Density and Subways, revisited

Since my last lengthy post went up, a number of defenders of subways in general, and the Sheppard line in particular, have raised some objections to my argument that Toronto’s inner suburbs (Etobicoke, North York, Scarborough) don’t have the required population densities to support subways. Some objections have been from the most vociferous of Mayor Rob Ford’s supporters, and haven’t really engaged in the facts in any way. Others have been more thoughtful and deserve a real response. So here goes!

You say Toronto’s suburbs aren’t as dense as New York’s boroughs, but the Big Apple’s boroughs weren’t always this dense. Isn’t a fairer comparison between New York when it started building its suburbs, and the inner suburbs today?

I’m not actually sure this comparison will be any more accurate than my previous one, considering the number of moving targets that we’re actually talking about here. For one, in real terms subway construction was substantially cheaper back in an era when you didn’t have to pay a minimum wage. Nevertheless, I whipped up this chart of the densities of New York’s boroughs over the last century from Wikipedia data, with the megacity’s 2006 average (which overstates the average density of Toronto’s suburbs).

As we can see, all of New York’s boroughs have been denser than Toronto is today for a very long time, except for Staten Island (which isn’t connected to the main subway system, relying on surface rail instead). The first New York subways were built in the years before 1910, so this gives us a century worth of data. The other thing that happened after the New York subway was built was the introduction of New York’s first comprehensive zoning law, which had the visible effect of radically decreasing density in Manhattan from 1920-1980, which is nevertheless still very, very dense.

But this leads us to a second question, one raised by Dave Meslin in particular, which is this: Even if the inner suburbs aren’t dense enough to support subways now, can’t subways encourage the very density that would make them financially sustainable?

To which the short answer is: yes, very much. And you can see that happening in Brooklyn and the Bronx in the early decades after subways started making their way through New York. But there’s a problem: we don’t live in the 1920s or 30s anymore, which is mostly a very good thing (we don’t have to worry about prohibition or fascism!) but for subway advocates poses a few real problems. The long answer to Dave’s question ends up being… well, no.

Costs

Subways cost a lot of money, and they cost more—in inflation-adjusted terms—than they used to. When Toronto built the Bloor-Danforth line, it did so at the same time as it built the University line, joining Union to St. George and building the east-west subway line from Keele to Woodbine. I can’t find an exact number for the construction costs, but this site says the estimate was $200 million in 1960 dollars. Plug that in to our Bank of Canada inflation calculator, and that’s roughly $1.5 billion in 2011 dollars for about 15 kilometers of subway. If we had to build that today using the conservative $200 million that some optimists project for subways, it would be twice as expensive to build the Bloor-Danforth line in 2011 as it was in 1966. And if it went to the more likely $300 million per kilometer, we’d be talking $4.5 billion to tunnel under the downtown, or triple the cost of the original. (Which leads to the depressing conclusion that if we had to do it again today, we probably couldn’t.)

My point is that Toronto mostly built its existing subway system when it was relatively cheap to do so, in the context of a rapdily-growing postwar economy, with all the fiscal largesse that implied. Neither of those two conditions is likely to repeat itself in the Toronto area soon. So in fact, the business case for subways in the suburbs in 2011 is even more difficult, even if they were as dense as the downtown core.

And they’re not, they’re really really not, and they’re unlikely to get as dense as the core anytime soon.

Why? Because despite nominal commitments, this city is very hostile to density.

Toronto’s planning process, and how it works against transit

When I say that this city is very hostile to density, a lot of people look at me like I’ve grown a second head. “Look at the Places to Grow Act! Look at Toronto’s Official Plan! Plenty of density there!” (Yes, I’m the kind of person who actually brings this up in conversation.) But Toronto, like almost all North American cities, has a planning system that was never intended to allow density—or, for that matter, much change at all.

So for example, it’s true that there’s a Sheppard Avenue secondary plan that allows mixed-use density along parts of Sheppard. It’s also true that the same plan forces buildings to be set back from the sidewalk in case the city decides it wants to add a seventh (!) lane to Sheppard. (That alone should make subway advocates question how serious the city is about encouraging transit.) But the building heights envisioned in these plans are still laughably small (in the 5-7 storey range) but more importantly, the secondary plan disappears once you go any distance off Sheppard.

