Thursday, July 19, 2012

The missed opportunities of the Northern Branch

I had to admit that New Jersey Transit isn't planning to pave paradise for the parking lots on the Erie Railroad Northern Branch. They want to take over some existing parking lots, build on top of some others, and tear down some warehouses that are unremarkable and kind of shabby. It still makes me unhappy, though, because of what they're not planning to build there: walkable suburban downtowns.

There are a string of dense, walkable towns lining the western slopes of the Palisades: North Bergen, Fairview, Ridgefield, Palisades Park, Leonia, Englewood Tenafly, Demarest, Closter, Norwood and Northvale. They didn't happen by accident, nor did they spring fully formed from the head of some central planner. They grew up around the Northern Branch, and the New York Central's West Shore Line, and the trolley lines of the North Hudson County Railway. But after the opening of Route 46, and Route 4, and the New Jersey Turnpike, and Route 80, and the Palisades Parkway, it got harder and harder for the trains and trolleys to compete.

Now New Jersey Transit is planning to bring back some of the railroads. They're not talking about restoring passenger service on the West Shore, or rebuilding the North Hudson, but they are talking about restoring service on the Northern Branch, in the form of a Hudson Bergen Light Rail extension to Tenafly. As I've mentioned before, I'm concerned about the amount of parking. One reason is that I think it will encourage local driving and lead to more sprawl.

Another reason that the parking bothers me is the opportunity that is being missed. While Englewood, Tenafly, Cresskill, Demarest and Closter all grew up around the Northern Branch, the centers of North Bergen, Fairview, Ridgefield, Palisades Park and Leonia were all along Broad Avenue or even further east on Bergen Boulevard or Anderson/Bergenline Avenue. What developed along the Northern Branch in these towns was largely industrial. Since then much of the industry has been replaced with commercial development, but it has been sprawling, car-oriented development.

If we're not going to bring back the Broad Street trolley, but we are going to restore passenger service, people will be traveling from compact, walkable downtowns to rapid transit through a sprawling, car-oriented mixture of industrial and commercial buildings. You can see why many of those who have cars would prefer to drive.

If I were designing this, in the tradition of Jane Jacobs and Andres Duany, I would first rezone the areas around the stations for dense, pedestrian-oriented mixed-use development, no parking required. I would then look at likely commute paths from the downtowns to the new stations, including 91st Street in North Bergen, Edgewater Avenue in Fairview, Ruby Avenue in Palisades Park and Fort Lee Road in Leonia. I would widen the sidewalks on those roads, run shuttle bus service from the downtown to meet each train both ways, and rezone the properties on either side for the same kind of mixed-use development. For each of these towns, this would essentially allow a secondary downtown to grow around the new train station, extending it and the old downtown to meet each other.

Such a plan would allow these towns to return to their old transit-oriented pasts. Instead, New Jersey Transit has decided to lead these towns further down the car-oriented dead end, by building more parking and making it easier to drive and hard to walk. No wonder the leaders of these towns are upset. The planners didn't even try.


Matthew said...

The Northern Branch passes right through the town centers of Englewood and Tenafly. It also passes through the original town center of Leonia, but that has long ago shifted to Broad Ave. But that's just a short walk away.

These towns grew up as streetcar suburbs, and nobody would mistake them for anything more, but they do still fill up NJTransit buses every day. And there's no park-n-rides for those buses. People walk to those buses, and I don't see why they wouldn't walk to the train for better access to the ferry, Jersey City, and Lower Manhattan. If the "planners" don't get in the way.

busplanner said...

I fully agree that the municipalities along the Northern Branch should look to redeveloping the areas around the stations. And many of the stations have decent bus service from surrounding residential areas.

But I think you misunderstand the purpose of the parking at Route 4 and the end of the line. Most of northeastern Bergen County has no good public transit to the Jersey City/Hoboken Gold Coast or to the ferries to lower Manhattan. One has to head west to the Pascack Valley Line to find the first decent service to Hoboken. And the buses to midtown (to connect to subways to lower Manhattan) are maxed out at peak times due to Lincoln Tunnel and PABT capacity constraints. Accordingly, there is a very high "drive to work" number from northeast Bergen County to these Hudson County/Port Imperial ferry destinations. I believe the parking lots are designed to capture some of this traffic, decreasing auto VMTs.

Cap'n Transit said...

