Is There a "One-Way Epidemic"?

Recently, City Limits ran an op-ed by architect John Massengale about the need for safer street design. Under the photo at the top, the caption read: "First Avenue in Manhattan. Avenues used to run two-way, which is safer for pedestrians, but were mostly made one-way, to make life easier for drivers." This argument was elaborated in detail in the text. This week, Cap'n Transit followed up on his blog, extending the campaign against one-way streets.

These are familiar arguments. After beginning a discussion a while ago about one-way streets, it was suggested that I refer to "The One-Way Epidemic" section in Walkable City, which makes these same claims. These pieces all serve as good examples for discussion purposes. I wholeheartedly agree with the need to redesign New York City's streets to be safer for pedestrians, and I share the majority of the views of these safe-street advocates. I find some areas of common ground regarding one-way streets, as well, but it is useful to draw out the key differences as well as the overlap to illustrate why the rhetoric against one-way streets is overblown and counterproductive.

Let's start with the term "epidemic" used in Walkable City. This is rhetoric that invokes fear. The book explicitly compares the creation of one-way streets with an outbreak of influenza that killed 20 million people in 1918-1919. The not-so-subtle claim is that one-way streets are an imminent threat to your life. Avoid them like the plague!

Yet the notion there is an additional risk factor inherent from operating streets in a one-way configuration has rather weak support from any actual research. The complete range of available studies offer some mixed findings, with many indicating that one-way streets are probably safer for pedestrians than two-way streets under most, but not all, conditions. As previously discussed, the strongly-held belief among many urbanists that one-way streets are dangerous usually relies on somewhat specious logic.

Before going any further, I want to make sure one fundamental point is perfectly clear: my argument is that one-way streets are a useful tool that should not be dismissed without evaluating alternatives.  I am not arguing that one-way streets are inherently "better" than two-way streets any more than I would argue that a hammer is better than a screwdriver. Saying a hammer is a useful tool doesn't mean you ought to start hammering in your screws. We need to improve the overall discussion both to empower planners to use a complete set of tools, as well as to improve public participation in our planning.

Like almost every general argument I see against one-way streets, both the City Limits op-ed and "The One-Way Epidemic" rely on the oversimplified and questionable premise that streets were converted to one-way operations because engineers decades ago chose to speed traffic through cities at the expense of pedestrians. That is not really true. Improving traffic flow was a significant goal, yet during the period when most streets were converted to one-way operations, there was an explicit consideration of pedestrian safety and a belief that the changes would improve the plight of citizens on foot as well. Perhaps that belief, and the studies on which it was based, have turned out to be incorrect; that would warrant discussion. Instead, what we get is only an unsupported assertion - one that is false - that one-way streets were created solely for cars by people who viewed pedestrians only as obstacles for vehicles. As a narrative, it works well enough, but the implied conclusion that reclaiming streets for pedestrians requires reversing the change from two-way to one-way operations is unfounded.

The other recurring point involves misunderstandings about speed and ways it can be controlled. Massengale's claim that one-way streets are inherently less safe because they make cars go faster is a prime example.  He states:

In fact, making cars go faster is always worse for pedestrians, for three reasons: 1) a pedestrian hit by a car going 30 miles per hour is eight times as likely to die as a pedestrian hit by a car going 20 miles per hour, and a pedestrian hit by a car going 40 miles per hour is 17 times as likely to die; 2) a driver going 20 miles per hour sees three times as much as a driver going 40 miles per hour; and 3) a driver going 20 miles per hour has more time to react and can stop in approximately one-third the distance of a driver going 40 miles per hour.

The facts about speed are basically correct, and this is the way arguments against one-way streets are invariably constructed. Unfortunately, the use of these basic facts is oversimplified and deeply flawed. The connection between one-way streets and speed is misunderstood, and other factors affecting the risk of a collision are overlooked.

To understand "fast" one-way streets, we need to disentangle a few different issues. First, it is important to distinguish between point speeds (which are what matter for survivability in a collision), and travel time. One-way streets employ signal progression, which minimizes the amount of time drivers spend waiting at red lights. Vehicles travel "faster" by reducing stopped time, not by increasing their speed at any point on the street.

Another important issue is whether speed is an inherent characteristic of one-way streets, or something that can be controlled. This depends on the type of street we're talking about. If we're talking about a suburban couplet with long distances between signal lights, the one-way configuration almost certainly can induce higher speeds. There are compelling arguments about the friction of two-way streets slowing drivers on this type of arterial street (although it certainly is not the only means for reducing speed). When talking about dense urban street grids with regular signal lights, on the other hand, speeds can be well regulated with signal timing. When engineers wanted arterials to move fast, they used a signal progression set for a high speed limit. By the same token, to emphasize safety, they can set a slow signal progression to calm traffic. Often, signal timing to calm traffic is more malleable on one-way streets than on two-way streets. There is great potential for things like a "green wave" on cycling streets, which keeps cyclists moving instead of making them stop frequently for red lights. It is unfortunate that people who obviously care about making streets safer, better places don't do a better job trying to understand the great work being done by transportation professionals before condemning them.

