Recent Published Papers

​Previous work on agglomeration economies ignores the built environment. This paper shows that the built environment matters, especially for commercial sectors that dominate city centers. Buildings are specialized beyond random assignment, in part because externality-generating anchor tenants skew a building's other tenants towards the anchor's industry. An anchor elsewhere on the blockface has a much weaker effect, and one that is weaker still if across the street, suggesting rapidly attenuating agglomeration economies. Attenuation is pronounced for retail and information-oriented office industries but is absent for manufacturing. Building managers have incentives and capacities to partly internalize local externalities, contributing to urban productivity.

The causal effect of city size on urban wage premia has been difficult to measure because unusually skilled workers may select into large city labor markets. We propose a new approach to this challenge. For single-peaked wage distributions, if individuals left of the mode disproportionately select out of large city labor markets, the CDF evaluated at the mode shrinks as city size increases. Among college trained, white full-time US workers, evidence of selection is present even after conditioning on extensive observable attributes. Among individuals with a high school degree or less, selection is absent. Additional estimates indicate that for college trained workers, 3.5% is an upper bound on the modal worker's wage elasticity with respect to city size. For those with limited education we can be more precise: modal wage elasticity is 3.9% for men and 5.2% for married women.

How close do firms and workers need to be to benefit from proximity to each other? This question is is implicit throughout the literature on agglomeration economies and is the focus of this paper. More colloquially, how close is close?

​Tall commercial buildings dominate city skylines. Nevertheless, despite decades of research on commercial real estate and horizontal patterns of urban development, vertical patterns have been largely ignored. We document that high productivity companies locate higher up, with less productive offices lower down and retail at ground level. These patterns reflect tradeoffs between street access and vertical amenities. Vertical rent gradients are non-monotonic, independent of nearby employment, and large. Doubling zipcode employment is associated with a 10.7% increase in rent, consistent with the presence of agglomeration economies. Moving up one floor has the same effect on rent as adding roughly 3,500 workers to a zipcode.

Although there is broad recognition that cities differ in their tendency to experience house price bubbles, most studies assume away any possibility of within-city heterogeneity in response to a bubble. We develop a model that suggests that this assumption may be appropriate when markets are rising but can be far from reality on the bust side of a bubble. During a housing boom, new construction and related supply adjustments by developers ensure stable relative prices between low- and high-quality segments of the housing market. On the bust side of a bubble, however, reduced housing starts allow demand-side forces and mortgage default to create pressure for relative prices to diverge across market segments. Absent a change in technology, as markets recover and new construction rebounds, relative prices should revert back to pre-crash levels. Evidence based on 2000–2013 single-family home sales in Phoenix, Arizona supports this modeling framework. Additional evidence also suggests that high rates of mortgage default contributed to divergence in relative prices when markets crashed.

​Previous studies have struggled to demonstrate that higher taxes deter business activity. We revisit this issue by estimating the effect of changes over time in cross-border differences in state tax conditions on the tendency for new establishments to favor one side of a state border over the other. Identification is enhanced by taking account of previously overlooked reciprocal agreements that require workers to pay income tax to their state of residence as opposed to their state of employment. When reciprocal agreements are in force, higher personal income tax rates lure companies from across the border, while corporate income tax and sales tax rates have the opposite effect. Where reciprocal agreements are not in place, the results are largely reversed. These patterns are amplified in heavily developed locations, and differ in anticipated ways by industry and corporate/non-corporate status of the establishment. Overall, results strengthen the view that state-level tax policies do affect the location decisions of entrepreneurs and new business activity, but not in a way that lends itself to a one-size-fits-all summary.

​This paper considers the empirical literature on the nature and sources of urban increasing returns, also known as agglomeration economies. An important aspect of these externalities that has not been previously emphasized is that the effects of agglomeration extend over at least three different dimensions. These are the industrial, geographic, and temporal scope of economic agglomeration economies. In each case, the literature suggests that agglomeration economies attenuate with distance. Recently, the literature has also begun to provide evidence on the microfoundations of external economies of scale. The best known of these sources are those attributed to Marshall (1920): labor market pooling, input sharing, and knowledge spillovers. Evidence to date supports the presence of all three of these forces. In addition, there is also evidence that natural advantage, home market effects, consumption opportunities, and rent-seeking all contribute to agglomeration.

