Saturday, June 28, 2014

Is the building height debate mistaken? Should width, not height be the focus of the human scale?

There are few subjects that evoke so much passion in urbanist circles as building height. Some urbanists point out that height is necessary to add housing supply to built-out cities and to reduce pressure on housing prices, while others say that high buildings are basically an abomination against the human scale, with the most ardent defenders of this point of view outright saying that no building over 4 or 6 stories should be allowed.

One thing that plays against tall buildings is the absolute architectural and urbanist failure of the "towers in a park" concept, popularized by modernists in the mid-20th century, and picked up by many public housing projects on both sides of the Iron Curtain (but as most housing in communist countries were public housing, they were extremely common, hence the name "commieblock").
Towers in a park, in a former communist country
"Neighborhood" in Anjou, a borough of Montréal: a variant of the towers in a park, towers in a parking lot... proof that even a terrible idea can get much, much worse
These developments absolutely destroy the urban fabric and in their attempt to combine the advantages of green suburbs and dense cities, they instead combined the disadvantages of both. All the problems of density without the advantages. Walking from one place to the other is long, uninteresting and unpleasant. There is no connection to anything. The result to many of these developments was poverty, isolation, crime and other social ills of the kind.

Experiments like this turned many urbanists against the concept of towers altogether. Except...

The problem with the "towers in a park" concept wasn't with the "towers" part, but with the "in a park" part.

Take a classic "towers in a park" concept, replace these towers with 6-story buildings, or even 4-story buildings, and I think they would look just as bad. In fact, these places exist in many suburbs around the world, but here is one case around Stockholm that exemplifies this well:
Low-rise apartment buildings in a park, Stockholm suburb
Same area seen from the sky
Another example, this time in Montréal, I almost rented out this place before I decided for an apartment in a multiplex, a choice I do not regret
Thinking about this issue, I've had a thought: maybe height is not really what's important in terms of the human scale. Maybe the real important factor is building width, not height.

Width, the unspoken factor

There is a lot of discussion about height, but really little about building width. The two may be conflated somewhat because tall buildings tend to also be wide buildings, at least in North America. And building width may not seem important in the case of many European cities where buildings are attached, forming extremely wide "walls" of buildings on either side of the street, even though individual buildings tend to be generally narrow.

Paris, though the shared walls and uniform height gives at first an impression of an unbroken building more than 100 meters long, each "wall" is made up of 13- to 15-meter wide individual buildings
The people who argue against height tend to mention the human scale, that higher stories are isolated from the street by height, people on the street cannot visually identify people at balconies at more than 4 or 5 stories. That's true, but not everyone likes being near to the street when in their homes, some people do like having their own bubbles, it's quite easy to allow balconies on the first few floors, then setbacks for higher floors for people who want more isolation in their private home. It's better than pushing those people to suburbs to get the same privacy.

I personally think that what makes the human scale is the same thing that makes an area walkable. And what do you need for walking to be useful, comfortable and pleasant? You need density and mixed uses so that things are close to people. You need to have many things to see around you, variety makes every walk an unique experience. One way to have variety is to focus on details, not shape. The car-scale is often associated with very wide buildings and large empty spaces, but when you get close to the buildings, there is no attention at all to details, because no one is within close distance of the buildings except on their way in or out. You also need to give people the impression that they're going fast, and one way of doing that is to make sure that people's immediate environment frequently changes.

The importance of width here should be evident. The human scale is a point of view 1,5 to 1,9 meter-high and going along at 5 km/h (around 1,5 meter per second, 5 feet per second). So if a building is 10-meter wide, it takes only 6 seconds to cross it, at which point one's environment changes, you were in front of one building, now you're in front of another one. But if a building is 60-meter wide (200 feet), it takes you the most part of a minute to cross it (40 seconds). It takes quite a bit of time for the environment to change around you, you have the impression of being quite slow.

Blank walls: a specialty of wide buildings

It's important for buildings to have interesting features to keep people interested during a walk. And nothing is more interesting to people than other people. In that regards, I think doors and balconies are great, because they offer the promise of human activities. Theoretically, windows could do the same, but in reality, they're not that great as, especially in dense areas, residents will tend to make them opaque to preserve their privacy.

Shuttered windows in Tours, France
Now, the big problem with wide buildings is that they tend to have only one door, or maybe two, and to otherwise offer one long blank wall the rest of the side of the building. A blank wall is a wall without any feature or evidence of human activity, it is a plight on cities as it gives people no reason to be in the street and creates a boring, sometimes even dangerous, streetscape.

