« Back to Results

Structural Analysis of Transportation Policies and Urban Structure

Paper Session

Sunday, Jan. 9, 2022 3:45 PM - 5:45 PM (EST)

Hosted By: Econometric Society
  • Chairs:
    Fernando Ferreira, University of Pennsylvania
  • Panle Jia Barwick, Cornell University and NBER

Efficiency and Equity Impacts of Urban Transportation Policies with Equilibrium Sorting

Panle Jia Barwick
,
Cornell University and NBER
Shanjun Li
,
Cornell University and NBER
Andrew Waxman
,
University of Texas-Austin
Jing Wu
,
Tsinghua University
Tianli Xia
,
Cornell University

Abstract

We estimate an equilibrium model of residential sorting with endogenous traffic congestion to evaluate the efficiency and equity impacts of urban transportation policies. Leveraging fine-scale data on household travel and housing transactions with home and job locations in Beijing, we jointly estimate residential location and travel mode choices and recover households' preference for the ease of work commute and other attributes of residential locations. Counterfactual simulations show that while different policies can attain the same level of congestion reduction, their impacts on residential sorting and social welfare are drastically different. First, a driving restriction intensifies income-stratified urban structure where high-income households live closer to subway and jobs. Distance-based congestion pricing reduces the spatial separation between residence and workplace across income levels, while subway expansion does the opposite. Second, residential sorting strengthens the effectiveness of congestion pricing in improving traffic conditions but undermines that of the driving restriction and subway expansion. Third, the driving restriction is welfare reducing as it leads to large distortions on travel choices. Congestion pricing improves welfare but is regressive, highlighting the need to recycle revenue to address the associated equity concern. Finally, congestion pricing and subway expansion deliver the largest congestion relief and efficiency gain when combined. This policy mix is self-financing in that revenue from congestion pricing could fully cover the cost of subway expansion.

Estimating the Economic Value of Zoning Reform

Fernando Ferreira
,
University of Pennsylvania

Abstract

We develop a framework to estimate the economic value of housing regulations, and apply the framework to a zoning reform in the city of S˜ao Paulo, which substantially altered maximum permitted constructed area to land area ratios at the block level. Using a spatial regression discontinuity design, we find that developers swiftly reacted to increases in allowable building densities by filing for more multi-family construction permits. We incorporate these microestimates of developer responses to zoning reforms in to an equilibrium model of housing
supply and demand to estimate the long term impact of zoning changes on construction, house
prices, residential location decisions and resident welfare. The model predicts the reform will
produce a 0.6 percent increase in the total housing stock, leading to rather small reductions in prices, and consequently small gains in welfare. Neighborhoods with the largest increases in
permitted construction to land areas receive more construction (5%) and lower prices (2.4%).
College-educated and higher income households gain the most from the reform, because more
of those families can now move from the suburbs to the more central parts of the city. Finally,
welfare gains from price changes increase five fold once we account for changes in the built
environment, as density increases and the age of housing units go down. Counterfactual simulations of more aggressive zoning reforms - e.g. doubling allowed densities - produce much
larger gains in welfare.

The Welfare Costs of Urban Traffic Regulations

Isis Durrmeyer
,
Toulouse School of Economics
Nicolas Martinez Franco
,
Toulouse School of Economics

Abstract

We compare the short term impacts of alternative transportation policies to reduce road traffic. More specifically, we measure the welfare costs associated with using alternate day travel schemes, a regulation that has the advantage of being easy to implement and has been often used recently. We compare them with the welfare costs of reducing traffic using standard price instruments. To do so, we develop a structural model of transportation mode choice with endogenous road congestion level. The model considers car trip durations are endogenous equilibrium outcomes from transportation mode choices and road congestion technology. We estimate the model using rich survey data on individual transportation decisions combined with data on expected trip durations from Google Maps and TomTom and high frequency traffic data from road sensors for Paris area. Our results suggest that all the policies are costly for individuals: the benefits of relaxing road congestion does not offset the costs of substituting away from cars. The tolls can be a globally welfare improving regulation if the tax revenue is redistributed.

Refugee Influx and City Structure: Evidence from Mobile Phone Data in Jordan

Konhee Chang
,
University of California-Berkeley
Michael Gechter
,
Pennsylvania State University
Nick Tsivanidis
,
University of California-Berkeley
Nathaniel Young
,
European Bank for Reconstruction and Development

Abstract

Forcibly displaced people are an important feature in many developing countries. Today, one out of every 108 people in the world are displaced. Yet we understand very little about the lives of refugees and other displaced persons within their host communities. In this paper, we examine important decisions made by refugees within a host country, such as where to live and work. In terms of high frequency dynamics, what are the patterns of migration (temporary, permanent, circular)? We examine the role of social networks in shaping these decisions. How important are social networks to refugees in deciding on places to live or for finding work? Do these networks evolve over time, and are there features that facilitate or hinder the integration of these networks with the host communities? We investigate several dimensions of equilibrium impacts on cities from a large influx of refugees. What dynamics emerge across neighborhoods in terms of housing prices for both refugees and incumbents? Do long term residents sort into different neighborhoods when refugees arrive, and at what point does this occur? How might employment opportunities evolve?

We examine these questions in the context of Amman, Jordan. Jordan’s population grew by 13.3% due to the influx of Syrian refugees. Most refugees live in urban environments, with 35% residing in Amman. To answer high frequency and spatial granularity questions, we analyze the metadata from all transactions (voice, SMS and data activity; i.e. CDR data) on one of the largest mobile phone operators in Jordan from June 2015 to the present. Our dataset is unique in that we are able to merge a panel survey of 4,000 respondents with these high frequency cellphone metadata. We validate the use of CDR data to determine home and work locations based on survey responses. We then classify these locations out of sample for the remainder of the network. Applying machine learning methods, we classify customers as long-term residents or refugees/migrants. These results allow us to look deeply into migration dynamics and interactions between host and refugee communities. Social networks can be traced across time using voice and SMS transactions. Neighborhood characteristics and long-term changes in outcomes, such as housing prices, employment and shares of refugees are calculated from administrative data from before and after the refugee influx. We will fit these various data to a structural model in order to measure the indirect impacts of the refugee influx on the host community and better understand the equilibrium implications for all residents of the city.
JEL Classifications
  • R2 - Household Analysis
  • R4 - Transportation Economics