Thursday, November 17, 2011

Job sprawl and the multipolar city

Back in 2009 I had an interesting discussion with Alon Levy, Jarrett Walker, Jeff Wood and Damien Newton about transit and multipolar cities. Jarrett, in particular, highlighted the efficiency benefits of multipolar layouts:
If you want a really balanced and efficient public transit system, nothing is better than multiple high-rise centers all around the edge, with lower-rise density in the middle, because that pattern yields an intense but entirely two-way pattern of demand. If balanced and efficient transit were the main goal in Los Angeles planning, you'd focus your high-rise growth energies on multiple centers such as Westwood, Warner Center, Burbank, Glendale and perhaps new centers in the east and south, while continuing to add density in the middle as opportunities arise.

It's expensive to have your jobs centralized in a small area and your housing scattered around it. It results in large crowds at rush hour and empty trains and buses in the reverse peak, or else large railyards and/or bus garages in the middle of the city. In a multipolar city there are jobs and residences everywhere, so that in the rush hour there is a greater use of the transit system but it is not highly unbalanced in one direction. In Paris, for example, you have people who live in Nanterre and Noisy-le-grand commuting in to work at the Bourse or one of the government offices, but at the same time you have people who live in the Marais and the Latin Quarter commuting out to La Défense and Montreuil.

In a multipolar city like that, you want circumferential routes, not just radial ones, because it's more efficient for someone who lives in Saint-Ouen and works in Nanterre to go directly west instead of going south into the city first. New York could stand to be more multipolar, and David Giles has pointed to job sprawl as evidence that of such multipolarity emerging. In our current transportation network, circumferential highways like I-287, I-278 and the Belt Parkway are much more functional than any similar transit routes. Given this, David argues that job sprawl favors drivers over transit users. He argues that we should build "bus rapid transit" to even out this disparity.

The thing is that there's a big difference between a multipolar city and sprawl. Multipolar cities have, you know, poles. Of higher density, near major transit hubs. Sprawl has highway interchanges, but the density is still fairly uniform. You can get off the train at La Défense and walk to your job, but you can´t do that for most of the jobs in the White Plains area: they're too far from the station. Because of this, job sprawl forces most people to drive and is much less efficient than a multipolar city.

David did consult me and other smart people for his report, and on the whole I like the way it turned out. But I disagree with his ultimate conclusions, especially the emphasis on "bus rapid transit." There are two problems with that. The first is that buses are always more expensive to operate, per passenger, than trains. The second is that it won't ultimately solve the problem of job sprawl. That second point deserves a whole post, so look for one soon!

13 comments:

Matthew said...

Boston has a similar problem with radial transit. There has been rumblings of an "Urban ring" line for decades, but so far only buses are available.

Notably the MBTA created several rush hour buses to serve this transit pattern: CT1, CT2, and CT3. These are relatively close to the core. Moving slightly further out, the 66 bus covers a wide swathe of territory between the Red Line and the "Silver Line", also crossing every Green Line branch. It has become (as of 2010) the most popular bus route by weekday ridership.

However, many of the jobs are much further out still, near the Rt 128 beltway, a badly sprawled area. I know from friends that the traffic jams out there can get pretty bad, especially if you need to drive some portion around the ring.

giles said...

This is a nice discussion. And thanks for referencing my report. I think the big point in the report was that multipolarity already exists in NYC in terms of jobs (and job growth) but the transit component is missing in a lot if not most instances. So who would have thought that 25,000 people commute into East Flatbush every day? It was built quite obviously as a residential neighborhood, but as the three hospitals there have grown up, vertically and economically, the area has begun to pull in commuters from all over the city, Staten Island, Long Island, South Brooklyn, Northeastern Queens... I take it as a given that a new subway is not in the offing in East Flatbush; dollar vans and black cars could perhaps be leveraged more efficiently (people, especially patients are already using them); but the most obvious solution is ramping up bus service, dramatically, and creating connections with the LIRR on Atlantic ave, the subways at Atlantic terminal and other BRT lines moving east/west across the center of the borough. If DOT and MTA had been smarter they would have created a network of BRT lines that dramatically reduce commute times into job centers like this one. Flushing is another huge job- and commercial center with woefully inadequate transit. Jamaica and JFK are up there as well. Hunts Point, Maspeth, College Point, and Staten Island's north shore need massive amounts of help. And so on. It's interesting how little these growing polarities have played into the planning and communications strategy for SBS. And so in East Flatbush and other areas all the attention at the community level is paid to the merchants abutting the proposed SBS routes, while the larger needs of the community as a whole go unheeded. Obviously i don't think creating BRT by itself will halt or dramatically slow down exurban growth without some corresponding land-use policies in exurban regions, but it will most definitely create massive growth opportunities in the polarities themselves. SUNY downstate folks told me quite explicitly that several satellite projects, including a biotech facility were being endangered by commuting issues.