Blank wall in Stockholm
This is an example from Stockholm, an example taken not far from the first images of apartment buildings in park I've shown above. They essentially took the same building, but put it next to the street instead of lost in a park. Despite correcting this flagrant flaw of putting buildings far from the street, facing each other instead of the street, the result is not that much better, because the wall at eye-level to pedestrians is a featureless concrete wall with a door every 30 meters or so (100 feet).

Meanwhile, let's look at the opposite of this blank wall, triplexes in Verdun, a neighborhood of Montréal.
Multiplexes in Verdun: narrow buildings with lots of doors and balconies, and small setbacks
Anyway, the point is that wide buildings, even in the "urbanist-approved" low-rise apartment blocks, can result in blank walls that hurt the walkability of areas by making walking less pleasant and less comfortable.

Skinny towers: a special feature of Japanese cities

Up to now, I've shown a lot of examples of what not to do, about wide buildings that keep walkability down, but here is now the radical alternative: narrow but tall buildings.
Skinny towers in Ginza
Buildings in Gotanda, a neighborhood of Tokyo


Also Yokohama

A part of Yokohama seen from Landmark Tower, you have wide towers to the left, but narrow towers to the right
These buildings vary quite a bit, but they tend to be around 8 to 10-story tall. However, they are quite narrow, maybe around 10 meters on average (33 feet). They are quite common in older Japanese cities, as the Japanese tend to respect private property a lot. Meaning that not only will they tolerate their neighbor building higher buildings, but also that buying many neighboring lots to build a big, wide apartment block tends to be quite difficult. Lots tend to be quite narrow too because of earlier buildings. All of these images tend to be of wide streets because of the zoning rules that make building height dependent on street width. Also, note that unlike European cities, shared walls are extremely rare in Japan, even with narrow lots, buildings tend to leave some space between the walls, this is likely because shared walls require quite a bit of regulations to work in a context where each lot is owned by a different owner and there is constant construction going on with buildings being brought down and rebuilt.

Anyway, the result of such skinny towers is that they don't impact the walkability of the streets much, if at all. They're not disturbing like the wide buildings seen earlier as they offer a varied walking environment and their narrowness means that almost all of their front is dedicated to an entry point for the building, without any blank wall. It results in streets formed of a multitude of doors, one after the other, or even of small stores.

I've spoken of Sapporo recently, a much more recent city, and its modernity means that it tended to have wide streets, but also relatively wide lots. Meaning that their towers were wider on average.


Sapporo, some wide buildings, some less so
Sapporo from the air
Sapporo from the air
Still, there is a way to avoid the curse of blank walls with wide towers, as you can see in the next image:
Sapporo, the wide building to the left is actually split in two narrow stores on ground level
In this image, the wide building actually is split up at the ground level, with two narrow stores offering a better environment to pedestrians. This is unlikely to have been thought out by urban planners of the city, given how Japanese planning is less invasive than European planning, it may just be about the company that built the building doing what it thought was best to maximize rents.

Here is another example:
Still in Sapporo, this building has actually 4 or 5 stores plus a door for the apartments that are visible on this photo

In conclusion

I don't think building height matters much in terms of providing an interesting and human-centric built environment for people. The reality is that human beings tend to look down, not up, when they're walking, what's most important is populating streets at eye-level with human-friendly features: doors, balconies, verandas, store fronts, etc... and to make sure that there is a great variety of buildings on the way to provide for interesting and diverse environments. For that purpose, the important thing is to have narrow buildings, whether they have shared walls or not, it doesn't matter.

The entire debate on height is probably from a mistake, from misidentifying what makes the large, tall buildings so deleterious to street life. It's not that they are tall, but that they are wide. Wide low-rise buildings are no better.

Wednesday, June 25, 2014

To hell with punctuality! A personal rant

So, this is a short rant about the obsession with punctuality of some transit operator. 

First of all, I wrote an earlier article talking about bus speed and ridership, about how as the number of boardings increase for a bus, the more it has to stop and the slower it gets. But how can you predict exactly how many times a bus will have to stop at its bus stops? Or at traffic light? Or how fast traffic will be moving? The answer is, you can't, but you can expect a certain range of values.

