Trail underpass in Greenbelt, Md.

A century ago, cars first started appearing on city streets and subsequently began running over everyone else. The 1920s epidemic of traffic fatalities, particularly among children, led to a fad in town planning to rigorously separate paths for cars and people, and to turn over the roads entirely to cars.

This philosophy was implemented in several New Towns — most notably in Radburn, New Jersey, but more extensively in three suburban Washington communities (Greenbelt, Columbia, and Reston). For the most part, though, existing cities proved too expensive to retrofit, and most other American suburbs just built the car-only roads without their pedestrian counterparts.

Greenbelt Museum

In both Radburn and Greenbelt, houses were even “turned” to face the walkways, with back doors facing alley-like auto access roads. Yet as these places have matured, most households have rotated their houses and enclosed their public front yards into private backyards; communities have added sidewalks to the streets. (Above, the museum/house at right has been restored to its original appearance, but the house at left is typical.)

Trail crossing street in Lake Anne Village, Reston, Va.

Examples built from the 1960s-1980s, like Reston, typically threaded trails past backyards. LIFE magazine wrote in 1965: “The need for automobiles in the village is virtually eliminated by a network of footpaths and bridges which connect residential areas with the shopping center.” (Indeed, my own 1960s neighborhood was replatted with a half-completed secondary grid, which I do use occasionally.)

Interestingly, two of my Streetsblog colleagues, Jeff Wood and Steven Vance, both lived in Kingwood, outside Houston, and both made extensive use of the trails as kids. In an email exchange, Steven writes:

Those trails, the Greenbelt, connected every subdivision with homes (there was an entrance 50 feet from my house), community pools, and passed behind every school. They also took me to the shopping areas. I “delivered” myself and my sister to school and the pool.

Jeff Wood made a distinction between walkability and runability on a recent episode of the Streetsblog podcast (sorry, can’t find the link!). While the trails in these communities are still prized as a recreational amenity, they are not a useful travel option for adults because they’re not really “irresistible”: distances are far (because densities are low), travel paths are circuitous, and there’s frankly not much to look at. While those attributes were suitable for Jeff’s track and field team, they don’t suffice for people walking to destinations — and who have the option of driving instead.

It later occurred to me that neighborhood-level household travel data is in fact available for one of these New Towns. In 2011, MWCOG did a geographically focused household travel survey aimed at understanding community scale differences in travel behavior. The chart above shows weekday trips for all purposes, arranged in distance from the regional core. The first five are within the Beltway, and the second five are outside the Beltway, but only Logan Circle is prewar fabric (within the L’Enfant City, even).

For a control, though, I’ll focus on White Flint. Both the Reston and White Flint survey areas have high household incomes, a high proportion of multifamily (49% in Reston, 61% in White Flint), upscale regional retail and substantial office employment at their core, and began development in the 1960s. White Flint has an established Metro line, higher density (6,000 residents/sq. mi.), and smaller households (1.4 average size). Reston’s Metro stop only opened recently, but it has had local and express bus service to other regional centers, larger households (2.2, still small by suburban standards), and generous open spaces that reduce overall density to 4,000/sq. mi. Most importantly, though, Reston has an extensive trail network, while White Flint has better street connectivity.

Sure, Reston has a substantial walk mode share of 14% — more than 50% higher than the regional average of 9%. Yet it’s not much higher than Woodbridge or Frederick, both of which have large expanses of haphazardly planned subdivisions surrounding a walkable “old town.” And it’s substantially below the 18% of trips that White Flint residents make on foot.

Steven points to Boulder and Davis as counter-examples: [T]hat kind of trail network that helps Boulder, CO, and Davis, CA, have high bicycling rates… [Boulder’s] bike map emphasizes their 10% bike-to-work mode share, while in Chicago we can’t seem to nudge past 1.6% (although for some neighborhoods it’s pushing 4.5%).

Yes, extensive trail networks are a great thing to have at a regional level, and certainly have something to do with the relative popularity of cycling in metro areas like the Twin Cities, Denver, and DC. Yet in explaining Boulder or Davis, I’d give less credit to infrastructure and more credit to favorable demographics, culture, and geography — they’re compact, connected, relatively high density, and mixed-use.

Every once in a while, I’ll encounter advocates who argue that trails are the only acceptable environment for pedestrians or cyclists, occasionally pointing to surveys showing that off-street trails are rated in most surveys as the gold standard of bike facilities, especially for those who don’t already regularly bike.

Sure, a trail network is great to have, and should be connected or created at every opportunity. However, we already have a fantastically extensive bike/ped circulation system that touches everyone’s front door — it’s just that our streets are filled with dangerous cars.

Yet the same surveys show that excellent on-street infrastructure is almost as good at enticing people onto two wheels. And since most new development will happen in places that are already built, it’s kind of a moot point, anyways.

Focusing on building up the fabric, and adding the best infrastructure available, is also the approach that Reston and White Flint are taking, and I expect that future household travel surveys will show that the approach will bear fruit for both areas.

[based on an email exchange]