skip to main content

→ Top Stories:
Clean Power plan
Safe Chemicals

Kaid Benfield’s Blog

The definitive study of how land use affects travel behavior

Kaid Benfield

Posted June 4, 2010

, , , , , , , , , , , ,
Share | | |

Enraged at the spill in the Gulf and the American appetite for oil that ultimately caused it?  Stop land development on farmland, forests and other fringe locations and direct future development to close-in opportunities.  A massive new study, years in the making, makes it crystal-clear that it can make a big difference.

In particular, transportation uber-researchers Reid Ewing (University of Utah) and Robert Cervero (UC-Berkeley) have published a painstaking ‘meta-analysis’ of nearly 50 published studies on the subject of land use and travel behavior.  Writing in the Journal of the American Planning Association, the two return to a subject to which they have dedicated most of their careers, in this case updating their previous meta-analysis from 2001. 

What they found: location matters most when it comes to land use, driving and the environment. 

The study's key conclusion is that destination accessibility is by far the most important land use factor in determining a household or person’s amount of driving.  To explain, revitalization in Columbia Heights, DC (by: the courtyard, creative commons license)'destination accessibility' is a technical term that describes a given location’s distance from common trip destinations (and origins).  It almost always favors central locations within a region; the closer a house, neighborhood or office is to downtown, the better its accessibility and the lower its rate of driving.  The authors found that such locations can be almost as significant in reducing driving rates as other significant factors (e.g., neighborhood density, mixed land use, street design) combined.

The clear implication is that, to enable lifestyles with reduced driving, oil consumption and associated emissions, environmentalists should continue to stress opportunities for revitalization and redevelopment in centrally located neighborhoods.  As Ewing and Cervero put it:  'Almost any development in a central location is likely to generate less automobile travel than the best-designed, compact, mixed-use development in a remote location.'

The authors carefully examined each study, applying statistical analysis to tease out which land use factors had the biggest impacts on travel behavior when extraneous factors such as income were controlled.  After discussing destination accessibility, the authors continue:

‘Equally strongly related to [vehicle miles traveled] is the inverse of the distance to downtown. This variable is a proxy for many [other factors], as living in the city core typically means higher densities in mixed-use settings with good regional accessibility. Next most strongly associated with VMT are the design metrics intersection density and street connectivity. This is surprising, given the emphasis in the qualitative literature on density and diversity, and the relatively limited attention paid to design. The weighted average elasticities of these two street network variables are identical. Both short blocks and many interconnections apparently shorten travel distances to about the same extent.’

it works (via Federal Highway Administration)My designer friends have educated me considerably over the past few years on the potential effects of a good street network, and they are thrilled that the study has confirmed their expectations.  A whole bunch of Facebook and Twitter updates popped up on the subject earlier this week (mostly referring to this good summary of the study’s discussion of intersection density).  None of the updates mentioned the even more critical location findings, but maybe that’s because designers can’t design a location and thus have less professional interest in that aspect.  Those who work at the neighborhood scale can affect local street networks, and we are fortunate that some of them are very good at it.

Speaking of street networks, Ewing and Cervero found that they have a huge influence on how much walking is generated from a given location:

‘Among design variables, intersection density more strongly sways the decision to walk than does street connectivity. And, among diversity variables, jobs-housing balance is a stronger predictor of walk mode choice than land use mix measures. Linking where people live and work allows more to commute by foot, and this appears to shape mode choice more than sprinkling multiple land uses around a neighborhood.’