So let's take this graph I made for the earlier article, showing a theoretical relation between average speed of a bus line and the number of boardings per kilometer.
Average speed of urban buses versus the number of boarding per km, supposing bus stop spacing of 200 meters
Now let's suppose that you expect the average number of boardings to be 3 per kilometer, plus or minus 0,5 boarding per kilometer. Visually, it means that you can expect this range of speed:
Average boardings/km and range of possible values
So in this case, you expect speed to vary between 19 km/h and 16 km/h. It may not seem like much, but on a 10 kilometer route, that's a 6 minute difference. So what is a transit agency to do?

Maybe the spontaneous answer is to schedule the bus line for the average and let the chips fall where they may. On average, buses will be on time, but some will be a bit early, others a bit late. Except that this is often seen to be a terrible idea. A bus can be a bit late, no worry, most users even expect buses to be a bit late. But a bus shouldn't be early, because then anyone who followed the schedule and arrived at the scheduled time will rage at missing their bus as it was too early. On less frequent lines, missing a bus can impose quite a delay on users, up to 1 hour on some suburban lines, but even on urban lines, 10-15 minutes until the next bus aren't rare. So you don't want buses to be early.

The solution is to then schedule the bus for the average, or even the worst possible scenario, and tell the bus driver to slow down if he's early to follow the schedule. In effect, scheduled times mean "the bus will arrive no sooner than...". Visually, it gives something like this:
Range of speed for the scheduled average with directive not to be early
So in this case, speed varies less, but it does so by being capped to a maximum of 17 km/h. So though you improve predictability, you also slow down service and increase travel times for most people. Yay?

That is why if you have ever taken a bus during off peak times, you will notice the bus driver is driving quite lazily, at 40 km/h (25 mph) in 50 km/h zones for instance (30 mph). It's not that he wants to be slow, it's that he's being told not to be fast.

But that's not all. Administrators often love punctuality, it's an easy metric to calculate and to boast about (our bus service has 99% punctuality! We are #1!!!). So what they will do then is schedule the worse possible case, assume the slowest bus possible is the rule and tell all buses that have either less passengers, face less traffic or hit more green lights to slow down to match its speed. Meaning in the previous case that all bus drivers will be told to have an average speed of 16 km/h and no faster.

So... to avoid some buses being 2 or 3 minutes late, you impose a delay of 3 to 6 minutes on ALL users. If you care about good service, that is a terrible, terrible deal that will waste dozens of hours for users each day and make it more expensive to run the bus. But hey, buses may be slow as hell, but at least they're punctual!

So that is why I say: to hell with punctuality. OK, punctuality on a transit line with the means to be punctual without slowing down, like subways or trains on their own right-of-way, is pretty great. But punctuality achieved by making the entire service worse? That's insanity. The next time you hear a transit agency head boast about their record of punctuality, hear what they are really saying: "we have slowed down all our buses and made our service crappy to game the stats about on-time service".

What brought this about is that my transit agency has just modified my buses' schedule to deal with a temporary construction zone, adding a 10-minute buffer in their schedule that wasn't there before. However, the people doing the construction have given buses a lane to avoid all the congestion and traffic, so that in effect, buses are only slowed by 1 or 2 minutes.

The result is that, this morning, my bus went at 30 km/h (20 mph) in a 70 km/h zone (45 mph), but even this wasn't enough, so the bus driver had to stop 2 minutes at an intersection not to be early on his schedule. The poor guy looked apologetic and really embarrassed, but I don't blame him. I blame his idiot bosses. This means that my trip to work has just gone from 30 minutes to 40 minutes each and every single fucking day, all so a transit operator can boast about his on-time performance.

I'll take a bus that sometimes is 2-3 minutes late over a bus that is on-time but ALWAYS takes 10-15 minutes more to get to destination. Maybe I'm the only one, but I had to get it off my chest.

Monday, June 23, 2014

How density limits bring about gentrification

It had to happen, I had to talk about gentrification sometime. For those not in the know, gentrification is a process in which a community, either a poor one or a middle-class one, sees an influx of new residents of a significantly higher income level than current residents. This often comes with new housing, services and stores made to cater to these new residents. Anti-gentrification activists warn of possible displacement as housing for the poor and middle class is destroyed to make way for more luxurious housing, and the loss of local stores and community.

Now, to be quite frank, I find anti-gentrification activists to be merely another form of NIMBY. It's funny how the rich don't want the poor in their backyard while the poor... don't want the rich in their backyard either. Anti-gentrification activists blame the rich for fleeing urban neighborhoods in the 50s and 60s (white flight), and now blame them for coming back.

