Monday, March 25, 2013

The case for yield streets in New York City

Back in December I noted that there are actually three ways to adjust the balance of a road to favor pedestrians over cars, or vice versa. You can have expensive, permanent infrastructure: curbs, barriers and stop lights. You can have regulations marked with signs, paint and enforcement that can be changed relatively easily but take some time to get used to: one-way traffic flow, curbside parking and bus and bike lanes. You can have regulations that change with the time of day: reversible lanes, signal timing adjustments and rush hour parking restrictions.

Beginning in the 1950s, transportation planners rolled out more and more one-way streets. This approach has been discredited more and more over the years, and its Waterloo was probably Mike Primeggia's Park Slope one-way proposal that brought over 400 people to a community board meeting in 2007.

Pedestrian advocates here in New York may have stopped that one-way pair, but they have had less success reversing existing one-way pairs such as the Eighth Avenue-Prospect Park West pair that terrorizes park visitors. In other cities like Portland, Oregon and Albuquerque, New Mexico, however, one-way streets have been returned to two-way flow, increasing pedestrian safety and comfort.

We should examine restoring all our one-way pairs to two-way. Four of the streets on Tri-State's Most Dangerous Roads list are one-way: Amsterdam Avenue, Second Avenue, First Avenue, and Bedford Avenue. Other pairs that make for a lousy pedestrian environment include Crescent and Thirty-Third Streets in Astoria, Skillman and Forty-Third Avenues in Sunnyside, Court and Smith Streets in Cobble Hill, and Lafayette and DeKalb Avenues in Fort Greene.

The Manhattan avenues are roughly a hundred feet wide. The one-way avenue pairs that I mentioned above in Brooklyn and Queens are seventy feet wide, with two lanes of traffic and two parking lanes. But we need to go beyond those. The vast majority of one-way streets in New York City are fifty-foot streets with one lane of traffic and two parking lanes. It's harder to imagine those as one-way streets, but you remember this picture of West Tenth Street in Greenwich Village, from the time when it was two-way.


It's easy to make a fifty-foot street one-way, as was done on Fifty-Eighth Street right here in Woodside, by removing a parking lane. That goes counter to the hierarchy of street allocations that I posted earlier, because moving cars are worse for our goals than parked cars. But what if you kept the parking lane? Well, there's a precedent for that.


This is Saint Johns Road in Newport, a pretty town on the Isle of Wight off the coast of England. It's thirty-five feet wide, and it's two way with occasional parking on one side of the street. The car in the foreground is waiting for the light to change; the cars in the background are parked. What happens when cars are coming in opposite directions? One of them yields to the other. The parking spaces are distributed so that there is always room for one car to pull over. This is what Andres Duany, Elizabeth Plater-Zyberk and Jeff Speck call a "yield street."

You may be thinking, "gee, that sounds like an invitation to head-on collisions!" I didn't see a single head-on collision during my visit, and the UK Department for Transport says that "road layout" was a factor in only two to three percent of crashes in 2011 (PDF). Keep in mind that "road layout" includes bends and hills in addition to "narrow carriageway."

Why doesn't every driver on the Isle of Wight die in a fiery head-on collision as soon as they get on one of these roads? Because they are aware of the risk and adjust their driving accordingly, going slower and paying a lot more attention to what's around them. The same as with any traffic calming technique.

We should go beyond returning hundred-foot and seventy-foot avenues to two-way flow, and examine some of our fifty-foot streets. Any street in a neighborhood slow zone that sees a crash should be a candidate.

The biggest political obstacle to implementing yield streets is the need to remove some curbside parking spaces to make the yield zones. Some of our alternate-side-obsessed neighbors may not appreciate this. And that in turn points to the need to adequately price curbside parking in the city. Curbside parking is a privilege that should never have been given away for free.

Saturday, March 23, 2013

When is the MTA going to stop sending limiteds all day long?

Last night the 92nd Street Y hosted a debate for candidates for the Republican nomination for Mayor of New York. Members of the Y submitted questions to a camera crew on the sidewalk in front of the building, and the questions were then projected onto a screen behind the candidates. A lot of people seemed to be amused by a question posed by an Upper East Side resident, which you can see at 39:10 on the video.


My name is Selig Alpern. I've been living in Manhattan all my life. I am 89 years old. We have 4 to 5 limiteds flying down Lexington Avenue, and we have to wait 20 minutes for a local. From 10 o'clock in the morning to 4 o'clock in the afternoon there should be no limiteds. There should only be limiteds from 4 to 7 and 7 to 10 in the morning. When is the MTA going to stop sending limiteds all day long? They are not necessary from 10 in the morning to 4pm. It's a pet peeve of mine, and I wish it would be corrected.

Alpern was described as "ranting," "railing," "ornery," "upset," and "angry," but I think a lot of it was just him speaking loudly and clearly in order to be heard over the noise of Lexington Avenue. I thought his concern was valid, but his prescription was wrong, and then he undermined the whole thing and set himself up to be taken as a crank by calling it a "pet peeve." Transit frequency is not a pet peeve.

