Corridor Zoning is not Transit-Oriented Development
Corridor zoning is a style of development, in which major arterial streets are zoned for high residential and commercial density, while side streets remain zoned for low density, sometimes even for single-family housing. Yesterday, a discussion on social media brought up transit-oriented development, so I’d like to explain why corridor zoning is not TOD, in intent or function.
For reference, here’s Yonge Street in Toronto:
And here’s Cambie, in Vancouver:
Both streets have subways running underneath them. Both streets – Yonge far more clearly than Cambie – are characterized by continuous high-density development along the corridor, only extending a block out from it. This is billed as protecting single-family neighborhoods from apartments while also providing opportunities for housing growth; Toronto has fairly YIMBY and Vancouver has very YIMBY housing growth rates, with housing shunted to such corridors and to redevelopable nonresidential sites like Marine Landing. In the United States, corridor zoning can also be found in Seattle, which has nearly the fastest apartment growth rate per capita in the country.
And this is not TOD. The reason is simple: TOD doesn’t care about corridors but about the underlying transit, which is a subway, not a bus stopping every block. From the perspective of a transit user, the walk from the Canada Line 300 meters north-south along Cambie is the same as the walk 300 meters diagonally along Cambie, King Edward, and side streets. Good TOD is therefore measured in terms of walk radius from the station.
Take the following example. TOD zoning might allow a floor area ratio of 3. This is roughly the floor area ratio of the central residential neighborhoods of Berlin, or the inner suburbs of Paris, or much of Upper Manhattan and inner Brooklyn. It can be realized with a six-story building, a common maximum height for North American woodframe construction, using half of a 30 meter deep lot. If these are built 500 meters from a train station, using the taxicab metric, then the resulting 0.5 km^2 area can host around 500,000*0.6*3/45 = 20,000 people at 45 m^2/person with the lots occupying 60% of the area and the streets 40%. This is a cluster that can support a few neighborhood-scale commercial functions, probably at or right next to the station.
The main benefit of corridors is not residential, but commercial. A residential gap in an otherwise commercial corridor will absolutely reduce spontaneity. But continuous commercial development is not achieved through corridor zoning – to the contrary, if the TOD density of FAR 3 can extend sufficiently far from the station, then that density can naturally be developed exclusively away from the corridor, while the corridor itself has the demand for one- and two-story commercial buildings, which can take some residential floors above them in theory but not in practice due to different development needs and different profitability rates. But that’s not corridor zoning – rather, it’s pre-automobile urban development styles, which had isotropic density across an entire neighborhood.
I don’t expect a TOD regime to produce Kreuzberg and Neukölln above. But I do expect it to produce considerable quantities of housing designed around train stations, and if those are modern stations then it’s likely they’re fairly far apart with low-density gaps between high-density pods (but, as much as possible, no residential gaps in commercial corridors). This is not really what’s going on with Canadian corridor zoning, because people can walk in multiple directions, and the explicit purpose of corridor zoning is to concentrate both air pollution and the perceived social pollution of apartment dwellers in quarantine zones that single-family homeowners can avoid. Good TOD practices will therefore drop corridor zoning, and allow much higher density within a radius of several blocks from streets like Yonge, even if it means some single-family homeowners will have to share blocks with mid- and high-rise apartment buildings.
Pedestrian Observations 
Yes, corridor development is just a high-rise version of strip-malls except with even less life. I don’t know either of those streets but any equivalent I have seen are pretty lifeless. Modern high-rise zones have trouble developing ground-level urbanity, despite the paradox of ostensible high density. One reason is that such zones today are entirely corporate and eschew the kind of owner-operated small businesses that bring interest and life rather than corporate-preferred tenants like banks, real-estate offices and fast-food franchises. Another reason is that the high-rise will have lots of parking in often multiple subterranean levels, and this means the residents tend to descend in elevators to their parking and drive away. On returning the same happens and it short-circuits any tendency to linger at shops and cafes etc or even do supermarket shopping locally rather than some giant drive-in mall.