Anyway, anti-gentrification activists tend to be on the front line of NIMBY movements to block density and any kind of construction in established communities. Height limits, regulations making it hard to renovate, downzoning (making zoning even stricter and allowing even less density), all are tools they tend to use. In a way, it's understandable, they see all the new buildings and renovations all meant to appeal to a richer, higher class type of people than current residents and they simply think that the way to avoid gentrification is to simply avoid construction and renovations. It can somewhat work, as the rich won't tolerate living in rundown tenement housing amongst the rats and bed bugs... but it's absurd because you also condemn current residents to living in rundown tenement housing amongst rats and bed bugs. Talk about the cure being worse than the disease.

I think clear thinking is needed on the issue of gentrification, and that these construction and density restrictions in fact make gentrification worse, not better.

Dynamic of the housing market

One of the basic facts of the market is that luxury products tend to have higher profit margins than mainstream products. So the first market segment that will be addressed is usually the richest people, who are the "low hanging fruits". As the upper end of the market get saturated (which doesn't take long as there aren't that many rich folk around), producers will start to make products for lower sections of the market, until they reach the bottom and the entire market is saturated, or at least until it's as low as they can do without becoming unprofitable. So the market satisfies the rich's needs first and goes down to the list after.

So what happens if there are restrictions on how much of anything can be built? Well, then the upper end of the market will not be saturated, so only the richest will have their needs served.

I don't know who said it, it is not me and I would like to attribute it correctly, but anyway, there is this great quote I've read: 

"If carmakers were allowed to build only 10 000 cars per year, they would only be making luxury cars".

We even have a confirmation of this principle in real life. In the 1980s, Japanese companies were compelled to accept "voluntary" quotas on the number of cars they would export to North America, under threat of facing extremely high tariffs. Since the number of cars they could make for the North American market was limited, they responded by creating Lexus, Acura and Infiniti, luxury brands of their otherwise mainstream, middle-class cars. The Japanese automakers responded to quotas by going after the luxury market to get as much profit from this fixed amount of cars as possible. It's not just because profit margins are higher for luxury cars, but also that the profit per vehicle is larger even if the profit margin is the same. 10% profit on a 20 000$ car yields 2 000$ profit, 10% profit on a 60 000$ car yields 6 000$ profit.

The fact is, the housing market is exactly the same.

If you limit the number of housing units people can build in a desirable spot, then promoters will automatically build housing for the highest possible end of the market that would be interested in living in that spot. So when an area becomes newly desirable or a building restriction (due to zoning or parking minimums) is lifted, naturally the first buildings that will be built will be for the richest people who want to move there.

This is normal, and it should be temporary, because there just aren't that many rich people around. If you continue accepting new constructions, at one point the luxury housing market will be saturated and the new constructions will be more affordable as the promoters have to target poorer demographics.

Unfortunately, when the first constructions occur, there is often panic about "gentrification", many seem to believe that any new constructions will inevitably be for the richest, and so try to block any construction, or at least limit them. But in fact, this is a self-fulfilling prophecy, because by restraining supply increases, they keep housing in severe shortage and make sure that supply isn't able to increase enough to ever saturate the upper end of the market. So the reason new constructions aren't targeted to the middle class and below is because we don't allow enough of it.

Still, there is a way for the poor and the middle class to outbid the rich for the available land in a neighborhood...

Higher density: the mass' ace in the hole

I've said it often and I'll repeat it: multi-family uses tend to generate more profit per square foot of land than single-family use. Even if the profit margin is lower, the profit associated with one condo building tends to be higher than that of houses built on the same terrain, by sheer numbers alone. So that's how people can outbid the rich, by tolerating higher densities and smaller units than the rich will.

One of the tactics used by the anti-gentrification crowd is to limit density to what exists currently. The reasoning is that blocking the construction of anything is quite hard, but if you limit density, you make it much harder to build anything because any replacement of a building with a new one will not be able to dilute the value of the older building amongst more units. So the new units will have to incorporate the old units' value, a cost that is a severe drag on profits and so may make new projects harder to justify financially.

The law of unintended consequence is in full force here. These limits also impact current residents by making housing more expensive. As I said in an earlier article, what limits the price of housing is the ability to add new housing, the cost of building new units is a ceiling to how high the price of current units can be. If you restrict additional supply, you are in a shortage situation where prices are likely to explode. It also results in keeping housing around well after its "best before" date. Furthermore, it makes sure that the ONLY new units that can be built will only be for the richest people who want to move there, because these density limits deprive the middle class and the poor of their one ace in the hole that allows them to outbid the rich: higher density and smaller units.

