The Environmental Impacts of Land Development Depend Largely on Where We Put It

08/11/2015 08:53 am ET Updated Aug 10, 2016

There's a trendy meme emerging in progressive city planning circles to the effect that whether land development is harmful "sprawl" or benign "urbanism" is a matter not of location but of design. I recently saw a tweet expressing this sentiment, written by an influential city planner and picked up quickly by other urban designers. Not long after, I saw a Facebook post along the same lines: "It isn't where, but what that makes a place urban or suburban."

This rings true in some situations: bad neighborhood design can certainly turn an otherwise well-positioned development into the equivalent of sprawl; and there are pockets of what are essentially urban neighborhoods in, say, suburban downtowns.

But our nomenclature gets tricky when applied to new development located on or beyond the fringe of metropolitan areas. Outlying newer developments are typically built on what was formerly farmland or forests, sometimes "leapfrogging" over available closer-in sites to do so. Those are classic characteristics of suburban sprawl. Nevertheless, many of them today are designed to mimic the feel of older city neighborhoods, with higher densities, more walkable streets, and more diverse land uses than conventionally sprawling "pod" subdivisions? Does that make them "urban" and environmentally benign?

I suppose that the extent to which such a development qualifies as "sprawl" or somehow "urban" despite its outlying location is a matter of semantics. (I'm on record as saying that a development doesn't have to be low-density to be sprawl, and that even internally "green" sprawl is still sprawl.) And, whatever terms we use, I would certainly concede that well-planned outlying development is somewhat better for the values that I care about than poorly planned outlying development.

But we must be very careful about the implication that location - where we put new development - doesn't also matter a great deal to those values, particularly with respect to a development's environmental performance. Both design and location matter.

Let's start with the elements of neighborhood-level design: Whether a development has a good street pattern, is walkable, uses land efficiently, provides shops and other lifestyle conveniences and amenities conveniently accessible to homes, has green buildings and infrastructure, and so forth will have an effect on how it performs with respect to carbon emissions, air and water pollution, ecosystem conservation, and other environmental indicia. By and large, the architects and planners associated with the New Urbanist school of design get these neighborhood principles right, and those of us who are environmentalists owe them a great deal of appreciation for bringing good practices and influence to the world of real estate development.

However, it must be emphasized that, as much as the design of a development matters to the environment, its location can matter even more. The best, most compact and internally walkable design in the world is unlikely to overcome the impacts associated with leapfrog sites that pass over developable land to settle on locations even farther outside the existing footprint of a metro region.

The research on development location

Those impacts are many, but perhaps the most measurable is automobile dependence and associated emissions of carbon dioxide, a potent greenhouse gas. The research demonstrates that location is by far the most significant indicator of how much driving typically takes place to and from a given neighborhood.

This is because of something called "destination accessibility": outlying locations have fewer jobs, shopping opportunities, schools and other typical trip destinations within easy reach than do more central locations, causing average driving distances to be longer. (It is also generally easier in more central locations to substitute transit and walking for what would otherwise be driving trips, but such "mode shifts" are statistically less significant to vehicle miles traveled than are driving trip distances.) As a result, carbon emissions from outlying locations, per person and per household, tend to be higher - typically a lot higher - than those from closer-in locations.

On everyone's short list of leading researchers on the relationship between travel and the built environment are Professors Reid Ewing of the University of Utah and Robert Cervero of the University of California at Berkeley. The two have been studying the effects of various neighborhood location factors and design elements on vehicle trips and vehicle miles traveled for decades, in 2001 and 2010 publishing two painstaking "meta-analyses" of all the published literature they could find on the topic. (For their 2010 article they examined some 200 studies, performing quantitative analysis on over 50 of them.)

Ewing and Cervero's analysis has found, among other things, that destination accessibility, particularly as indicated by distance from a given location to downtown, is nearly as influential in determining vehicle miles traveled per capita and per household as three factors generally associated with urbanism - neighborhood density, land-use diversity, and design factors such as the street network and sidewalk coverage - combined. They conclude that "highly accessible areas such as center cities produce substantially lower [vehicle miles traveled] than dense mixed-use developments in the exurbs."

