By Jerry Walters

This is the third in a series of reviews of recent studies on transportation, land use and climate change.  The first two reviews addressed:

• Moving Cooler — by Cambridge Systematics for ULI, EPA, USDOT, et al
• Driving and the Built Environment – by a group of academics and national experts for TRB.

Each review compared the study findings with those of Growing Cooler – The Evidence on Urban Development and Climate Change, a 2008 ULI book co-authored by Fehr & Peers.

This third installment reviews:

• Travel Behavior, Residential Preference, and Urban Design: A Multi-Disciplinary National Analysis, by University of North Carolina, September 2009.

This report states in its summary and conclusions that Neo-Traditional Developments (NTD) generate more trips than typical suburban development. 

It is a very ambitious study that involved identifying 17 cases of paired neighborhoods, each within the same community, and comparing the neighborhood designs and travel behavior between members of each pair. The 17 communities included Carlsbad and Chula Vista in CA, Clackamas OR, and Colorado Springs, Longmont and Fort Collins CO as well as cities in Texas, Illinois and eight in the southeastern US.  Within each city, they measured and compared the two neighborhoods in terms of design (including the 5D’s of density, mix, urban design, infill vs greenfield, transit vs no transit) and they surveyed residents on their daily trip making.  To quote their stated findings:

“We found that residents of NTDs neighborhoods make more trips, more car trips, more non-motorized trips, and more trips internal to their neighborhood than residents of typical suburban neighborhoods. We found no difference in vehicle mileage, and thereby conclude that trips taken by NTD residents tend to be shorter in length than trips taken by their suburban counterparts. Furthermore, we did not find a statistical difference between neighborhood type residents in the frequency of external trips. This suggests that the difference in overall trips detected is the result of greater internal trip capture by the NTDs”.

However, careful read of the report indicates a very mixed message at best.  Among my observations:

• In most of the pairs there is very little difference between the two neighborhoods.  In some cases the use mix was better in the “suburban” neighborhood than the NTD, in others aspect of the suburban street connectivity were better than the NTD, and in others the differences even in density were too close to matter.  By my assessment, only four of the neighborhoods were significantly different from their mates, eight were only moderately different from their mates, and in five cases the differences were negligible. 
• By their nature, both members of each pair have the same transit service and both members of each pair were either greenfield or infill locations.  So the distinctions with respect to two of the primary determinants of trip reduction were nil.
• In the 17 total, only four of the pairs were infill and only nine had transit service.
• All of the neighborhoods, both NTD and suburban, were entirely single family residences.  On average across all of the neighborhoods, about 20% to 25% of the residents are retired. This may suggest that some of the pairs, especially in the Southeast, cater to retirees and are less representative of the full-spectrum developments we usually study.
• In spite of heroic efforts to survey household trip making, the study reports the reliability of its trip-making estimates at only 51%.

Finally, in spite of the bold statement in the report’s summary, Table 11 which contains a comparison of average trip making for the group of NTD versus suburban indicates that the survey means for NTD exhibit about:

• 12% lower VMT than suburban
• 5% lower vehicle trips than suburban
• 20% higher internal trips than suburban
• 120% higher non-motorized travel than suburban

They did construct models from their data and attempted to parse the differences to a more specific set of individual paired comparisons that may have lead them to their summary conclusions, but bear in mind that their travel survey was only 51% reliable, and only 4 of their 17 individual pairs had notable differences from one another with respect to density, diversity or design, and none had differences with respect to destination accessibility or transit.  It seems questionable that one would want to parse such data too finely. 

While not directly comparable to this study, others (including some of our own research) are finding that those who reside in remote locations tend to combine their trips more than those closer-in with all of their destinations nearby.  Trip chaining can result in lower vehicle trip generation for remote development, although household VMT is still considerably higher than for infill, mixed and transit-oriented development.