So though density limits may reduce the pace of construction, it will also make sure that all new constructions will be the type the anti-gentrification activists hate, for people at an income level far above the current residents'. They have just made it impossible to build new housing for people similar to current residents.

An example of what density limits to prevent gentrification brings about: the "monster home"

There is no greater demonstration that density limits are useless to prevent gentrification than the phenomenon of so-called "monster homes" or "monster houses" that are quite common in Vancouver and Toronto. In both of these cities, there are exclusive single-family detached zones quite near to the center of the metropolitan area, zones in highly desirable locations that would normally attract low-rise, if not mid-rise, multifamily developments. However, these developments are banned and density is severely limited as lot sizes are imposed by zoning and only single-family developments can occur. In terms of density limits, you can't do much better.

So what happens? Essentially, either rich people themselves or developers buy older houses, that are already expensive but that are not suitable for the rich (too small, not enough amenities, etc...), they tear them down and build modern houses in their place, houses that are as large as allowed in zoning. These houses are tremendously expensive, not only because they are luxury housing, but because they must include in their price the cost of the earlier house that got torn down.

In Vancouver, older houses to the right, monster house to the left
Surrey, a Vancouver suburb, an older bungalow in the foreground with a recent monster house in the background
Monster houses tend to be less talked about because communities in these areas are much weaker than in developed urban neighborhoods, as these are car-dependent areas, the direct environment where people live matters less as few people walk and bike. Still, it is gentrification at its worse, because since these are low-density middle-class developments being replaced by luxurious low-density developments, the displacement is really one for one, if not two for one if lots are merged. And the new houses are so expensive they are unaffordable for all but the richest.

When a neighborhood becomes highly desirable, there really is no real way to keep it from gentrifying, the rich will outbid everyone on a level playing field. The only solution is to allow higher densities, which can allow current residents to keep residing there by accepting higher density housing.

But what happens if the rich also tolerate higher density housing? Many rich people go for condos in mid-rise or high-rise towers, does that mean that they are utterly unbeatable when they accept high-density housing?

Yes, it does. But it's not a problem, because the great thing about density is that it doesn't take a lot of place. And there aren't that many rich people to go around, so if they accept to live in high-rise condos, they will inevitably take very little space and the danger of displacement will be negated.

In conclusion

It is normal for some place that is reviving after years of neglect to have an influx of people who are richer than current residents as promoters will first satisfy the desires of the richest people, since there is more profit to be made with them than with other people. However, that is temporary as the upper end of the market gets saturated very quickly, if people don't try to maintain an housing shortage by limiting construction in a misguided attempt to prevent displacement. It's like tearing off a bandage, the slower you go, the more it will hurt.

So the solution to gentrification is to allow even more construction to allow the upper end of the market to get saturated quickly, to relieve pressure on house prices and to allow current residents to outbid the rich by accepting higher density housing.

At least, that's the solution is the problem anti-gentrification activists fight against is effectively the displacement of current residents... if the real unspoken problem they seek to prevent is the presence of richer people, that they seek to maintain homogeneity in a neighborhood, then that is no "solution", but neither is social mixity a problem in my view.

Saturday, June 21, 2014

Ode to the multiplex: Québec's traditional urban housing

Like most kids in Québec born after the 1960s, the residential neighborhood where I was raised was defined by one particular type of housing: the bungalow. 

Canadian bungalows: one floor, wide...
...but shallow (note, that's 40 km/h, not 40 mph)
A one-story single-family house with a similarly big basement, a basement often sold unfinished yet finished by the owner later on, on a large lot. This type of housing was endorsed by the CMHC, the federal Canadian institution on housing and mortgages, that made architecture contests to encourage architects to think up new small single-family homes, then making the blueprint available to builders. These contests were pretty honest in admitting that they were attempts to invent single-family homes for the age of the automobile.

However, exploring the old neighborhoods of Québec, I could only notice that this bungalow's domination on housing in the suburbs is a recent development. The real traditional housing type in Québec cities is actually the multiplex: duplexes, triplexes and more. In fact, the CMHC did one contest only for multifamily housing, and that was for duplexes, a contest they did explicitly for Québec, where duplexes were common.

I've touched on this in the article comparing Montréal's and Toronto's modes of development, but I may have given the impression that this type of housing is only present int Montréal, which isn't the case. Though neighborhoods of multiplexes with shared walls is common only in cities, the truth is that detached multiplexes are common in essentially all Québec cities, big and small.