Let's look at some examples of how this plays out in the real world, particularly with respect to carbon emissions. Fortunately, we have great data, organized by the Center for Neighborhood Technology, to draw from and make some comparisons. CNT's remarkably detailed, nationwide database of transportation and housing data (see also here) allows us to hone in on just about any given US location and learn its per capita driving rate and associated per capita emissions of carbon dioxide from transportation. We can then compare these data points to those for other locations and to the averages for the relevant metro regions as a whole.

I did this for five master-planned neighborhood developments that are all well-known for their exemplary internal designs, all based on the principles of New Urbanism. All seem, to my taste, to be very nice places to live, at least if you discount the long commutes associated with some. All are highly walkable and mixed use; all use land efficiently and contain a diversity of housing types. But they differ in their locational characteristics: three of the five are located on the outer reaches of their metro areas while two are located within central cities, though not downtown. One of the three suburban developments has a light rail transit station.

The performance of the five developments does indeed vary significantly by location; the data are generally consistent with what one would expect from Ewing and Cervero's research.

Carbon emissions for outer suburban locations

Our first development, New Town at Saint Charles, is located some 29 miles from downtown St. Louis, Missouri. I have never been there but it looks beautiful in photos. New Town is not only outlying but transit-deprived: To go downtown by transit, Google Maps suggests that you drive seven miles from the development's center toward St. Louis to pick up the #35 bus and eventually transfer to light rail. Your entire drive/transit journey should take about an hour and 19 minutes, if you were really determined to pursue it or had no other choice. Now ten years old, New Town is the most isolated of our five developments, still surrounded by farms and other open land.

According to CNT's database, a typical household in New Town would emit 0.89 metric tons of carbon dioxide from transportation every month, or close to eleven tons per year. That's a lot: it is 16 percent higher than the average emissions rate for households in metropolitan St. Louis as a whole (0.77 tons per month). For purposes of comparison, New Town's emissions rate is also significantly higher than that for the somewhat closer-in but more established outer St. Louis suburb of Chesterfield 21 miles from downtown (0.6 tons per month), and fully twice as high as that for an average household in the revitalizing neighborhood of Old North St. Louis, about a mile from downtown (0.44 tons per month).

Our second development, Kentlands in Gaithersburg, Maryland, is one of the country's best-known New Urbanist communities. It was originally built on farmland beginning in the 1980s, but it is now completely surrounded by other suburban development, much of it sprawl by anyone's definition. To me, Kentlands looks and feels very much like a suburb, but much better than classic sprawl. I really like it because it gives suburban consumers a more walkable choice. Kentlands is in the outer suburbs of Washington, DC, some 28 miles from the White House. It is served by bus transit, which links up with the region's Metrorail system.

The carbon emissions from transportation for a typical household in Kentlands are 0.69 metric tons each month, a little over eight tons per year. That's a lot better than for New Town at Saint Charles but not significantly better than the emissions for an average household in the sprawling DC region as a whole (0.71 tons per month). Kentlands' carbon emissions from transportation are almost identical to those for the similarly distant (27 miles from downtown DC) but more conventionally sprawling suburb of Germantown, Maryland (0.70 tons per month); and they are a little less than twice as high as those for an average household in DC's truly urban and Metrorail-accessible Columbia Heights neighborhood two and a half miles from the White House (0.37 tons per month).

This brings us to Orenco Station in Hillsboro, Oregon. Orenco is an interesting case because it is definitely outlying, about 15 miles from downtown Portland. It, too, feels very suburban to me, if much more compactly designed and walkable than most new suburbs. Like Kentlands, it is now surrounded by other suburban (and some industrial) development. But unlike New Town at Saint Charles and Kentlands, Orenco Station was designed from the start to be transit-oriented; its light rail station sits on metro Portland's MAX transit system.

Before looking up the numbers, my hope was that the presence of rail transit would prove a significant enough factor to bring Orenco Station's carbon emissions down below what would otherwise be expected. In fact, carbon emissions from transportation for an average household in Orenco Station are 0.54 tons per month or about six and a half tons per year. That is a very respectable 23 percent below the Portland regional average (0.7 tons per month) and a terrific number for new suburban development. (It is nonetheless significantly higher than the 0.34 tons per month emitted by an average household in Portland's highly urban Pearl District, adjacent to downtown.)

New developments with in-town locations

Now let's look at a couple of well-planned, New Urbanist developments that are located inside central cities. First, Denver's Highlands' Garden Village sits on a 27-acre site that once was a closed and deteriorating amusement park, about three and a half miles from downtown. As with the other New Urbanist developments, HGV has great walkable density, mixed uses, a diverse array of housing choices, green space, and attractive architecture. I'm a big fan. Unlike Orenco Station, it does not have rail transit; but downtown Denver is a 15-minute bus ride away.