Busplanner, I've heard that reasoning, but as I've written before, it's stupid. short-sighted thinking to focus on replacing two twenty-mile daily driving trips with two-mile driving trips, while continuing to subsidize all the other driving trips throughout the week.

Cap'n Transit said...

For example, if the Northern Branch extension was built without parking, I would imagine a lot of the commuters would take the buses to Tenafly and transfer to the light rail. That would leave more room for Midtown passengers who want to get on in Tenafly.

BruceMcF said...

The big flaw with too many park and ride commuter lines is the presumption of one size fits all.

Indeed, in your sketch for towns that were streetcar suburbs for another alignment, there is likely to be room for some parking a block away from the railway station on the side opposite the existing downtown. It can, indeed, serve as pooled parking a no-mandatory-parking zone around the station.

Zmapper said...

A few places where park 'n' rides can be useful:

- Base of a very low density populated mountain. Generally, it impossible to provide decent service within walking distance to the residents, and nearly everyone will own a car, so the only real feasible way to run service would be with a centralized park 'n' ride at the base.

- When a line-haul route intersects a Freeway. Such an arrangement allows for the freeway to increase the catchment area of the line-haul route, without congesting city streets. The Washington Metro has even built park 'n' rides with dedicated off-ramps from the freeway, so traffic doesn't need to even stop for traffic lights and the like.

I am not that familiar with the Northern Branch line except for a cursory Google Earth look, but it appears that neither of these examples above apply, except for the Rt 4 station.

Alon said...

The Northern Branch suffers from one critical flaw: it doesn't get people to Manhattan. Trying to build commercial TOD in Jersey City - which means transit access in multiple directions, including north - is intriguing, but the main walk-to-bus ridership in the area is toward Manhattan.

busplanner said...

@Cap'n - You have long been critical of park-rides; but if you look at the area north and north/northwest of the proposed end of the line, you will see a primarily built out area with residential density of five or fewer residences per acre. This density (per Jarrett Walker) does not usually support transit.

While I do not have current information, a now outdated license plate survey of autos driving to the Port Imperial ferry and the Hudson Waterfront offices showed a sufficient number of addresses from this large area, if the drivers could be focused on one transit route. Thus, there appears to be some merit to at least a few park-rides on the Northern Branch route.

busplanner said...

@Alon - There is a tremendous amount of Class A office space along the Hudson River in New Jersey served by Hudson-Bergen Light Rail. I once consulted on a company expanding from Manhattan to Jersey City and a large number of their employees resided in this corridor. I believe there will be more than adequate demand for peak period service. (Whether there is sufficient demand for reasonably frequent off-peak service is another question.)

Cap'n Transit said...

Yes, I've walked through many of those towns. They're a lot denser and more walkable than you'd think if you just look at the numbers. Passenger service should go all the way to Nyack.

I respect Jarrett, but he doesn't know it all. The idea of "density to support transit" is pure bullshit.

NJ Transit could get a lot more riders by convincing the towns to allow new walkable residential development than by chasing after people who would rather be driving down the Turnpike.

Zmapper said...

Density is one of the challenging issues to overcome when expanding transit to the suburbs. Generally, East Coast suburbs are less dense than equivalent Midwest or West Coast suburbs.

For instance, compare the suburbs
of Denver, Colorado to those of the East Coast. Notice how surprisingly urban the Denver suburb is. Density is roughly 6-10 houses per acre, and there aren't many permanent dead-end streets. When the commercial areas along 104th are built out, people could safely walk to go shopping. Such a community could easily support additional transit service.

New York on the other hand has very low-density suburbs, on the order of 1-4 houses per acre. Worse yet, because the main roads were historically cow paths, the ROW is constrained such that there is little space for bike lanes or sidewalks. Whereas main arterials out here have 100-150 ft wide rights-of-way to use, providing ample space for every user, suburbs along the East Coast have to cram everything into 60-80 ft wide streets.

Cap'n Transit said...

What? No, Zmapper. Please read the post I linked to above. Repeat after me: you don't need density to have transit. You don't need density to have transit.

I don't know where you get your information from, but the Bergen County suburbs north of Tenafly are all pretty dense and walkable.

Your interpretation is off, too. You don't need wide streets for walkability. You can have it with 36 foot wide streets, and it feels much more pedestrian-friendly.

Zmapper said...