Montréal triplexes

Duplex in Boucherville, a former village turned into a Montréal suburb

Duplexes in the Old Terrebonne, a northern suburb of Montréal
Multiplexes and duplex in Shawinigan, in the Mauricie region

Multiplexes in Trois-Rivières

Duplexes in Victoriaville, a former rail station town between Québec and Montréal
Multiplexes in Beauceville... reminiscent of the low-rise apartment blocks in Japanese cities

Multiplexes in Charlesbourg, a Québec City suburb

Multiplexes in Saint-Georges, a small city in the Beauce region, south of Québec City

Duplexes and multiplexes in Alma, a small city in the Lac-Saint-Jean region

A triplex in Rimouski, in the Bas-Saint-Laurent region
Essentially, if you go on Google Maps, search for "church" in all Québec town and explore the area around them, you will quickly find multiplexes like these.

What defines the Québec multiplex?

Stratified multifamily housing in relatively small buildings

Yes, it's pretty evident, yet it's worth noting nonetheless. In much of North America, single-family homes are the norm, even in old working class cities. Cheaper housing is identified with smaller lots and buildings rather than opting for multifamily buildings, with these reserved only for the poorest of the poor. This may be a cultural thing, with Anglo-Saxon cultures generally looking warily towards multifamily housing. In the US, multifamily housing has sometimes been called "unamerican" or even immoral.

Same thing over in the UK, with typical working class housing being row houses rather than stratified housing. Which seems to indicate a probable cultural cause for this distinction.

The multiplex is a multi-family housing type, but its dimensions are more similar to single-family housing than to tenement blocks. Multiplexes tend to be 7 to 15 meters wide only (24 to 50 feet) and 10 to 20 meters deep (33 to 66 feet). They are built face to the street, just like single-family houses, whereas large tenement blocks, especially in suburbs or small cities, tend to be built far from the street, often facing other blocks rather than the street.

Sometimes owner-occupied

Large multi-family apartments in North America are often owned by relatively wealthy landlords, who will own many such buildings and have little to no interaction with the occupants, leaving a concierge to collect rents and deal with renters.

Québec multiplexes tend to function quite differently. For one thing, in many cases the owner of the building will actually reside in it, in the ground floor unit of the building, and rent out the upper units (and sometimes, the lower, basement units). This owner will have a regular job, owning only this building, dealing with renters on a personal basis. For the owner to reside in the building shows that there is little to no social shame in inhabiting it, otherwise, the owner would rent out the ground floor unit and use the proceeds to live elsewhere.It is not unheard of either for a multiplex owner wanting to sell to first approach long-term tenants and offer them to buy the building off of them.

The presence of the owner tends to have a good impact on the building's maintenance, as if the owner lets the building go to waste, he will be trashing his own house. Also, in many large multi-family apartment buildings, the yards tend to be almost completely ignored, tenants have no real claim on them and use them very rarely, if ever. The owner could manage them, but as he desn't live there, what's the point? So yards become only buffers of grass without any active management, only the most basic maintenance. In multiplexes with the owner on the ground, the yard is almost like a single-family house's yard. The owner may have a small garden or some nice arrangement as they live on the ground and would enjoy it. Even when the owner isn't an occupant, the smaller size of the building tends itself better to an appropriation of the yard by the tenants, especially those on the ground level.

I have a suspicion that many cases of multiplex were ways for people to afford to buy a lot and build the house they wanted, by getting a supplementary income from renting out part of their house. In a way, multiplexes have similarities with secondary suites or granny flats and may be considered a particularly advanced form of such housing.

Each unit has its own door, or shares a door with only one other unit

This is one of the defining characteristics of Québec multiplexes. I mentioned it in the earlier Montréal vs Toronto article. Québec multiplexes tend to have little to no interior common areas, which allows units to have doors directly to the outside, giving them a more private feel. This often results in exterior staircases to the upper units. I mentioned how this was perhaps a result of regulations in Montréal in the earlier article, but this doesn't explain the presence of so many exterior staircases in multiplexes of many cities outside Montréal. It seems quite unlikely that the existence of these stairways are a result of Montréal regulations regarding setbacks, as there are similar multiplexes all around Québec, even in small towns where space was far from lacking.

The exterior stairway may be a result of owner occupation, as the owner would have wanted to have the impression of living in a single-family house, taking the stairway outside allows the owner to have a veranda for his ground level floor, a veranda that will not be shared with his tenants. This doesn't explain why they would build exterior staircases on triplexes that look like the owner never occupied the ground floor, except maybe "monkey see, monkey do". But not all multiplexes have exterior staircases, some have a secondary entrance on the side or even on the front.

