By Chris McCahill
Low density, disconnected development patterns—or sprawl—peaked in the mid-1990s, then declined by as much as 9 percent in the following decades, according to a new analysis of street patterns published by the National Academy of Sciences.
Because of its innovative approach—looking at street patterns rather than density or geographic extent—this new study suggests that sprawl development slowed earlier and declined more substantially than was indicated in prior reports. It also offers compelling evidence that specific policies intended to reverse sprawl have played an important role, particularly at the local level. Those policies are typically put in place to mitigate the higher vehicle use, infrastructure costs, energy consumption, and emissions associated with sprawl.
For this study, researchers from the University of California in Santa Cruz and McGill University in Montreal relied on measures of street network connectivity to characterize different development patterns—knowing that dead ends, cul-de-sacs, and three-way intersections typify sprawl. They then tracked residential development back to 1920 using tax records and census data, and classified development patterns based on their nearby street network.
They found that the average nodal degree—the number of roads entering an intersection—for new developments began dropping around 1920 (Figure 1). That trend accelerated after 1950 and then peaked in 1994. Similarly, fewer developments were built around four-way intersections and more were built around dead ends until the late 1990s when those trends reversed.
By looking at where the nodal degree increased most, the authors conclude that some change is due to urban infill, but that development patterns are changing across entire metropolitan regions, including suburban areas. They also found that the largest changes occurred in the areas around Austin, Texas; Charlotte, North Carolina; Gainesville, Florida, and other cities that have implemented policies encouraging connectivity, suggesting that those policies have been at least partly effective.
County-level data from the study are publicly available as a supplemental online component.
Chris McCahill is a Senior Associate at SSTI.