatial equilibrium model, in which agents choose between owning and renting, and which I extend by incorporating the evolution of agents’ beliefs about housing fundamentals. In anticipation of rising housing prices, optimistic agents want to move to booming cities and to own larger homes. Housing supply is constrained by local regulations and geography. I apply the model to examine the reallocation of labour during the 2000-2006 US housing boom, a period when housing prices soared in certain metropolitan areas, such as San Jose and Las Vegas, while in others, including Pittsburgh and Atlanta, housing prices remained notably flat. I calibrate the model using US MSA data and find that, in response to a simulated housing bubble, it predicts a sizeable reallocation of population away from booming MSAs with unaffordable housing and towards MSAs with booming housing prices and affordable housing, as well as MSAs with flat housing prices.
Returns to Capital Across US States and Over Time, with Daniele Coen-Pirani