What remains constant is the multiplication of doors, with each unit having its own door to the outside. Even when there is a secondary ground floor entrance with an interior staircase, the upper floor unit or units will often have a door on the upper floor, opening on a balcony. Balconies (on upper floors) and verandas (on the ground floor) are present on almost all multiplexes. This provides semi-private outside areas for units and they seemed to have been very frequently used at the time based on contemporary sources.


There is much to like about traditional Québec multiplexes. They are an extremely versatile housing form, able to offer from 30 to 100+ units per hectare, from 6 000 to 20 000 people per square kilometer. They can be duplexes, offering two big housing units like two bungalows one over another, or they can be multiplexes with 2 or 3 1-bedroom units per floor above one large 3-bedroom unit on the ground floor, etc... The multiple doors and balconies that characterize them and their narrow size means that they allow for a lot of individuality on residential street, every building may be different from their neighbors and offer housing choices that simulate single-family housing thanks to each unit having their own door to the outside. They avoid the blank walls of large apartment blocks.

A bloc of apartment buildings with parking lots
A bloc of multiplexes with driveways with roughly the same density and number of parking (in the more spatially efficient form of driveways)

The fact that they are small buildings create density without the risk of the "tower in a parking lot" syndrome. They also have distributed driveways, avoiding the plight of huge parking lots caused by tall and wide blocks.

There is also a potential economic advantage, as the greater number of buildings and the smaller size of buildings make the threshold to becoming a landlord much lower. This increases competition amongst landlords, as when there are only a few apartment blocs, there will necessarily be few people owning most of the rental housing stock, and building new blocs will require deep pockets, putting them out of reach of most people.

Monday, June 16, 2014

Sapporo: a relevant Japanese model for North America

I've spoken a lot about Japan and what they do differently for urbanism. However, one criticism I often hear when I bring up Japan is that comparisons to Japan are not actually relevant because Japanese cities are very old and have been built in another era, with Japan living in an essentially medieval feudal state until the Meiji Era in the mid 19th century. The assumption is that, more than just a cultural factor, Japan just has a lot of old city grids and streets, and that the Japanese inherited their cities from their forebears, and since we inherited different cities, Japan is just not that relevant as a model for urbanism in North America. And that's not just true for Japan, mentions to Europe or older North American cities have the same reactions, as if new cities couldn't possibly be built other than for the car.

To be fair, that argument is accurate to an extent. Tokyo, even before the Meiji Restoration, had a population of over 1 million people, a milestone it reached in 1721, making it the largest city in the world even during that time. Consequently, it has an enormous number of streets and extremely small blocks, especially compared to American cities:
A section of Tokyo compared to the street grid of Manhattan

The same section compared to the notoriously small Portland blocks
So even compared to Portland, Tokyo's blocks are much smaller, as little as one quarter of Portland's blocks. The streets are extremely narrow in those blocks, as little as 4 meters across (around 13 feet). In such a context, car speed is significantly reduced by the narrowness of roadways, the presence of pedestrians and cyclists and the number of intersections, making walking and biking much more attractive in comparison.

But Japan isn't only Tokyo (Even if a quarter of the population resides in the Greater Tokyo area). In fact, the cities in the north of Japan developed much more recently. Sendai, the main city of the Tohoku (North-East) region, was founded only in 1602, predating by just 6 years the foundation of the city of Québec. Even better, the northern island of the archipelago, Hokkaido, has only very recently been "colonized" by the Japanese population. There used to be only semi-nomadic people in the area, the Ainu, and colonization of Hokkaido has only started in earnest in the mid-19th century, when Japan was opening up and the Japanese quickly learned all they could from Europeans and Americans to catch up to them. The biggest city of this island, Sapporo, now has a population of nearly 2 million people, yet was founded only in 1868, at roughly the same time as the main Texas cities (Dallas, Austin and Houston).

Sapporo is the one I find most relevant, as it was built following the planning ideas prevalent in America at the time, and many Americans are involved in the early years of the city. The result was a city built on a rectangular street grid with wide streets and big blocks:

Sapporo versus Manhattan block sizes
Sapporo versus Dallas block sizes, both are almost the exact same size
The street grid of Sapporo, the skeleton of the city, is thus extremely similar to 19th-century American cities, what Nathan Lewis on his blog refers to as an "hypertrophic city". And the streets are pretty wide, in the downtown area, the width of streets, including sidewalks, are around 20-25 meters wide (66 to 82 feet), some are as wide as 35 meters (nearly 120 feet).