Distance to a store was the second most influential factor in influencing walking, with location and the accessibility of transit next.  To an extent, this departs from the findings of the huge SMARTRAQ study of neighborhoods in Atlanta, which found land use mix to be more significant.  SMARTRAQ was included among the studies in the meta-analysis, though; perhaps it did not do enough to analyze other variables, or maybe it was just an outlier for some reason.

walking in Columbia Heights, DC (by: Elvert Barnes, creative commons license)Interestingly, neighborhood density, when separated from the other factors, was found to be significantly less significant than other factors in influencing both miles traveled and vehicle trips, although still influential.  On its face, this would seem to contradict the substantial body of literature that associates increasing density with reduced driving, including the source of one of the graphs in my post earlier this week on stormwater runoff.  Ewing and Cervero suggest that perhaps measures of density are inadvertently acting as proxies for other significant factors (‘i.e., dense settings commonly have mixed uses, short blocks, and central locations, all of which shorten trips and encourage walking’).

Transit accessibility is correlated with both reduced miles traveled and more walking, though not to the extent that location and street networks are.  Transit use was most closely associated with, in addition to distance from a transit stop, the street network – especially the presence of four-way intersections (for a good explanation of why this may be so, go here).

Ewing and Cervero spend a good deal of space attempting to address the issue of ‘self-selection’ or whether, to give an example, people walk more in places with a good walking environment because they are predisposed to walk and choose to live there, rather than because the environment entices them to.  Reid Ewing explained this to me at some length over the phone a few years back, and I understand it, but I’m not entirely sure it matters.  driving, Newburgh, NY (by: Daniel Case, Wikimedia Commons)All indications in the market suggest that we have a large, growing, unmet demand for close-in, walkable neighborhoods and an emerging surplus of automobile-dependent environments; research consistently shows that, where walkable neighborhoods in smart locations exist, walking goes up and driving goes down.  Unless that unmet demand for close-in, walkable environments somehow turns into a surplus, which isn't happening anytime soon, building more of them will reduce driving and increase walking.  The environment doesn't care what the psychological motive is.  In any event, the authors found in this case that applying research controls for self-selection might, if anything, show an even more significant influence of land use on behavior.

Since we tried hard to reward good, accessible locations (while keeping the most remote ones out) as well as intersection density in LEED-ND, I think this research validates our approach.  It does make me wonder, though, if we are giving enough emphasis to revitalization and street networks when we advocate in the arena of transportation policy.  It also makes me wonder why more environmental groups, clearly incensed at BP and the Gulf oil spill, aren't paying more attention to land use.

Move your cursor over the images for credit information.

Kaid Benfield writes (almost) daily about community, development, and the environment.  For more posts, see his blog's home page


Share | | |


Barbara McCannJun 4 2010 05:25 PM

Great discussion of the details of density and design. Does it all boil down to 'location, location, location?'

Granted this was not peer reviewed, but I got a little validation from a recent Grist article on 'ten ways cities and towns can kick the offshore oil habit" Link:

Kaid @ NRDCJun 4 2010 09:37 PM

Nice plug in that one for your work on complete streets, Barbara.

Don CJun 4 2010 10:35 PM

I'm very glad to see this updated meta-analysis. I agree that while self-selection is a very important research issue, the benefits of building walkable communities are pretty uncomplicated. If I were a mayor, a county executive, or even a governor, I'd certainly want fans of walkable urbanism to self-select into my jurisdiction for a whole range of reasons--to have a healthier population, fewer auto fatalities, greater retail and commercial foot traffic, etc. Great piece!

Steve MouzonJun 5 2010 09:37 PM

Great article, Kaid... Thanks! Articles like this should be everywhere, reminding us that the primary culprit isn't BP... it's us! If our behavior doesn't change, our machines won't save us. Matter of fact, so long as our appetite for oil increases as easy reserves are exhausted, we're going to have more and more situations like this as oil companies have to go to even more hostile environments to extract oil.

Laurence AurbachJun 7 2010 09:40 AM

The Ewing and Cervero study is extremely valuable and informative; it incorporates many data points and covers numerous land use issues. Choosing which finding is most important depends on one's perspective and priorities.

If one is focused on VMT, the finding about location will be the most important. However, it is not entirely new information; it confirms the research conclusions that have been rolling in for more than a decade from Cervero and other investigators.

