2003年-世界发展银行全球_The_Impact_of_Urban_Spatial_Structure_on_Travel_Demand_in_the_United_States_60页_2mb

Summary of "The Impact of Urban Spatial Structure on Travel Demand in the United States"

Core Content

This paper investigates how urban spatial structure and public transit supply influence travel demand and commute mode choices in the United States. The study uses data from the 1990 Nationwide Personal Transportation Survey (NPTS) combined with urban form and transit supply measures for 114 urbanized areas. The objective is to understand the relationship between urban form and travel behavior, particularly focusing on the impact of public transportation on reducing vehicle dependence.

Main Questions

1. How do measures of urban form—such as city shape, road density, population centrality, and jobs-housing balance—affect annual vehicle miles traveled (VMT) and commute mode choices?
2. How does the supply of public transportation—measured by route miles and proximity to transit stops—affect VMT and commute mode choices?

Key Findings

• Urban Form Impact: Measures of urban form, including population centrality, city shape, and road density, have a statistically significant but small effect on VMT and mode choice. A 10% increase in population centrality reduces the probability of driving by 2.1 percentage points, while a 10% increase in distance to the nearest transit stop raises the probability of driving by 1.6 percentage points.
• Public Transit Impact: The supply of public transportation, particularly rail, has a smaller effect on VMT compared to urban form. However, the presence of rail transit significantly influences mode choice.
• Combined Effects: Altering urban form and increasing public transit supply simultaneously can have a more substantial impact. For example, moving households from a city like Atlanta to one like Boston reduces annual VMT by 25%.
• Elasticity of VMT: The elasticity of VMT with respect to urban form variables is small, ranging from 0.10 to 0.20 in absolute value.
• Endogeneity Addressed: The study uses an instrumental variable approach to account for the endogeneity of proximity to public transit, using average transit usage in potential residential areas as an instrument.

Measures of Urban Form

• City Shape: Measured by the ratio of the minor to major axis of an ellipse that encloses the urbanized area. A ratio of 1 indicates a circular city.
• Road Density: Calculated as the total miles of road (divided by road width) per square kilometer of urbanized area.
• Population Centrality: A measure derived by comparing the distribution of population within the city to a uniform distribution. It is calculated by plotting the cumulative percentage of population against the cumulative percentage of distance from the CBD and measuring the area between the curve and a 45-degree line.
• Jobs-Housing Balance: Measured using a Gini coefficient-like index, which assesses the evenness of job distribution relative to housing. A higher value indicates less even distribution.

Measures of Transit Supply

• Transit Route Miles: Calculated as the total bus and rail route miles per square kilometer of urbanized area.
• Proximity to Transit: Measured as the distance to the nearest transit stop, and instrumented using the average transit usage across potential residential areas.

Methodology

• The study estimates two types of models:

  • A commute mode choice model using multinomial logit (McFadden 1974) to analyze the choice between driving, rail, non-rail transit, and non-motorized transport.
  • A model of vehicle ownership and VMT to understand how urban form and transit supply affect the number of vehicles owned and miles driven annually.

• The NPTS data includes information on:

  • Automobile ownership and annual VMT.
  • Commuting behavior, including mode of travel and distance traveled.
  • Demographic and socioeconomic characteristics of households.

• The sample includes 6,470 workers from 26 cities with rail transit, with the majority (79.7%) using private transportation. This is lower than the overall NPTS average (86.5%) due to the inclusion of New York, which has a high proportion of public transit users.

Conclusion

The paper concludes that while individual measures of urban form and transit supply have a small but significant impact on travel demand, the combined effect of altering urban form and increasing transit supply can lead to substantial reductions in VMT. The results suggest that efforts to reduce auto dependence through urban planning and public transit expansion are likely to have modest effects on travel behavior, but they can still contribute to improved urban mobility and reduced environmental impact.