Thursday, December 15, 2011

Intersection Density, Disinvestment, and Decay

On twitter this morning and in recent discussion about the West Dallas Plan and its relation to the contentious Sylvan:Thirty project I've been writing/talking a good bit about block structure. Particularly, block size and the surrounding network, its complexity and interconnectedness as measured by intersection density. I have several contentions regarding the West Dallas Plan, but as it regards to Sylvan:Thirty, I found the biggest glaring weakness is the lack of definitive block size. There are no maximums. Therefore, there is no certainty that a complex, interconnected, safe, desirable, walkable urban neighborhood will ever emerge.

This centers on the work by Prof. Norman Garrick at the University of Connecticut, who has shown direct correlation between intersection density and the safety of an area, modal share, and walkability. I find this to be not just correlated, but causal and predictive as towards the future value of a neighborhood. Limited network density, means decreased mobility, increased risk of severe injury, therefore less desirable, and ultimately less valued by individuals, the market, and eventually investors. They often don't realize it well until the place has cleared out and its lost critical mass and then real estate values plummet.

This fall, the City of Dallas initiated a new planning project focused on five DART station area plans. All are in rather desperate areas and the plan is that after initial planning efforts, each will have a first phase of affordable, workforce housing for the kind of people that need to be near DART station in order to have mobility so as to not have the mandatory expense of car ownership.

However, DART is only one form of mobility, regional mobility. There also must be local mobility. So when going after this project, I put together the following studies showing intersection density and real walkability as opposed to the typical 10-minute walk circles. (The project has been given to Fregonese/Calthorpe)

Also, it should be noted that distance-based walkability is strictly an objective measurement. How far can you get in ten minutes? What is within convenient reach? In the following imagery, I'm only showing that study. I've left out the subjective analysis (as you can see pages are missing), which shows imagery of the character of the environment: boarded up buildings, vacant lots, parking lots, broken or non-existent sidewalks, big bad roads, etc. etc. i.e. the kind of things that make walking feel or seem unsafe, unpleasurable, and undesirable.

Each study area gets three pages. The first shows the demographics in the study area. The second shows the real world ten-minute walking distances in relation to the theoretical ten-minute walk circles. The third page shows intersection density. It was my point that the areas became run-down because of the lack of real, quality public infrastructure in the way of an interconnected, dense network of streets and blocks.

See (and click to embiggen):

Yellow shows theoretical walking distance. Green shows actual 10-minute walks.

Blue dots show intersection density and on each page I show the calculation of intersections per square mile. Garrick's work shows that walkability, safety, and in turn desirability, really jump at around 225 intersections per square mile and up. All of these are much, much lower. Most of which are sub-100.

As a contrast, State-Thomas neighborhood, one of the best examples of neighborhood revitalization in the country after the area was gutted by S&L speculation in the 80s, has 253 intersections per square mile. It's average block size is about 350x325'. The Pearl District in Portland, another example of revitalization, has a typical block size that is even smaller, 225'x225'. In turn there is a very tight network of streets, yielding a highly walkable intersection density of 447(!).

Thus, we can begin putting together some predictions that any revitalization must be tempered and potentially unsustained if we don't significantly raise mobility and safety of areas through increased intersection/network density.