​This paper estimates the value firms place on access to city centers and how this has changed with COVID-19. Pre-COVID, across 89 U.S. urban areas, commercial rent on newly executed long-term leases declines 2.3% per mile from the city center and increases 8.4% with a doubling of zipcode employment density. These relationships are stronger for large, dense “transit cities” that rely heavily on subway and light rail. Post-COVID, the commercial rent gradient falls by roughly 15% in transit cities, and the premium for proximity to transit stops also falls. We do not see a corresponding decline in the commercial rent gradient in more car-oriented cities, but for all cities the rent premium associated with employment density declines sharply following the COVID-19 shock.

Although well-known that high current loan-to-value ratio (CLTV) is necessary for mortgage default, the amplifying effect of high payment-to-income ratios (PTI) that can force families to move has received limited attention. Using the 1985–2013 American Housing Survey panel, we show that high CLTV by itself has little effect on mobility, but high PTI prompts families to move and especially so when CLTV is high. Evidence also indicates that high PTI and CLTV discourage home maintenance. Our estimates suggest that loan modifications that lower PTI will likely be more effective at helping underwater families to remain in their homes and avoid mortgage default as compared to policies that lower CLTV.

This paper examines vertical patterns of employment density and agglomeration economies within tall buildings. Theory suggests that vertical density should depend on the interplay of street access, height-related amenities, and productivity. Based on suite level data, we show that density patterns are u-shaped, with high density at ground level and high floors. Furthermore, factors associated with productivity, including nearby employment and firm-specific characteristics, have positive effects on employment density. Vertical density patterns are consistent with productivity spillovers that are strongest on a company’s floor and attenuate rapidly with vertical distance. Similar evidence is obtained based on sales for law firms.

This paper measures the impact of individual-level housing capital gains on transitions into and out of self-employment. Drawing on special features of the 1985–2013 American Housing Survey (AHS) panel, our most robust models control for recent expenditures on home maintenance, MSA-by-year fixed effects, lagged proxies for wealth and other household attributes. Net of home maintenance, a 20% real increase in home value over a two-year period raises the likelihood of entry into self-employment by roughly 1.5 percentage points; housing capital losses have little effect on exits. Controlling for house fixed effects, self-employed homeowners are also more likely to hold a HELOC, facilitating easy, low-cost access to home equity that could be used to cover business expenses. These and other estimates suggest that links between homeownership and self-employment are strong enough to be important when home prices are rising rapidly, but modest when housing capital gains are limited or negative.

​This chapter reviews recent literature that considers and explains the tendency for neighborhood and city-level economic status to rise and fall. A central message is that although many locations exhibit extreme persistence in economic status, change in economic status as measured by various indicators of per capita income is common. At the neighborhood level, we begin with a set of stylized facts and then follow with discussion of static and dynamic drivers of neighborhood economic status. This is mirrored at the metropolitan level. Durable but slowly decaying housing, transportation infrastructure, and self-reinforcing spillovers all influence local income dynamics, as do enduring natural advantages, amenities, and government policy. Three recurring themes run throughout the paper: (i) Long sweeps of time are typically necessary to appreciate that change in economic status is common, (ii) history matters, and (iii) a combination of static and dynamic forces ensures that income dynamics can and do differ dramatically across locations but in ways that can be understood.

While filtering has long been considered the primary mechanism by which markets supply low-income housing, direct estimates of that process have been absent. This has contributed to doubts about the viability of markets and to misplaced policy. I fill this gap by estimating a "repeat income" model using 1985-2011 panel data. Real annual filtering rates are faster for rental housing (2.5 percent) than owner-occupied (0.5 percent), vary inversely with the income elasticity of demand and house price inflation, and are sensitive to tenure transitions as homes age. For most locations, filtering is robust which lends support for housing voucher programs.