This is driven by developers, NIMBYs and city politicians who want density (and to justify building the Metro) but are ignorant about what drives good urbanity.
One thing I cannot tell easily from the pics is what the distance between Metro stations is.
The interstation on Cambie is a mix of 800 m and 1.6 km, depending on cross streets; King Edward is 1.6 km from both adjacent stops (Broadway north, 41st south), but then the subsequent stops are not so far (Olympic Village north, 49th south).
As for your other points:
1. The corridors in Vancouver have cafes.
2. While Yonge is high-rise, the corridors in Vancouver actually aren’t – Cambie, Broadway, etc. are mid-rise; the high-rises in Vancouver are TOD at a radius from a SkyTrain station, and corridor zoning on the Canada Line was proposed as a supposedly more human-scale alternative to the high-rise TOD in East Van and Burnaby.
I think you’re right about high rise zones in the West, but in China I’ve seen high rise zones with good ground-level environments for pedestrians (not so much for cyclists, and not trams except for HK). E.g. big Chinese cities will usually have at least one pedestrian mall in a high rise area, with a decent amount of small to medium business presence as far as PRC goes (Chinese small businesses seem to look to expand rapidly if they are doing well, so these kinds of ped malls are often filled with a mix of small to large chains and some single location stores/restaurants).
The “Vancouver Podium” style skyscraper is actually conducive for a good street level urban form. It allows the skyscraper to be set back from the street, and a tighter, more enclosed feeling to be possible at street level with human scale architecture.
Unfortunately it seems that the actual reason for bad urban form is due to economic reasons, not strictly because of high rise or low rise buildings. Ground level commercial stalls must be small in area so rent isn’t too expensive for indie businesses. This is how you get vibrancy in a neighbourhood regardless if a building is high rise or not. This model seems untenable due to the high cost of construction at the moment.
Rosslyn-Ballston is what actual TOD looks like, with the highest densities where the potential for access is highest, with density decreasing away from the station in a concentric fashion. The “corridor” style high density only makes sense where stop spacing is low, like for a tram/brt line, but even then the high density can extend further out from the main road.
What are good examples of this kind of corridor transit density?
Looking at a couple BRT lines, Curitiba seems to be a good example while Medellín looks to have plenty of medium height density outside the BRT corridors. HK tram does have a lot of density along it, but that is probably due to geographical constraints.
I guess you could take a look at the old streetcar suburb pattern style of development. Unfortunately it seems a lot of the development patterns have changed since so it’s not really obvious anymore. I think there were streetcars along wisconsin ave and connecticut avenue in DC/Maryland and you can still see some evidence of higher density along the street even between the metro stops. However, the newer metro focused “target” style of TOD optimised for metro is now obvious too.
There’s also a bit in Tempe with their streetcar but it’s very minimal (it’s Arizona what do you expect). But given how much more auto-orientated our society is, I think this style of TOD is less important now because streetcars don’t compete with the automobile in speed. This sort of development pattern assumes that the bus/streetcar is your primary transit mode, which it shouldn’t to get good ridership. Ideally think the best situation would be the one in DC suburbs, with an old streetcar corridor higher density corridor overlayed with a newer development pattern focused around metro stops.
https://www.mta.info/map/5256
> Vancouver Podium
Could you point me to some good example locations? I took a look at Google Streetview and Downtown Vancouver + the Skytrain routes look to be quite sparse on street facing businesses compared to, say, downtown San Francisco let alone midtown NYC or a Hong Kong/Paris.
> Streetcar suburb
I’ve seen these in Washington DC, Los Angeles, Toronto, and other places in NA. Some of the ones I’ve seen are excellent for urbanism, some not so good; often in the same city/metro. Usually they at least have decent housing density, but several instances where the former commercial spine along the streetcar route now has stroad characteristics, e.g. Rockville Maryland.
Adirondacker, my impression of NYC is that many of the good urbanism corridors are along former streetcar routes, not just current metro routes. E.g. 125th Street, 10th Avenue, 9th Avenue are strong commercial corridors without subway lines.