Kaid, you wrote, "None of the updates mentioned the even more critical location findings, but maybe that’s because designers can’t design a location and thus have less professional interest in that aspect. I can't speak for any others, but I highlighted The Power of Intersection Density because (1) the findings are new, and as the authors note, surprising; and (2) the elasticity of intersection density is large -- almost twice as large as the elasticity of destination accessibility for VMT. If one is concerned primarily with walking and transit use, then the findings about intersection density will be most important.

And if one is a scholarly researcher, the most important aspect of the study is that it establishes a database of studies that can serve as a seed kernel for more extensive and rigorous research.

Laurence AurbachJun 7 2010 09:50 AM

To clarify: the elasticity of intersection density for walking is much larger than the elasticity of destination accessibility for VMT. But within the VMT category, intersection density has a smaller elasticity than destination accessibility.

Kaid @ NRDCJun 7 2010 11:13 AM

I completely agree with your second statement. If our public policy goal is to reduce VMT and oil consumption, destination accessibility is by far the most important factor and even the best design in the wrong place will not help. If our policy goal is to increase walking and promote public health through physical activity, intersection density is by far the most important factor.

We can agree to disagree on which of those is more important to communicate to the public, but surely we can agree that having a great location and a great street network is more powerful than either, alone, for sustainability.

LucyJun 7 2010 04:56 PM

I would say that it's not "location" that really matters, it's "place" that determines value, reflecting the decreased car use enabled by appropriate design as well as appropriate location.

Ewing and Cervero have (once again) shown that urbanism is more functional than sprawl. Or as we say at CNU, it's not a style, it's not about the porches, it's about places.

Kaid @ NRDCJun 7 2010 05:33 PM

Thanks, Lucy. I would say that a sustainable 'place' must incorporate both. Great design in the wrong place will not create sustainability, and a great location that fails to fulfill its promise with the wrong design is basically a missed opportunity. I stay with the location theme as strongly as I do because I have encountered so many people who think that design is all that matters.

In fact, for a true sustainable place, we need all three of the major elements recognized in LEED-ND: great location, great neighborhood design, and great green features. The time is long past when we could settle for less.

JarrettJun 7 2010 09:45 PM

Thanks for the link. As I argued in the post, I think that new urbanist greenfield development may not add up to a very legible city. Think about how easy it is to remember the geography of Midtown Manhattan, where you can describe everything as a set of grid coordinates (Second Ave at 73rd St).

Grids can help to make an entire city feel legible. The little sample of well-connected but not regular street network that appears in your post can be headache-inducing if you replicate it across the city, especially if you're trying to define bike paths, etc, along the collector and minor streets.

More on this here, along with a very very smart comment thread on the subject, here:

Jarrett Walker,

Tim TormaJun 8 2010 09:21 AM

Kaid, Thanks for another thoughtful and informative posting. I also wanted to take this opportunity to let folks know that this study was funded by the EPA Smart Growth Program.

Laurence AurbachJun 8 2010 03:19 PM

Kaid, you surely are stating the truth of the matter. When all of the elements of sustainable development are present and working in combination, they have synergistic benefits that are much greater than those of any single element by itself. That may be the most important lesson for elected officials and policy makers.

LEED-ND has a three-part structure of smart location and linkage, neighborhood pattern and design, and green infrastructure and buildings. To some extent that's a reflection of the three partners who created LEED-ND: the Natural Resources Defense Council, Congress for the New Urbanism, and US Green Building Council. But it's also gratifying to see research confirming that these three themes working in combination is the best, most effective way to achieve the environmental and societal benefits of sustainable development.

Comments are closed for this post.


Switchboard is the staff blog of the Natural Resources Defense Council, the nation’s most effective environmental group. For more about our work, including in-depth policy documents, action alerts and ways you can contribute, visit

Feeds: Stay Plugged In