This, for example, is how the city envisions the intersection of Sheppard and Willowdale:

In case you haven’t wasted your life learning city zoning prefixes, what this map shows is that going north 400 metres from Sheppard, you’re allowed to build a mixed-use, commercial/residential building until you hit Hollywood. Then it’s all RD, which stands for residential-detached. Can you guess what is an allowable use of land in an RD zone? From Chapter 10.20.20.10 the city’s handy by-law guide:

The following principal uses are permitted in an RD zone:

• Park
• Dwelling Unit, if it is located in a Permitted Building Type in Clause 10.20.20.40.

If the suspense hasn’t killed you yet, the Permitted Building Type is “detached house.” It specifically forbids semi-detached homes, apartment buildings, townhouses, and, well, anything other than the kind of houses that already exist there.

This isn’t an anomaly. Zoning systems were invented to put the brakes on development, and Toronto’s does it very well.

The long and short of it is that, as currently structured, it would be illegal to add the density subways require along the Sheppard corridor.

Ah, but developers can appeal the official plan and zoning, right? Of course they can. But let’s be clear about what this is: the city will, if a developer pays enough money and drags the city to the OMB, and wins, allow density to increase. There are many words for this system, but “density-friendly” isn’t one of them. So yes, the system as structured is hostile to density. It’s intended to be.

Density does not equal height (alone)

But surely the condo boom along Sheppard is adding enough density to make a difference, right? Well, it will certainly add more people and that won’t necessarily hurt. But (a) relatively few of these condo-dwellers will be taking the subway (there’s also the 401 nearby, drawing commuters to their cars) and (b) Sheppard is unlikely to support enough condo towers, especially at the 7-storey maximum councillors prefer, to make the subway a winning proposition.

But that’s actually not a deal-breaker, because the far, far better way to add density to the area is not to build a canyon of condo towers but to intensify the much larger area of residential streets north and sourth of Sheppard. This is a real bonanza: just by allowing three-storey townhomes, or semi-detached duplexes, density could be added cheaply, quickly, and without radically altering the nature of a suburban residential street.

So that’s a winner, right? Wrong. We’ve already seen that the existing zoning forbids anything but detached homes, and when it comes to planning appeals townhomes on residential streets are at least as controversial in Toronto as towers on main streets.

To pick one real example without having to try very hard to do research: this file that’s going before the North York Community Council is a good example. I pick it not because it’s perfect or because the developer is noble and virtuous, but because it arrived in my email inbox this week and frankly, there’s always one example on any committee agenda I could point to.

It’s not even a suburb story, which shows that this is a city-wide problem, not at all an issue of narrow-minded suburbanites: a developer finds two homes north of St. Clair that have unusually deep back yards. He buys the lots, and wants to turn it in to a row of 18, three-storey townhouses. The area is zoned to allow townhouses, so there’s no problem, right?

Wrong. City staff are concerned that the building style doesn’t conform with the existing neighbourhood of a) two-storey semi-detached homes and b) five-storey apartment buildings. Also, the fact that the townhouses won’t face on to Winona street is an issue. From the city’s perspective, the only acceptable development is to.. replace the buildings with exactly the same kind of building.

For a row of three-storey townhomes, the developer will have to go through a community consultation and a city staff report that they estimate will be ready in “the third quarter of 2012 provided any required information is submitted by the applicant in a timely manner.” If he’s truly lucky he won’t have to drag the city to the OMB, costing more time and money, before getting permission for this development.

I could go on, but maybe one more anecdote will suffice: a few months back, Shelley Carroll got up in front of city council and said she’d never been lobbied so hard to release a hold in all her years at council. What item was she holding? An OMB appeal in Josh Matlow’s ward where the city wanted to try and stop a developer from turning two detached homes in to three townhomes. (The specific issue, if I recall correctly, was the developers’ preference for below-grade parking.) Shelley Carroll is, of course, a former budget chief so when she talks about intense lobbying we should perk up our ears.