Unknown said...

I love this concept. On a baseline level, the central city concept works and makes sense. But as a city grows, coming up with multiple poles as you describe them is a great way to go.

It actually gives me hope for LA, if they can get development patterns in check. The transit plan they're currently building out would actually serve such a model pretty well. And they already have centers like Burbank, Hollywood, Culver City, Century City, etc that are poles.

Back here in NYC, we also have poles, we just need to bolster the transit between them so you don't always need to go into Manhattan to get between them.

Joseph said...

" buses are always more expensive to operate, per passenger, than trains"

That's only true if the trains are full, and if you count by passenger-mile. Here in the Portland area, we have FRA-compliant diesel train between several suburbs, which costs several times as much to operate, per passenger-mile, compared to the average bus.

In Los Angeles, the BRT orange line costs about $0.60 per passenger mile (last time I checked), a little more expensive than light rail ($0.50), but cheaper than the local-stop buses, and much cheaper than the regional Metrolink trains. In Los Angeles, a light rail vehicle costs almost $300 per vehicle-mile to operate, including maintenance costs and salaries, versus about $100 per vehicle-mile per bus, so you need at least 3 times as many people on the average train car to make it work.

I'm sure that in New York, the operations costs of subways are lower than the buses, due to faster speeds and higher capacity. But is that even true of the regional rail system, like Metro North and the LIRR? Considering the ridiculous pension and disability fraud that has been going on at the LIRR, I wouldn't be surprised if the total costs per passenger mile are higher for those trains than for a city bus.

Cap'n Transit said...

Thanks for the comments. David, I think it's important to not take it as a given that new train lines will not be built. As long as the road builders feel comfortable talking about widening the BQE and building giant new tunnels to Westchester, we need to keep talking about new train lines. The Tribororx is one that would serve the area you describe for very low construction costs.

Brandon said...

The comparison between Paris and Los Angeles is missing one thing: scale. This is particularly true when talking about suburban job growth in New York. Multipolarity within New York City would be fantastic. Build up some of the outer borough mini-cores to Parisian density (7 story standard), but avoid more jobs in the 'burbs. New York has less job sprawl than almost anywhere else in America, and this is a very good thing. Job sprawl is one of the most insidious forces out there.

George K said...

@Brandon:

I think you should expand the definition of "outer boroughs" to include suburbs built up along rail lines in the region. That way, not only do you have increased reverse-commuting, but people in the suburbs are also able to access jobs closer to home.

For instance, if Hicksville became a larger job hub, people out in Ronkonkoma would be able to commute there instead of going all the way to say, Jamaica, Queens.

Cap'n Transit said...

Joseph, what I should have said is that all other things being equal, buses are more expensive than trains. The main thing, actually, is that maintenance of the right-of-way is usually not counted as part of the capital budget for bus lines, but it is for rail lines.

You make a good point about the insane FRA regulations, but honestly, if transit advocates put as much energy into overturning those regs as they do into bloviating about "BRT," the regs would be gone and it would be no contest between buses and trains.

giles said...

The discussion you reference at the beginning of this post is interesting and obviously important. But it also has an extremely long time horizon--30 years at least--which is not to say that people shouldn't be discussing it, just that our arguments occupy very different spheres. Something approaching a BRT system could be implemented in 3 or 4 years, if the political will were there. Every single SBS line proposed to date can and probably will be implemented using DOT/MTA's current capital budget, even amid huge shortfalls. And not only that but they could afford to do a lot more than they've proposed. Queens, for example, has been taken off the table not for technical or financial reasons but political ones. This is the situation our report was designed to address and most of our discussions with political leaders and advocates coming out of it have centered on how to mobilize support and tap constituencies that stand to benefit from the changes (i.e. CBOs, unions, business groups,etc.). That said, I'll definitely continue to educate myself about opportunities for rail enhancements.

giles said...

Also, the Triboro RX doesn't come nearly close enough to East Flatbush and hospitals to make much of a difference there. A really cool proposal, however, and it wouldn't have hurt to mention a few more of these in our analysis of commuter patterns.

Alon said...

Triboro comes exactly to Brooklyn College, though. I don't know about the hospitals, but the college is both a major source of employment and the busiest station in Brooklyn outside Downtown Brooklyn.

Also, if the FRA were reformed, the cost of implementing and running regional rail would not be much higher than that of running a bus.

George K said...

@Alon: I think that honor goes to the Utica Avenue station on the (3)/(4) trains (which is also in that general area). Both stations are one of the first subway connections you reach when you're coming from those subway-less areas of southern Brooklyn.

Alon said...

@George K: yes, you're right. I'm sorry. Utica is much higher-ridership - 8.5 vs. 6.3 million. This is what I get for not double-checking. And Bedford on the L has 7.4 million. But Flatbush is still the third busiest off-downtown Brooklyn station, and the sixth busiest in the borough.