Yes, in the case of the Northern Branch the suburbs are fairly dense and pedestrian friendly, I agree. What I was trying to get at is how the newest East Coast suburbs have rather low densities, lower than suburbs in the Midwest and West built during the same year. The typical reason given for this involves water usage; East Coast cities have water pretty much everywhere and can afford to have large lawns, while West Coast residents don't have that much water that just falls from the sky, and therefore have to pipe it in. When you have to use expensive pipes to deliver water, you have a very good reason to limit the amount of water that you use. Denver's suburbs aren't even the densest; I believe someone ran the calculation a while back that the newest development on the far reaches of Las Vegas has a higher density than the Richmond district of San Francisco.

As an example, most of the development in Rockford County isn't that dense, despite being rather new. Further complicating matters, the right-of-way that the government owns is rather narrow, meaning that you can squeeze in barely 2 traffic lanes, but then you really don't have that much remaining space for bike lanes, parking, or sidewalks. While narrow roads would be ideal, people like to get where they are going within a reasonable amount of time, thus leading to the idea of primary and secondary roads. If you tell people that they have to drive 10 mph in order to accommodate the relative few who want to walk down the middle of the street, don't be surprised when they overwhelmingly vote you out of office come November.

Just to clarify, I don't mean that I necessarily support what Charles Marohn calls STROADS. What I mean is that given the current choice of a 2-lane road lacking sidewalks, or a 4-lane road with sidewalks set behind a 10' tree median, the 4-lane road is more pedestrian friendly. If people don't feel safe walking to the bus stop, they aren't going to take the bus.

What correlation that exists density and transit viability holds true, but is rather weak. As an extreme example, you couldn't run a bus every 10 minutes to a 1000 acre ranch in Montana and expect to get any riders. Likewise, Lower Manhattan is capable of supporting a dozen subway lines, almost all of which run at least every 5 minutes during the peak.

But most of America doesn't live in rural Montana or Lower Manhattan, they live somewhere in between. At these densities typical of suburbia, transit ridership can vary widely, but is mostly dependent on the service quality provided. You could run crap bus service in a large city and get few riders, or you could run very good bus service in a small city and get many riders.

I read your post, and have to agree with the part about Wyoming. While I-80 is debatable, there was and still is no real need for I-25 North of Cheyenne to be a full interstate. Only 5000 vehicles daily use the road, which is less than many 2-lane streets. Typically, the state DOTs like to upgrade a rural road to 4-lanes at 10,000 ADT, which means that trips on I-25 would need to literally double in order for the road to even pass the very lenient demand standards of today. A 2-lane modern highway would be more than plenty for the Casper-Cheyenne corridor.

In the specific case of Casper-Cheyenne, the rail line is too curvy and congested with coal traffic to make passenger rail travel competitive with coaches. The same holds true for Cheyenne-Laramie, another relatively busy corridor. Union Pacific runs at least 100 mile-long freight trains over Sherman Hill with just 2 tracks, meaning that there isn't that much spare capacity to insert useful passenger trains. While it may seem like a good idea at first to move people between cities by train, the reality is that the lines are already congested with freight trains. Any effort to expand intercity passenger service would require additional track capacity, thus negating the environmental and fiscal benefits saved from switching modes.

Alon said...

Sure, in principle you can run trains in completely unpopulated areas. But if you want people to ride them, you'll want to focus on thicker markets.

busplanner said...

@Cap'n - I neither agree nor disagree with all of Jarrett's or all of your postings; however, to deny the importance of density as Jarrett and I use it ignores the experiences both of us have had in designing transit routes.

I have designed transit routes for both service expansion and purely political reasons. And, in general, those routes serving areas with low residential unit numbers per acre do not meet any sort of financial or ridership goal, whether feeder routes or longer through routes. On the other hand, park and rides to serve a trunk route for these areas are often successful.

The above statement does not deny that there are some walkable areas in the communities beyond the proposed terminus of the Northern Branch service. It also does not deny the value of mixed use TOD developments at proposed station sites. And it does not deny that there are a number of existing bus routes that could feed the service, including in Tenafly and Englewood. Park/rides add to the total value of the investment.

busplanner said...

@Cap'n - You have suggested extending the Hudson-Bergen Northern Branch extension to Nyack. There are many possible reasons why NJ Transit has chosen not to do so. (I do not know to what extent each of these led to the decision to propose to end the route in Tenafly.)