Now all of this is a real shame, because most of the social benefits to density don’t come from adding 40-storey condo towers. The most rapid increase in the benefits in density come from the first 3-10 storeys (depending on which planning literature you read.) So adding more condo towers downtown is a game of chasing diminishing returns—but the residential areas of the suburbs offer an incredible amount of low-hanging fruit, if only we were willing to pick it. But the legal and political facts are that we’re totally willing to let this fruit wither and rot.

Again: maybe you like this system. Maybe the developer is a dick. And who really wants a system with less public input? But whatever you think about it, the system is hostile to density at every turn and it should trouble anyone hoping for rapid increases in density to support subways.

Imagining suburbs where subways worked: they wouldn’t be suburbs

What would subway-sustaining density look like? Well, for one, it would extend away from Sheppard a great deal further. The attitude of many suburban residents seems to be that any density added along Sheppard needs to be quarantined there, and kept away from streets north of it (with minimal exceptions, like the one I pointed out on Willowdale). I’m not saying that as a mind reader, but as an evaluation of the councillors (like David Shiner) that they keep electing.

But this isn’t how subway-sustaining density works. You need really large areas of moderate density, not a valley of towers on Sheppard and then sprawl elsewhere. Look at the areas downtown where the subway travels through, especially from Pape to Keele: three-, four-, five-storey buildings along Bloor and Danforth are the norm, with apartment buildings found frequently on the streets running north-south from Bloor, and walkup apartments common even on smaller streets like Annette in the west end. Meanwhile, the houses are vastly more likely to be semi-detached or townhouses, and the detached houses are much more likely to be duplexes. The “catchment area” for the Bloor-Danforth line is vastly larger than just the buildings that front on to those streets—and it’s this kind of density that is expressly forbidden in the city’s planning along Sheppard.

When the TTC talks about the density needed to sustain subways, they don’t talk about people per hectare—they talk about people and/or jobs per hectare. What planners understand is that if you’re going to balance a transit system (that is, if it’s not just going to be full trains running east, and empty trains running west) you need to have a balance of people and employment spread through the region. The TTC, according to Steve Munro (who would know, after all) uses a rule of thumb of 100 people and/or jobs per hectare to support subways, which is an appealingly round number. Using the city wards as a unit of measure again, it allows us to judge them on their subway-fitness by adding the people and jobs in a ward and seeing how they measure up.

So here is subway-fitness index for the three wards currently served by the Sheppard subway:

Ward 23: 52 people per hectare + 28.5 jobs = 80. But Ward 20 was already growing rapidly from 2001-2006 (for which the latest data is available, sadly) so I’m actually going to give this one a presumptive 100. Anyone who’s seen Yonge and Sheppard can see a pretty vibrant centre there. But a single node does not a subway line make.

Ward 24: 32 people and 16 jobs per hectare, without the double-digit growth of Ward 23. Total score: 48. #subwayfail

Ward 33: 52 people and 27 jobs per hectare, and Shelley Carroll’s ward actually shrunk between 2001-2006. Score: 79.

So you’ve got basically one ward that qualifies as subway-sustainable today, and two that rather dramatically do not.

And extending the Sheppard line east in to Scarborough? Ward 40: 50 people and 24 jobs per hectare, predictably less dense than Ward 33.

(Obviously, these numbers will have changed with the 2011 census. It will be interesting to see what they do, and I’ll be very happy if the business case for Sheppard has improved. I suspect it will still be a dubious case, and still won’t support an extension past Victoria Park.)

Now a lot of people have pointed to the development along Sheppard and accused me of ignoring the rapid growth there. I swear I’m not, but numbers like those for Willowdale show that the density would have to double since 2006, and I guarantee you that it hasn’t. Ward 33 may be closer to the 100 mark than it was in 2006, but I doubt it’s added 40% of the Ward population in five years.

