Which factors matter, and why
Research tells us that CO2 from transportation is the result of a location’s accessibility to major destinations and the design characteristics of an area. Thus, both where development goes and how it is designed matter. The following neighborhood design factors have consistently been significant in predicting carbon dioxide emissions in analyses in King County. Socioeconomic variables such as income, age and auto ownership — known to affect household transportation decisions — have been taken into account.
Land-use mix and street connectivity. Both of these are associated with reduced trip distances — a mixed land-use pattern puts common destinations close to home and work, and an interconnected street network creates more direct paths between destinations. When distances get short enough, auto trips can be replaced by less polluting modes of transport such as bicycling or walking.
In our 2007 study, funded by the Washington State Department of Transportation, we found that having shops and services near workplaces, and having good conditions, influenced mode of travel to work and increased the likelihood of walking for mid-day work-based trips. In studies of employment centers in California, Robert Cervero linked an increase in land use mix at employment centers to higher shares of employees taking transit and ride-sharing to work, and fewer vehicle trips.
Location within the region. To measure and control for the influence of regional location, we have employed measures such as average commute time or distance. The sheer size of an urban region, how much it sprawls, and the distance from a household to major population and employment centers significantly influence how much people drive and, in our analyses, the volume of transportation-generated greenhouse gases. A 2002 national study by Reid Ewing and others, Measuring Sprawl and Its Impact, concluded that “the degree of sprawl was the strongest influence on vehicle-miles traveled per person” — more than metropolitan population and per capita income. Because a smaller urban area reduces average trip distances, containing development at the fringe of the urban area will be quite effective in reducing VMT (and by extension, greenhouse gases).
Regional transit service quality, as measured by transit travel time, is another significant factor in predicting CO2 from travel. Residents of even the most walkable neighborhoods will need to get to work, school, doctors’ appointments, or social outings outside their neighborhood. To lower the greenhouse gas emissions, more of these longer trips will likely need to be by transit.
Residential density. Generally, residential density is not statistically significant once land-use mix and street connectivity are considered. This is probably because places that are denser, tend also to be more mixed and more connected. From a practical standpoint, residential density is needed to support neighborhood services, and is the mechanism enabling many residents to be within walking distance of local-serving commerce. Density alone is not enough to create a walkable neighborhood — all the factors above, and probably others (such as an attractive streetscape) are needed.
Looking at these strategies from an implementation point of view, it is important to note that they are extremely interconnected. For example, convenient transit service will help to serve regional travel needs, but it requires supportive land use patterns and a high-quality pedestrian environment to be effective. Taking the bus to work in a single-use office complex — with little access to services, restaurants or other offices — means errands, lunch and meetings are difficult or impossible without a car. — Lawrence Frank and Sarah Kavage
