Land Use–Transportation Scenarios and Future Vehicle Travel and Land Consumption: A Meta-Analysis

Abstract

Problem: Since the late 1980s, U.S. metropolitan regions have increasingly engaged in a style of land use–transportation scenario planning that merges techniques borrowed from military and business strategic planning with long-range transportation systems planning and project-level alternatives analysis. Aside from occasional anecdotal descriptions, the effectiveness of such approaches in generating compact growth plans has not been evaluated comprehensively.

Purpose: We analyzed a wide range of scenario planning studies to determine how far compact growth scenarios are predicted to reduce vehicular travel below existing trends.

Methods: Using hierarchical modeling, we developed a regional vehicle-miles-traveled (VMT) model based on 85 scenarios in 23 planning studies from 18 metropolitan areas.

Results and conclusions: Using coefficients from this model, we conservatively estimate that compact growth scenarios reduce VMT in 2050 by 17% below scenarios assuming a continuation of existing trends.

Takeaway for practice: Existing transportation models remain largely insensitive to changes in land use and transportation policy. This must change for scenario planning to achieve its full potential. In the future, scenario planning should incorporate the best current knowledge about how global economic and environmental trends will affect regions.

Research support: Collection of the primary data used in this article was funded in part by the Federal Highway Administration, U.S. Department of Transportation, under Cooperative Agreement No. DTFH61-03-H-00134. Any opinions, findings, and conclusions or recommendations expressed in this publication are those of the authors and do not necessarily reflect the view of the Federal Highway Administration.

Keywords:

• scenario planning
• land use
• transportation
• meta-analysis
• hierarchical model

Notes

1. An annotated bibliography of the projects is available at http://content.lib.utah.edu/u?/ir-main,101. Many of the project reports that were used as source data for this article are available at http://www.lib.utah.edu/digital/collections/highways/ At this site, click on “Digital Collections A-Z” and select “Land Use–Transportation Scenario Planning.”

2. A regional density increase of 50% is within the range of the scenarios studied in this article (see Figure 5). Although it is near the top end of that range, the median scenario study horizon is 25 years. Therefore, a 50% increase in density over 43 years is reasonable. Moreover, the major demographic shifts anticipated through 2050 suggest that this assumption may be conservative (see Ewing, Bartholomew, Winkelman, Walters, & Chen, 2008; Nelson, 2006).

3. Computed using the coefficient values from Table 2 as: (−0.074 × 50) – (1.50 × 1) – (4.64 × 1) – (0.068 × 73) – (2.12 × 1). The 73 represents a growth increment of 73%, or 43 years at an average growth rate of just over 1.28% per year.

4. In Contra Costa County's Shaping Our Future project, the study team used a spreadsheet-based sketch model to adjust the MPO's standard travel demand model for travel behavior changes resulting from changes in local land use patterns. The adjustments were based on travel elasticities derived from Bay Area household surveys, census data, and neighborhood paired-comparison studies. For each subarea zone in the study area the study team established an index combining population and employment density per acre; the mix of population, retail and nonretail employment; and a measure of pedestrian friendliness for each scenario. They then calculated the percentage differences between the indices for the scenarios in each subzone and applied the corresponding elasticities to create adjusted trip tables for each scenario (Bartholomew, 2005a).

5. Sketch models are broad-based decision support tools that use elasticities derived from a wide range of data sources. For estimating travel demand impacts of alternative land use scenarios, sketch models, such as EPA's Smart Growth INDEX Model, “can demonstrate the direction and magnitude of change and calculate rough estimates of relative impacts” (U.S. Environmental Protection Agency, 2003, p. 4). However, because sketch models are usually not calibrated to data from the region where they are applied, they are not considered appropriate for demonstrating compliance with regulatory mandates, such as those associated with air quality conformity.

6. The pricing policy assumes that all those in the area who drive alone to work will be charged $3.00 per day to park, and this revenue will be used to provide free transit passes to all commuters in the study area.