I’ve been accused of ignoring major development projects along Sheppard, but let me turn that around: ignore the visible, high-profile signs of increasing density and ask yourself: have something like 100,000 people moved in the those wards without anyone noticing? Are they all taking the subway and not driving? As big as those condo towers are, I guarantee you it’s a fraction of the needed population. (And if the 2011 census proves me wrong, than find me to point and laugh. Beer’s on me.)

By contrast, Wards 20 and 27 score 122 and 138. Even Ward 31, a relatively low-density ward in East York (and where I live) scores 88. (Better than the best along the Sheppard line, according to 2006 data.) And to illustrate my point about how important a large area of density is, Ward 17 (Davenport, entirely off of Bloor) scores in at 111. Ana Bailao’s Ward 18 is even denser. That is to say, based on these numbers there’s a stronger argument for a Dufferin subway, from the CNE to Eglinton, right now, today than there is for any extension of Sheppard, or even for the existing subway.

You’ll note that large areas in the wards I’m talking about aren’t solely the condo-tower clusters of the very downtown. Again, height does not equal density. But you can look at the long stretches of St. Clair, Dufferin, and Dundas West where there’s tons of density but, until very recently, very few condos (and still only limited subway access.) But they’re also light-years from what the suburbs are: mixes of housing stock, employment, and some already-existing transit options like streetcars that the density grew up around decades ago.

So my argument so far:

• The Sheppard corridor is not currently dense enough to support the subway it has.
• It could hypothetically be made more dense, without destroying the community, but:
• The zoning doesn’t currently allow sufficient density of the right type;
• the right type of density looks more like Ward 17 and 18 than like CityPlace;
• the existing appeals process makes changing the zoning slow and expensive;
• and actually adding sufficient density of the right type is likely to be politically implausible at best.

Now, it’s possible this will change. It’s possible that the city will, over the objections of residents on the cul-de-sacs north of Sheppard, allow developers to come in and build townhouses and walk-up apartments. And were it to happen that area could attract a lot of business, adding for the business case for more mixed-use along Sheppard. These things are all possible. I remain skeptical that they will actually happen. The prospects in other corridors in the suburbs (like Finch) are even more bleak.

The final issue is whether people are willing to subsidize further suburban subways more than they already do. The government can waste a lot of money if the voters support it for long enough (witness the only-recent death of the ethanol subsidy in the US.) It’s imaginable that Rob Ford could, in 2014, ask the people of Toronto to support a major property tax increase (around 10-20%) specifically earmarked for a Toronto Subway Fund. But that’s the level of committment that would be required, and I just don’t see Rob Ford doing that. Moreover, the same suburban voters demanding subways also, with distressing constancy, elect tax-cutting conservatives. To say “the suburbs want subways”, you also have to admit “the suburbs want low taxes.” Which is to say, “the suburbs want the impossible.”

And if the province or feds want to pony up? Well my preference would be for investment in the areas of the city that could already support, and desperately need better transit (Queen street, and another north-south line somewhere like Dufferin). But if they’re willing to fund a Sheppard extension and not something downtown, it should come only after the city has agreed to the kinds of major increases in density along the corridor, written into provincial law, so Ottawa or Queen’s Park don’t end up permanently subsidizing a money-loser. But again, I don’t see that happening along Sheppard. It’s far more palatable for the other levels of government to extend the existing subways in to the 905—hence the Spadina extension to Vaughan (which, in ridership terms, is just as dubious as the Sheppard line) or the long-rumoured Yonge extension up in to Thornhill.

So we’re left with: subways in Toronto’s suburbs can’t pay their way currently, are unlikely to do so in the medium-term future, and other levels of governments are unlikely to help. (Except in small, incremental ways.) The city could choose to support them with a massive tax increase, but neither voters nor politicians seem in the mood.

None of this argument, by the way, is because I’m personally opposed to Rob Ford’s politics. I’m just as opposed to the Spadina extension, but Dalton McGuinty didn’t ask my opinion. What I’ve tried to do, in this post and the last, is show that there simply isn’t the business case for subways in the ‘burbs and in many ways there cannot be one, unless Toronto’s politics change radically.