RESEARCH

Abstract: We characterize and quantify optimal urban transportation policies in the presence of congestion and environmental externalities. We formulate a framework in which a municipal government chooses among transportation equilibria through its choice of public transit policies---prices and frequencies---as well as road pricing. The government faces a budget constraint that introduces monopoly-like distortions and the potential need to cross-subsidize modes. We apply this framework to Chicago, for which we construct a new dataset that comprehensively captures transportation choices. We find that road pricing alone leads to large welfare gains by reducing externalities, but at the expense of travelers, whose surplus falls even if road pricing revenues are fully rebated. The optimal public transit price is near zero, with reduced bus and increased train frequencies. Combining transit policies with road pricing allows for higher transit frequencies and lower prices due to a slack budget constraint, increasing consumer surplus after rebates.

Abstract: This paper studies one of the largest spatially targeted redevelopment efforts implemented in the United States: public housing demolitions sponsored by the HOPE VI program. Focusing on Chicago, we study welfare and racial disparities in the impacts of demolitions using a structural model that features a rich set of equilibrium responses. Our results indicate that demolitions had notably heterogeneous effects where welfare decreased for minority households, especially those who were displaced from public housing, and increased for higher-income White households. Counterfactual simulations explore how housing policy mitigates negative effects of demolitions and suggest that increased public housing site redevelopment is the most effective policy for reducing racial inequality.

Abstract: This paper shows the endogeneity of amenities plays a crucial role in determining the welfare distribution of a city's residents. We quantify this mechanism by building a dynamic model of residential choice with heterogeneous households, where consumption amenities are the equilibrium outcome of a market for non-tradables. We estimate our model using Dutch microdata and leveraging variation in Amsterdam's spatial distribution of tourists as a demand shifter, finding significant heterogeneity in residents' preferences over amenities and in the supply responses of amenities to changes in demand composition. This two-way heterogeneity dictates the degree of horizontal differentiation across neighborhoods, residential sorting, and inequality. Finally, we show the distributional effects of mass tourism depend on this heterogeneity: following rent increases due to growing tourist demand for housing, younger residents---whose amenity preferences are closest to tourists---are compensated by amenities tilting in their favor, while the losses of older residents are amplified.

Abstract: Nested logit models represent consumers as agents that choose sequentially over product groups before choosing a final product, hence allowing for flexible substitution patterns across products. However, assuming knowledge of the nest structure has proven problematic in some applications. We propose a method that estimates both the nest structure as well as the structural parameters using product share data. We consider two different settings with price endogeneity: (1) longitudinal observations of products across a large number of markets, where conditional on a product fixed effect prices are exogenous and (2) single-market observations with a cost-shifter. In each setting, we develop estimators to recover the structure of the nest and the parameters and analyze its statistical properties. We propose two-step estimation strategies where in the first step we classify products and in the second step, we recover structural parameters. More specifically, in (1) we use the Bonhome and Manresa (2015) estimator to recover groups, and in the second step, we estimate the model conditional on the estimated nest structure. In (2) we make use of a control function approach to classify products using k-means clustering. We showcase the good performance of our method through a Monte Carlo experiment, and we apply it to the U.S. automobile market data first used in Berry, Levinsohn, and Pakes (1995).

Abstract: How much of the observed across neighborhood segregation is due restrictions in the neighborhood choice sets, such as restrictions imposed by discrimination, versus household preferences? To answer this question, we build and estimate a two-sided housing matching model that endogeneizes the differences in choice sets across demographic groups using a novel dataset that links households and developments to the historic street grid of 1940 Minneapolis metro area. Using an instrumental variable approach, we first document that having \textit{de jure} discriminatory restrictions in the neighborhood reduces the share of non-White and White Eastern European and Middle-Eastern immigrant households by 10.5 and 8.8 percentage points, respectively. Next we show, both theoretically and with Monte Carlo simulations, that traditional discrete choice models of neighborhood demand with mis-specified choice sets produce biased estimates of preferences for neighborhood characteristics. Next, we estimate our two-sided housing matching model that allows for the joint estimation of choice sets and show how mis-specifying those produce significantly biased estimates of preference parameters. Our results suggest that different demographic groups faced substantial different choice sets: On average, minority households are 26 percentage points less likely to have a neighborhood in their choice set relative to non-minority households even after controlling for de jure forms of discrimination. Moreover, ignoring these differences in choice sets over-estimates the role that de jure discriminatory tools play in shaping the location choices of minority households, and thus, segregation.