Jeff Speck has a great new book, Walkable City Rules, which is a prescriptive follow up to his terrific Walkable City. As the title Walkable City Rules suggests, the new book is a set of crisp, clear rules for achieving a healthy, walkable city. He’s been tweeting one rule every day or so since May. He recently tweeted what may be my favorite (certainly as it applies to Newton): Rule 50.
Here’s RULE 50: BUILD SLOW-FLOW AND YIELD-FLOW STREETS, in my new book, Walkable City Rules. I’m tweeting the whole book out May – July, one rule a day. Enjoy! Or go to: https://t.co/LVNgGNusqo pic.twitter.com/MCb9ickk48
— Jeff Speck (@JeffSpeckAICP) June 28, 2019
Speck conveniently provides the full text of Rule 50 in the tweet. Basically, it says to build narrow streets in residential neighborhoods.
The insight of Rule 50 is that residential streets don’t need to be designed to allow continuous side-by-side travel. In fact, designing streets to require traffic traveling in opposite directions to slow is a feature, not a bug.
Speck calls such streets yield or “yield-flow” streets, where there is a single, shared travel lane of 12′ or so. Cars coming in opposite directions have to yield, usually by one ducking into a parking lane. We actually have yield streets all over Newton, though the yield-ness of a particular street may depend on the parking at any given time. Behind my home is Walter St. It’s about 28′-wide. When cars are parked on either side — or, better yet, both — it is uncomfortable to impossible for two cars to pass without slowing significantly or stopping.
What possibly could Rule 50 have to offer Newton? We’re not building new residential streets. And, given the prohibitive cost, we’re not comprehensively rebuilding them either. Sure. But, press a little on Rule 50 and it provides the necessity and logic for evaluating our residential streets. Press a little harder and it provides guidance on how to make our streets safer and more livable.
While Speck’s framing of this rule suggests that it applies to building (or rebuilding) roads, it doesn’t require much of a leap to apply the rule to evaluate existing roads. If your road has peak volumes below 150 cars per hour, your road, according to Speck, should be 26′ (with parking on both sides) or 20′ (parking on one side). If it’s an existing road with volume below 150 cars per hour and it’s not 26’/20′, it’s too wide!
Applying such clear, objective criteria to our existing roads flips the script on how we think about investing in safety and livability. We shouldn’t have to force neighborhoods to submit an application for complete streets traffic calming. We shouldn’t put the burden on neighborhoods to make the case that their streets are dangerous and need some sort of remediation. By taking a tape measure to the street and counting volumes for a week, we can determine if a road is poorly designed. And, poorly designed roads should just go on a list to get fixed.
There are volumes and volumes on how to design roads and intersections properly, but there’s very little specific criteria to identify roads and intersections that are non-compliant with modern safety standards and should, therefore, be automatically considered for remediation. While he’s not doing it directly, Speck is providing very clear criteria to identify bad residential roads, of which we have tons in Newton. This is profoundly important.
Speck’s rule, while apparently a guide for building (or rebuilding roads), also gives us guidance on how to remediate our bad roads without rebuilding. The key is that Speck’s objective is not design for design’s sake, but changing behaviors. His insight is that the design of yield streets forces desirable behavior: cars slowing down. We can create the behavior of yield street, even if we can’t feasibility convert all of our residential streets to full-length yield streets. Arguably, you can build yield streets by building a little yield into each street. For residential streets with peak volume less than 150 cars-per-hour it’s probably enough to create a single section per block with a 12′, bi-directional shared travel lane. There are lots of ways to do that. Create mid-block bump-outs, which would also have the virtue of protecting the parking lane. As is done on Lexington Ave. in Cambridge, retain the existing width, but add bike lanes to reduce the width for cars to only 12′ for a shared, bi-directional travel lane. Put in bollards. Dig greenswales to mitigate stormwater runoff.
Newton has a bunch of 32-36′ residential streets, most of which appear to have been built in the middle of last century. They were designed to accommodate both good-sized parking lanes and side-by-side, uninterrupted two-way travel. They were, intentionally or not, designed for speed. To make matters worse, the homes on these mid-century streets often have two- or three-car garages and plenty of driveway, so cars don’t need to — and don’t — park on the street much. These streets are wide and largely empty. They are like residential freeways.
As residential roads are scheduled for repaving, they should be measured and any with peak volumes below 150 cars per hour and lacking a yield pinch-point should be identified as non-compliant. Those non-compliant roads should have a yield moment built, as a matter of course.
Measure. Evaluate. Fix.