First of all, what does the MTA intend to do, as documented by the bus schedules? Looking at the noon hour, the M101 Limited is scheduled to pass Bloomingdale's at 12:08, 12:16, 12:24, 12:26, 12:32, 12:37, 12:47, 12:51 and 12:55 - nine buses an hour. The frequency is uneven, but the average headway is six minutes.

For the local buses, the M102 comes at 12:11, 12:31 and 12:51, and the M103 comes at 12:06, 12:21, 12:36 and 12:51. If you're not going below Eighth Street you can take either one, so I'll treat them as equivalent. That makes seven buses an hour, for an average frequency of eight minutes. One problem is that the headways are uneven, ranging from zero to fifteen minutes. If one of those buses is even a little bit late, it could lead to exactly the situation that Alpern complained about: being stuck for twenty minutes at a local stop watching one limited after another fly by.

We should definitely not eliminate midday limited service. How does it look for a Jewish guy from East 68th Street to tell the MTA to add ten minutes to every limited bus trip going from Harlem downtown? But it would be appropriate to tweak the schedule and run one or two buses on the local route instead of the limited route.

Better yet, who's to say that we have to get rid of buses on one route in order to add buses to another? Why not add more M102 and M103 buses without getting rid of M101s?

Some of this depends on the "loading guidelines." You don't want to run too many empty buses. That said, you could run regular buses instead of articulated ("bendy") buses if you want higher frequency.

Steve "BicyclesOnly" Vaccaro tweeted that he cared about, "What are candidate's position on BRT?" And that's an excellent question. Why aren't there bus lanes on Lexington Avenue?

Alpern's question highlighted a major problem in New York today: the despondent tone of so much of transit advocacy. Why do so many people think it's a zero-sum game, where we've given over Lexington Avenue to the cars, and Jews and Puerto Ricans have to fight each other over the scraps of bus service that crawl along wherever they can fit in?

The candidates' responses to that question, and the Democrats' responses to another transit question, were revealing. None of them mentioned "BRT," or even buzzwordless bus improvements. But George McDonald did say he looks forward to the Second Avenue Subway opening, and predicted that it would relieve crowding on the Lexington Avenue buses. No word if he would do anything as mayor to help get the rest of the line built, or even finish Phase 1, but at least it was hopeful and forward-looking.

Monday, March 11, 2013

The difference a lane (or toll) makes

Stephen Smith has a great piece in the Observer showing how the Bloomberg administration has adapted its development strategy to avoid clashing with long-term residents, many of whom live near subway stations. The resulting disconnect between housing capacity and transit has encouraged high-rise towers near the Williamsburg waterfront, far from the overcrowded Bedford Avenue L station and even further from the Marcy Avenue J/M/Z station.

In the long term, NIMBY fears of "overdevelopment" are primarily based on fears of competition for scarce car space. To overcome it, we need to allow developers to build without encouraging people to keep cars, and we need to as convince the long-term residents that parking spillover is not as likely as they think, especially now when everyone is driving less.

In the short term, many of these highrises are already built, and more are in the works, so we need to improve their transit access to keep their residents driving as little as possible. Everyone wants to run ferries because hey, waterfront, but the ferries require large subsidies to compete with free bridges. It's hard to build subway stations right by the waterfront: the fabled South Fourth Street station is still five blocks from the river at Driggs Avenue. You could build an elevated subway or a streetcar, if you could convince people to build something that hasn't been built in this city in at least eighty years. You could also run buses.

Four years ago I mapped out some possible destinations that a Midtown Tunnel bus could get to from Penn Station in ten or fifteen minutes without traffic. The idea was that the tolls keep traffic from getting too congested in the Midtown Tunnel, and the 34th Street bus lane would give the buses priority. The way the 34th Street bus lanes have been executed leaves a lot to be desired, but the principle is sound.



What if we put an Exclusive Bus Lane on the Williamsburg Bridge and down Delancey and Chrystie/Forsyth streets to the Grand Street station? We could run the Nostrand Avenue Select Bus across it, and the Chinatown vans would be significantly faster.

But you could also run buses down Broadway to the waterfront. I agree with Jarrett Walker that loops are generally a bad idea, but Kent Avenue is now one-way, so you could have a South Kent Avenue loop going down to Flushing Avenue, and a North Kent loop going up to Greenpoint. In ten minutes (without traffic) from Grand Street in Manhattan, you could get to Morgan Avenue in East Williamsburg, either along the BQE or along Grand Street. You could go down Broadway to Flushing Avenue, or down Bedford and up Nostrand as far as Myrtle Avenue. And in fifteen minutes you could go further.



This is no longer serving existing high-rises, but rather improving service to long-established neighborhoods. That in turn could lead to people in those neighborhoods no longer thinking that cars are the only route out of poverty. But its usefulness lies in the dedicated lane for buses on the Wiliamsburg Bridge. That's what allows these buses to be competitive with driving.