1. Right-of-way issues
2. Lack of interest from MTA and/or Rockland County
3. Cost, especially with respect to net increase in ridership.
4. Too little projected ridership to meet guidelines for federal aid.
5. Opposition from New Jersey communities along the route. (Note that Tenafly has gone on record as opposing the line and it may need to be cut back to
Route 4.)
6. Fear of destroying/legal issues associated with competing with Rockland Coach/Coach USA, the private bus carrier that provides almost all service to the north and northeast of Tenafly.

Cap'n Transit said...

Busplanner, I'm not denying that you experienced something when you were designing those routes. I just don't think it was density. Density is just a distraction from the real issue: competition from facilities for private cars.

And yes, I'm sure there are challenges to reactivating passenger service all the way to Nyack. But it's nothing compared with maintaining or replacing the Tappan Zee Bridge.

Alon, you too? My point is that the thickness of the transit market is not a simple function of density.

busplanner said...

@Cap'n -

I agree that the thickness of the transit market is not simply a function of density. Age profile of the population, income levels, and, most importantly, number of people traveling in the same direction at the same time, are some of the other key factors.

But northern/northeastern Bergen County has a number of factors that make bus feeder service to a station-oriented service (whether BRT, light rail, or heavy rail difficult to sustain.

And in those situations, appropriately sized park-rides make sense, ideally as carefully planned parts of TOD developments at the station sites.

Cap'n Transit said...

No, Busplanner, it's not age or income, it's competition from car facilities. But maybe your salary depends upon your not understanding it.

busplanner said...

@Cap'n - Maybe you don't understand it. And my salary has absolutely nothing to do with it.

A public or private transportation company cannot support a service if there are insufficient passengers to use that service. And that applies to both local public transit and longer distance public transit.

Travel patterns in suburban areas are different than travel patterns in more urban (more population dense) areas.

If you have one person traveling per day per acre of land, that one person cannot support public transit (at a reasonable cost) directly from her/his door to wherever he/she is going. If, however, you can capture that person for part of the journey, you accomplish the following:

1. Improve the financial condition of the transit that person uses.

2. Perhaps, due to increasing ridership in a chicken and egg fashion, encourage increased frequencies/span on that transit which encourages more ridership which encourages more increases in frequency/span, etc.

3. Decrease auto vmt and pollution. Improve air quality.

4. Decrease pressure to widen roads used previously by those now taking public transit.

5. Reduce maintenance costs of those roads.

6. Increase political support for improved transit and decrease political support for road expansion.

These are worthy goals. They may not meet your "ideal world" view; but they are attainable improvements to the current state.

Cap'n Transit said...

Busplanner, if your salary has absolutely nothing to do with it, then why don't you actually read the damn thing and respond to it, instead of repeating the same patronizing platitudes over and over again?

Zmapper said...

The Public Transit Users Alliance in Melbourne has calculated what the minimum density should be in order to sustain a bus every 10 minutes, assuming $45 per hour operations cost and that 20% of the population makes one round trip journey by transit. As it turns out, the minimum density required is about 3500 people per square mile.

Cities in Northern New Jersey has a density of between 3000 and 9000, according to Wikipedia. Those cities most likely could support a feeder bus system, though in order to come close to an acceptable farebox recovery ratio the drivers would have to be paid much less. According to the NTDP, it costs NJ Transit $145 per vehicle hour. In comparison, Foothill Transit near Los Angeles, which is completely privatized save for the board staff, operates buses for $86 per vehicle hour.

On that note, does anyone have statistics on cost per vehicle hour for bus companies in the UK? 80-90% of bus routes in the UK are commercially operated and companies are expected to profit.

Cap'n Transit said...

Yes, did you read this part, Zmapper? "Of course, the crucial assumption here is that 20 per cent of people make one trip by public transport each day"

Zmapper said...

I did read that part, Capn. 20% is a very reasonable and realistic percentage to aim for in suburban areas. Remember, you have to crawl before you walk before you run.

Cap'n Transit said...

No, you're not getting it. The point is that there are other variables besides density involved. They're very clear on that.

Chris G. said...

the thing is that englewood actually already has the shuttle and a downtown that is way more walkable then drive-able... The other towns in the valley should build around making englewood/ fort lee the town center for the whole region... the route 4 stop is surrounded by factories and some of them are abandoned that can be new housing and jobs for all and help boost englewood