The final factor that gets overlooked when discussing speed is that a primary factor of risk is actually getting hit by a car. It is more dangerous when the car is going faster, but you can't ignore other factors that affect the basic likelihood of being struck. Two-way streets create more complex intersections with more conflict points. The fact of the matter is that the New York City Department of Transportation has found that most pedestrians killed or severely injured were struck at intersections, and at highly disproportionate rates on major two-way streets. These statistics simply don't reflect the contention that one-way traffic patterns on avenues are an imminent threat to safety, or that they would be improved by changing them to two-way operations.

Walkable City makes other claims beyond safety.  There is the"economic" argument. The evidence here is no more convincing. "The One-Way Epidemic" relies heavily on an unpublished study in Savannah conducted not by an economist, but rather an architect. I have twice requested a copy of the study from the City of Savannah, but they have not provided a copy. Needless to say, this is not a very compelling source.

It may be that for smaller cities it turns out the way proponents of two-way streets suggest. Nevertheless, the irony is inescapable when a book about "walkability" insists that the visibility of storefronts from passing vehicles is the critical factor for a vibrant downtown business. If your downtown has to rely on motorists getting a good view of the stores as they drive by, you don't have a walkable city.

The other major source for the claims about retail health, including "The One-Way Epidemic," comes from anecdotal information from Vancouver, Washington, an auto-oriented suburb of Portland. As a small municipality, there were limited resources for a systematic, in-depth study of the effects of the two-way conversion. Even without a real evaluation, the economic health of the downtown after the change seems clearly positive. Yet it is completely questionable how much of the success is due to the two-way traffic pattern rather than the full set of improvements that were made at the same time (general pedestrian treatments, etc.). It is a challenging question how the same businesses would have performed with the same enhancements on improved one-way streets.

While the evidence is rather flimsy, it has one compelling aspect. In this and the handful of other publicized cases, the actual outcome of a vibrant business district was consistent with the theory that two-way streets are supportive. This suggests that the two-way treatment may have been beneficial, or at the very least was not too harmful. Given the available information, it would seem wise to continue exploring the details around the improvements that have been done in smaller, more auto-oriented cities. I find the notion of extending this to the dense urban cores of larger cities quite dubious.

The final argument, made in both "the One-Way Expidemic" and on Cap'n Transit is about network efficiency, which seems particularly inapplicable to larger, more vibrant cities that are actually walkable. Especially in New York, nobody expects to park feet from the front door. If you are compelled to drive for whatever reason, the search for parking starts at least a block before your destination. It baffles me that the same people who love to quote Donald Shoup's figure that 30% of cars in congested areas are circling waiting for a parking space to open would realize that this would minimize the impact from the slightly indirect circulation on a grid of one-way streets.

If we set aside the issue of cruising for parking, then there is a tradeoff: one-way street networks are more efficient for longer trips, but two-way networks serve shorter trips more efficiently. So we have to ask if the tradeoff between short and long trips is desirable. Aren't shorter trips precisely those that are best served by walking, cycling, or transit? Surely a walkable city would put a preference on alternative modes for shorter trips. On the other hand, there is a question about the sprawl-inducing potential of facilitating longer driving trips. While the sprawl concern is valid, encouraging more driving trips within the city to reduce the capacity available for suburbanites surely is not the answer. Rather, the remedy is to curtail the ease of driving to the extent reasonable to maintain access for deliveries and other trips that lack reasonable alternatives to driving.

Of course, my previous post was criticized, as this one is likely to be:

@urbanresidue @victordover Your outlook ignores the fact that the 1-waying of 2-way systems has harmed 100s of places. Reverse not so.

— Jeff Speck (@JeffSpeckAICP) December 23, 2013

The assertion that one-way streets have harmed hundreds of places is questionable. The notion that there aren't streets that have not been harmed by bi-directional operations is even more dubious.

The specific quote may be nominally correct up to a point; there have been a somewhat limited number of cases of one-way to two-way conversions, and they have been implemented relatively recently. There has not been decades of experience with those cases, although it is fair to note that early experience has not yet raised significant red flags. Still, it is likely enough that there have been problems that have not been evaluated and documented yet. The issues pointed out by critics of one-way streets were not immediately obvious to the traffic engineers and downtown business interests who continued converting streets decades ago, either.

More significantly, the underlying notion that two-way operations have not harmed streets is most certainly incorrect. Whenever a two-way to one-way conversion was successful, it can be said the street was formerly "harmed" when it was a two-way street. Even the critics of one-way streets grudgingly recognize that some one-way streets work well.

My reluctance to accept this "everybody knows" urbanism is not entirely unique. Peter Calthorpe, for example, has noted, "There's this sad oversimplification going on where some people are just saying, 'One-way streets are bad." Calthorpe has experience using one-way streets to make San Elijo Hills Village Center more pedestrian friendly. "'If you're trading a six-lane arterial for two three-lane streets, that's a win," said Mr. Calthorpe. Nevertheless, he said the concept of one-ways remains 'heresy' among modern urban planners."
http://news.nationalpost.com/2012/01/08/taking-a-u-turn-on-the-one-way-street/

A familiar trope in discussions with one-way advocates is their reliance on comparing improved two-way streets to one-way streets designed decades ago. The resistance to an even-handed comparison of safe street improvements with different one-way or two-way operations can be dogmatic and detrimental to achieving the best results in each individual case.

Compelling proof of the value of one-way streets lies in the work of those places that have been most innovative in creating livable streets, places like Portland and New York City. They continue to convert two-way streets to one-way operations today, when the evaluation of options indicates that the change would support the goals of more livable streets for all users.

Kent Avenue in Brooklyn is an example of a recent two-way to one-way conversion. When a bicycle lane was first added to Kent Avenue, New York City DOT maintained the two-way operation on the street. However, as they continued to evaluate operations, it became apparent that the configuration was too limiting for access to the local businesses and did not provide as much safety for cyclists as possible. Trucks that needed to make deliveries resorted to violating the bicycle lane. This was unsatisfactory for the businesses making and receiving the deliveries (due to ticketing and hostility from passing cyclists), as well as the cyclists who were endangered by the illegal parking that forced them into traffic. By converting Kent Avenue to one-way operation for motor vehicles, it was possible to restore access for the local businesses while creating a two-way cycle track that provided much better protection.

NYCDOT document on the conversion from one-way to two-way operations on Kent Avenue:
http://www.nyc.gov/html/dot/downloads/pdf/kent_ave.pdf

Google Street View over the years:

2007


2009


2013

2014


Another potential example is South 4th Street in Brooklyn. Here, a wide stretch of street that encouraged speeding was narrowed, providing more space for pedestrians and ultimately a bike share station. This street could have been converted from one-way to two-way instead of creating the new pedestrian space. It's not clear what benefit would be created by converting this street to two-way operation. Pedestrians would not only have to navigate intersections with more conflicts, they would also have a much longer crossing distance. The value of the new pedestrian space is an open question, but it has proven quite useful for adding a Citibike station, which can be a highly contentious process given the competition for the limited space within New York's crowded streets.

South 4th Street improvement:
http://www.streetsblog.org/wp-content/uploads/2012/12/2012-12-20-borinquen-bqe.pdf

Before

After

Let me share some personal experience from The Bronx, as well. The home where I now live with my young son is on a narrow one-way street on a long block. I find crossing mid-block with him to be safer, because there are no turning conflicts and I have a long view of any traffic that is coming from a single direction. Because of the long stretch, lack of opposing traffic, and uncoordinated signal timing, there were problems with speeding until recently, especially as drivers tried to beat the light. With some advocacy, a little tenacity, and an unfortunate, photogenic crash, we were able to get a speed hump installed. The speed hump has been quite effective, without forcing me to find gaps in two-way traffic to get across the street with my son. When I return home on my bicycle, I have to circle the block on the one-way pairs, which I find little real inconvenience. I certainly prefer it to the idea of being forced into the mix between aggressive drivers on a "yield" street.

The other experience deals with the extension of the one-way pairs across Gun Hill Road. Until recently, the avenues remained two-way on the north side. I have noted increased safety while crossing the street as a pedestrian, especially at Decatur. Where there were two heavy flows of turning vehicles cutting across the crosswalk at the same time, I now only have to deal with one set of turning vehicles. Moreover, because they're not competing with the other turning vehicles for scarce gaps on a short block, I find the drivers now pay more attention to the pedestrians in the crosswalk. Gun Hill Road is still far from an acceptable pedestrian street, but the conversion of the two-way streets has made a noticeable improvement in my quality of life.

Some of the fierce critics of one-way streets do acknowledge that there are places where they work, but characterize them as "exceptions." Dismissing successful streets as "exceptions" is dangerous, because it creates a risk of missing opportunities to learn from their positive qualities. When planners swoop into a community with the predetermined view that its one-way streets are the root of its problems, they are likely to miss other issues and may recommend two-way conversions when improvements to the one-way streets would be more effective.

There is one point in Cap'n Transit's blog post that does stand out: New York City does not seem to have converted any one-way streets to two-way operation. As we have seen, there are many cases where one-way operation is more suitable, and the New York City Department of Transportation has made many good improvements that entailed converting two-way streets to one-way. But it does seem curious that they haven't found any cases that broke the other direction. I often casually wondered from time to time if some of the one-way streets like Rider Avenue and Canal Place in the South Bronx might be better as two-way streets. Perhaps the New York City Department of Transportation is missing an important tool.

Planners should be encouraged to consider a full range of treatments when they prepare their alternatives for evaluation. This must include improvements that include one-way configurations. I would hope that even their fiercest opponents could agree there is no harm in considering the relative performance of different alternatives; if their claims are correct, an evaluation of the alternatives should bear them out.

One-way streets have their place, and the rhetoric against them that has become popular among many architects and planners does nothing to help improve the quality of our cities. Creating a climate that is hostile toward a particular solution only makes it harder for planners to use that tool where it is the appropriate treatment. We do not have an epidemic of one-way streets, only an epidemic of rhetoric that recognizes only one way of designing streets.