Given its relative proximity to downtown, we would expect Highlands' Garden Village to have lower-than-average carbon emissions from transportation, and the development doesn't disappoint. An average household in the neighborhood emits 0.59 tons of carbon dioxide per month, about seven tons per year; that's 16 percent less than the average for the Denver region as a whole (0.7 tons per month), if a little more than we found for Orenco Station. HGV's emissions rate is exactly the same as that for the older Cole Neighborhood Historic District about three miles from downtown Denver.

Finally, I've saved the best-performing of these five new developments for last. One of my favorite New Urbanist developments is Atlanta's Glenwood Park, built on a former industrial site just a couple of miles from downtown. It is walkable, beautiful, compact and, to an extent, green. It is about the same size as Highlands' Garden Village and contains a similar diverse array of land uses. It is served by bus transit but not directly by rail.

Glenwood Park's average household carbon emissions from transportation (0.59 tons per month) are an impressive 28 percent lower than those for the Atlanta region as a whole (0.82 tons per month), the best of all five of the new developments I examined when considered in the context of their regions. (Highlands' Garden Village has identical household carbon numbers and Orenco Station's numbers are slightly lower; but metro Atlanta residents typically drive significantly more than do those in Denver and Portland.)

In this article, I have focused on transportation performance and associated carbon emissions because I had the data handy for those comparisons. But there's also a world of difference among various development locations when it comes to additional environmental issues such as fragmentation of ecosystems, spread of pavement across watersheds, loss of farmland and the scenic beauty of the rural landscape, and more. There's a world of difference among locations when it comes to economic and social issues, too. But if I get into all that, this will be a book rather than an article.

While what I have just done is far from a scientific analysis rigorously examining all relevant factors, the results are nonetheless consistent with what we would expect from Ewing and Cervero's far more rigorous inquiry. Even when we compare developments with roughly similar design characteristics - considered only internally, all are walkable and none would qualify as sprawl - the results vary widely by location: The three of the five that scored significantly better than their regional averages were located either well within the central city or on a rail transit stop. The suburban developments not on rail transit had emissions numbers that were, well, what we would expect from suburban developments. The most outlying development of the five, relative to its region, had the highest carbon emissions from transportation. (And truly urban, older and higher-density neighborhoods, close to downtown, scored best of all.)

I must stress that the performance differences among these examples, although stark in some cases, are in no way the fault of their designers, who in these cases are among the very best in the business. Development locations are chosen not by designers but by developers, working within the parameters of zoning and other elements of the regulatory framework established by municipalities and other government authorities.

So what's the takeaway?

What these data say to me is that, if we want to get serious about addressing carbon emissions and environmental quality, we had better think seriously not just about development design but also about development location. We should press for policies that prioritize urban infill and redevelopment sites, such as those in Denver and Atlanta that were transformed into Highlands' Garden Village and Glenwood Park. We should certainly pursue approaches that bring long-neglected older urban neighborhoods back to life with inclusive, sensitive revitalization.

We can't depend entirely on urban sites, however. While much new development will of necessity be suburban, the evidence suggests that we will minimize carbon emissions if policies favor suburban sites with strong transit service. We should also prioritize locations that are well within a region's existing development footprint, including suburban redevelopment sites, many of which are becoming available as aging commercial development goes out of service. (Sprawl "repair" or "retrofit" is a big topic within New Urbanist circles, as well it should be; I strongly agree with Rob Steuteville's recent article on the subject in Better Cities and Towns.) And public policy should be shaped to discourage development on leapfrog sites that, whether or not we call them sprawl, exhibit relatively weak environmental performance.

The environment fares best when we marry superior design to superior locations, respecting the importance of both critical issues, and especially when we infuse the result with green buildings, green infrastructure and good environmental management practices. The good news is that a lot of people are already on the job doing exactly that - many smart growth advocates, practitioners of New Urbanism, and green developers and designers among them. Let's give them all the support - and supportive policy - that we can.

Move your cursor over the images for credit information.

Kaid Benfield writes about community, development, and the environment on Huffington Post and in other national media. Kaid's latest book is People Habitat: 25 Ways to Think About Greener, Healthier Cities.

Related articles: