On 'Travel and the Built Environment'
Third, for a new, green, energy-efficient building in the United States, the carbon footprint of building users moving between the building and their other destinations is likely to be greater than the carbon footprint of the building itself. Like the building's direct carbon footprint, this building-access carbon footprint must also achieve long-term, very deep reductions.
[an error occurred while processing this directive]
As Kaid Benfield wrote recently:
"We know from research that where we put a building can have a bigger impact on the environment than how we design it, through transportation emissions and the impacts related to associated neighborhood infrastructure. Even the "greenest" building (judged internally) will hurt more than help the environment if it is placed in sprawl or an otherwise unwalkable location.
"In other words, my view is that sustainable architecture is only worthy of the name if it is in the right places, and includes design that respects and enhances the community around it, including neighborhood walkability."
Simply put, creating green buildings without understanding and controlling their impacts on travel demand is likely to be a self-defeating exercise.
The greenest green building in a sprawl location is probably more of a climate-buster than an average building built to current code in a central urban location.
In "Travel and the Built Environment," Ewing and Cervero significantly upgrade our formal understanding of how to weight different factors contributing to the travel demand induced by buildings.
Taking on a large body of existing studies, using the well-established approach known as meta-analysis, Ewing and Cervero find, in planner-speak, that:
• "[V]ehicle miles traveled (VMT) is most strongly related to measures of accessibility to destinations and secondarily to street design variables...
• "Walking is most strongly related to measures of land use diversity, intersection density, and the number of destinations within walking distance...
• "Bus and train use are equally related to proximity to transit and street network variables, with land use diversity a secondary factor...
• "Surprisingly, we find population and job densities to be only weakly associated with travel behavior once these other variables are controlled."
The top-line result, that location efficiency is a larger determinant for VMT reduction than neighborhood type, is appropriate, and as we at ArchitectureWeek would have expected. This is an extremely important finding.
It's as simple as putting new buildings close to metropolitan centers. This finding shows that, through the essential siting process, building development teams can have a substantial effect on the amount of driving induced by any given project.
If we get this right, collectively and over time, our cities may actually be able to grow their way toward shrinking per-capita VMT, just through the right land use planning, implemented with responsive architecture. Intelligent, evidence-based low-VMT urban development patterns may even lead us to shrinking absolute VMT eventually.
Certainly, urban development that doesn't take the geography of VMT into account will be leading in the wrong direction, expanding driving per person even as the number of people increases, too.
As Ewing said in a SmartPlanet interview, "The best way to minimize driving appears to be to develop in existing centers near the core of the metropolitan area, in areas of high destination accessibility... That's the most important single factor."
In another fundamental finding of their new deeper meta-analysis, Ewing and Cervero refine previous work by deposing "density" as the key factor determining the low VMT long associated with urban core areas.
"Conventional wisdom holds that population density is a primary determinant of vehicular travel, and that density at the work end of trips is as important as density at the home end in moderating VMT. This does not appear to be the case once other variables are controlled" (page 22).
It is great to move past "density" in and of itself as a credible VMT reducer, because as a false measure, it was pernicious.
The obsolete idea that simply packing residences closer together would shrink the driving of people who live in them has provided a pretense even in the hands of the well-intentioned to justify many new "green" neighborhoods, new towns, new outlying town centers, and new outlying mixed-use developments.
In fact, since the distance to the nearest metropolitan center is a several-times-stronger predictor of driving than density per se (stronger than density and mix of uses, combined), the creation of dense outlying new towns and developments can actually worsen the problem of excess driving, by putting even more people in high-driving locations than would be there in lower-density development patterns.
This shouldn't all be a mystery. But in planning circles, it has been until now. How did we get to the point that it's a surprise that density is relatively irrelevant? And where is this line of research likely to go in the future?
An earlier generation of research on land use and transportation relied heavily on aggregated data data that lumped together, for instance, all the information for each city in a comparative study.
This was an expedient approach for the field, but it opened the doors to some fairly obvious potential artifacts.
For instance, imagine that a non-geometric approach was used to compare two metropolitan areas which both show (hidden beneath aggregated data) the classic radial gradient pattern of VMT low VMT per capita in the urban core, well below average for the area, increasing gradually out to the suburban edge, where VMT per capita is much higher, often twice the area average.
Imagine that residential density in these two cities generally follows the same common pattern high density in the urban core, tapering out to low density where the metropolitan area meets the countryside.
Now, imagine that in the first metropolitan area, the city itself, City A, happens to have its city limits drawn halfway between the urban center and the urban edge.
And imagine that in the second metropolitan area, the city itself, City B, has its city limits way out at the urban edge.
The overall metropolitan pattern for these two cities is the same. However, comparing density and VMT for these two cities based on aggregated data by city, we would find that City A has high density and low VMT, while City B has low density and high VMT. It's a meaningless result, an artifact created by the arbitrary aggregation or combining of data by city.
Discuss this article in the Architecture Forum...