Manhattan’s strong commercial corridors are by design. The grid was laid out when the far suburbs where in what we now call Greenwich Village.
The population of itty bitty tiny little metro Vancouver is the same as Brooklyn. They aspire to have Ocean Parkway’s parkwayishness. Or six track of subway like there is under the former streetcar lines of Flatbush Ave. Places that have less population density than Queens are going to look like Queens not Manhattan. Places that have the population size of Portland Oregon or Austin Texas are going to look like that not Central Park West.
Is there a term for real transit oriented zoning along frequent transit routes? e.g. a 3-7 min walk buffer along a frequent bus line that allows considerably higher density, and has parking maximums (not minimums)? It is hard to advocate for something if it does not have a name.
Yes, I know that this would mean upzoning the vast majority of some municipalities. However, there are lots of inner ring suburbs where this would mean upzoning half or less of the residential land with any decent definition of frequent transit.
Where I live, Greater Victoria, there are now apartment buildings being built in places with terrible transit service and incomplete sidewalk networks. We can do better.
In your Toronto picture, the stations are well under 1km apart, which means that the entire corridor is in the 500m walkshed radius and is fit for TOD. So what is needed is to extend the TOD zone outward from the corridor, not to eliminate the corridor, which itself is a good start.
In other parts of Toronto like Yonge/Eglinton, there is already TOD in all directions around the station, noit a corridor.
Also “if the TOD density of FAR 3 can extend sufficiently far from the station, then that density can naturally be developed exclusively away from the corridor, while the corridor itself has the demand for one- and two-story commercial buildings” – ironically this is exactly what they have done in this part of Toronto (https://maps.app.goo.gl/kJ3wpQi2BoreZY4X8)!
North York Centre is 800 m north of the next station south (Sheppard) and 1.2 km south of the next station north (Finch). Then going south of Sheppard, the interstation is 2 km until Eglinton.
The 800m gap is entirely within a 500m walkshed.
The 1.2km gap is almost entirely within a 500m walkshed, and entirely within a more reasonable 600m walkshed.
The 2km gap is not part of a TOD corridor. There is TOD within about 500m of Sheppard, then a long gap with a freeway interchange and golf course, then some more TOD around York Mills (not Eglinton).
Okay, but the adjacent blocks to Yonge are largely within a 500 m walkshed but, on the side opposite Yonge, have low-density housing.
What is it zoned for? Just because there are single family houses doesn’t mean developers are interested in redeveloping it.
Single-family houses. If it’s a 300 meter walk from the subway in Toronto or Vancouver, developers are very much interested in redeveloping it; the house prices in these areas are high.
The important point of Alon’s post, in my opinion , is “The main benefit of corridors is not residential, but commercial.”
However, he overestimates the population density of inner Brooklyn (nearest Manhattan). Only one ZIP code (11226) area reaches a density of 20,000 residents/km^2. The average for the 20 largest density ZIP codes is close to 18,000/km^2.
That having been said, this density supports lively commercial areas, especially near subway stations and along bus lines, which typically stop every 3 blocks (245 m). A problem is lack of convenient supermarkets.
11226…. off the top of my head there is a Ctown masquerading as SuperFresh on Flatbush between Dorchester and Ditmas….
Flatbush Ave and Nostrand Ave through 11226 have supermarkets and greengrocers on almost every block. Maybe you’re thinking of the single-family zoned parts to the SW of 11226?
The southwest corner of 11226 is the Newkirk stop on the Brighton Beach line. Where there is a C-town supermarket. On the east side of the right of way. And a Walgreens on the west side. And a clump of other retail. parallel to the right of way, styling itself as Newkirk Plaza. With what appears to be apartments over it. New Urbanist’s wet dream of railroad/trolley suburb. Wikipedia says they’ve had service to Manhattan – via the Fulton Street Elevated and Brooklyn Bridge – since 1896.
*they