Frank 2000 - "Linking land use with household vehicle emissions in the central Puget Sound"

Frank, Lawrence, Brian Stone Jr., and William Bachman. 2000.
"Linking Land Use with Household Vehicle Emissions in the Central Puget Sound: Methodological Framework and Findings."
2000, Transportation Research Part D 5, 3: 173-96.
On the web
Relevance: high

Frank and colleagues used data from the Puget Sound Transportation Panel (a survey of 1,700 households taken every 2 years) to estimate the total amount of vehicle pollution (carbon monoxide, nitrogen oxides, and volatile organic compounds) generated by households in different kinds of neighborhoods.

They concluded that households in higher-density neighborhoods, with more interconnected street grids, and with greater mixes of land use, produced lower total emissions than households in more sprawling neighborhoods.  Also, as might be expected, long-distance commutes increased total household vehicle emissions.  Perhaps more surprisingly, commutes to places with very high employment density (e.g., downtowns) were associated with lower total household vehicle emissions -- though this effect that was seen mostly for the places with the densest employment.

  • Frank assumed that all households used the same kind of car -- an assumption that is probably too generous to sprawling neighborhoods, as residents of low-density places also tend to have larger vehicles [I think Golob and Brownstone say this].
  • The emissions model took into account the fact that vehicle emissions tend to be highest for the first few minutes after a cold start -- which could mean that places that generate more trips also generate correspondingly higher emissions.
  • Denser places with interconnected streets tended to produce more vehicle trips per capita -- essentially, people are more willing to drive to the store for a quick errand -- but still had lower total emissions because household members drove less.
  • The household density, street connectivity, and residential land-use mix patterns were similar -- big drops from the lowest-density or connectivity quartile to the second-most, and shallower but still significant declines for subsequent quartiles.
  • The employment density - vehicle emission relationship was non-linear: it was fairly flat for the bottom 3 quartiles of employment density, but a big drop for top quartile (e.g., people who work downtown).  Presumably, this is because people who work in places with high employment density can take transit; but it may also be because people who work in downtown tend to live in close-in neighborhoods that are denser, and have greater land-use mix & connectivity.
  • The relationships all seemed strongest for NOx.

Eric's post on Cascadia Scorecard.

Comments

Eric de Place

Below, a separate summary of this article that I wrote for the research team. (It captures most of the points made above. I'm including it here for the sake of collecting all our thoughts in one place.)

The study calculates three types of emissions--NOx, CO, and VOCs—by calculating driving distance, speed, travel time, and emissions from starting the car (adjusted for estimated engine temperature at start). Its findings are based on data from the Puget Sound Transportation Panel Travel Survey, which records travel for 1,700 households over a two-day period by giving each member of the household over 15 a diary for recording trips and their characteristics. The study controls for household size, household income (using a dummy variable), and the number of household vehicles.

Employment density of the work tract. The strongest land-use correlate to low household emissions is job-site employment density--not residential density. Frank reasons that this is because transit to places with high employment density, especially downtown Seattle, which is by far the most dense job center in Puget Sound, are cost- and time-competitive with driving because of better transit service, more auto congestion, and higher parking prices.

Household emissions do not observe a negative linear relationship with employment density. For the lowest three quartiles of employment density household emissions are about the same (they're a little higher in the lowest density quartile), but then they drop off sharply at the beginning of the highest density quartile. This suggests that there's a threshold of employment density--perhaps the density at which transit, carpooling, etc become viable--after which emissions drop quickly.

Household density of the residence’s tract. Residential density is less strongly correlated with lower household emissions than is employment density at the work tract. There is still a correlation--higher residential densities mean less vehicle emissions--but the difference, while significant, was relatively minor. Interestingly, the largest drop in household emissions appears to occur in the difference from very low residential densities to the next lowest quartile.

Employment density et al of the residence’s tract. Households located in census tracts with high employment density, greater mixes of land-use, and greater street network density generate more vehicle trips and more vehicle trips with a cold engine, which produces a disproportionate share of tailpipe emissions. Frank thinks this could be because there are more services and amenities nearby and there is less incentive to "chain" trips together as a typical suburban commuter might on the way to or from work. But even so, the households in higher densities produce fewer emissions simply because their trips are not as long as the trips taken by households in lower densities.

A couple of concerns about this study. 1) Frank controls for the number of vehicles per household. It is not clear that it’s fair to use this control, as one of the primary benefits of density is alleged to be lower numbers of vehicles per household. This may result in an understatement of the air quality benefits of higher densities. 2) Frank assumes that the vehicle fleet is identical across densities. There’s reason to believe that households in lower densities and households that drive farther actually have less efficient, and more polluting, vehicles than households at higher densities. Once again, this may result in an understatement of the air quality benefits of higher densities.

The comments to this entry are closed.