This is often considered to be the thing that shouldn't be done, as wide streets make it harder for pedestrians to cross them, they incite speeding and allow cars to move faster. How is Sapporo doing, is it yet another car-dominated and oriented city like all those North American mid-19th century cities?

Well, not so much. According to the transport studies of the Japanese government, the share of all trips that are made by car in the entire metropolitan region is 45%, higher than in Tokyo and the Osaka-Kyoto-Kobe region, but lower than in the Nagoya region, despite Nagoya being older and much more populous, and also lower than the Sendai region and the Fukuoka region. In comparison, the car's mode share in the Montréal region is 69% according to the AMT's most recent origin-destination study. Note that comparisons are a bit tenuous as there are different ways to calculate mode share, whether one includes the suburbs or counts just the central city, all trips at all times or just commutes to work, etc... I assume the numbers I found for Japanese cities include all trips in the city and its suburbs, which is what I calculated for Montréal, but the data I found for American cities were mostly commute-to-work numbers which are not directly comparable. I wanted to make comparisons to American cities, but couldn't find appropriate data. Note though that Montréal is considered one of the least car-dominated cities in North America, behind New York, but about the same level as Toronto.
Mode shares of Sapporo and Tokyo
Bus mode share tends to be really low in Japanese cities, so these numbers aren't the exception but the rule. The highest "bus" mode shares are in Nagasaki (12,5%) and Hiroshima (10,0%), which both happen to still have widely used tramway systems, which seems to indicate that trams are considered "buses" in the Japanese survey. You may have seen other mode share data on Tokyo showing a much higher transit mode share than this, this data is for the Greater Tokyo area, the data showing 50-60% of mode share for Transit is likely for the city of Tokyo itself or for commuting only.
Mode shares of Sapporo and Montréal
For this graph, I had to group together "walking" and "biking" into "active" mode of transport and "bus" and "rail" together for Sapporo, in order to make it comparable to Montréal's data which doesn't differentiate between walking and biking, nor between subways and buses.

Anyway, Sapporo manages some respectable numbers for sustainability. Its transit system is much less dominant than Tokyo's and actually doesn't perform better than Montréal's. Its advantage over Montréal comes almost exclusively from active modes of transport, walking and biking, which are nearly three times as high as Montréal's. So people in Sapporo walk and bike in great numbers. And Montréal's relatively poor performance can't be blamed on the climate, Sapporo is just as cold and even more snowy than Montréal.

So, how do they do it?

The streets

As I said, Sapporo's streets, especially in its downtown area, are pretty wide, but this is based on the distance from property line to property line, including sidewalk and roadway. Sapporo made the choice of giving only as much space to cars as they absolutely need, their lanes are essentially all 3 meters wide (10 feet) only, and most streets have 4 lanes (2 in each direction), rarely do they have more. On-street parking and large shoulders are rare.

What do they do with the rest of the street? They have sidewalks, very wide sidewalks. Their sidewalks tend to be around 3 meters wide (10 feet), sometimes even more. Now, some cities like New York and Dallas have similarly wide sidewalks in their downtown areas, but the main difference I see is that, in Sapporo, they respect the pedestrian space much more. Meaning that when they plant trees, add poles and signs, they do it at the very edge between the sidewalk and the street, whereas too often in North America, we tend to put these objects nearer to the middle of the sidewalk. Also, their curb cuts tend to be much less invasive on the pedestrian space, with cars facing a short ramp that takes them to the level of the sidewalk rather than taking the entire sidewalk to the level of the street.

The last image is at the limit of the subway system, in a suburban area, despite this, the sidewalk is comfortably wide and the street is only as wide as it needs to be. In all cases, every pole and tree is put at the extreme limit in order to preserve space for pedestrians and cyclists and to protect the sidewalk from cars. Also, the sidewalks are well maintained.

On the other hand, look at these examples of similarly wide sidewalks, wasted by trees and crap put way too close to the middle of the sidewalk, demonstrating how pedestrian space is seen more as a buffer between buildings and street.
Sainte-Catherine street in Montréal, note how the poles and trees are planted far from the curb compared to Sapporo
Dallas, the sidewalk is even built to indicate to pedestrians that half of the sidewalk is for objects, not people
Dallas again, with an otherwise large sidewalk poorly maintained and with poles put almost in the middle of it
The result of the lack of on-street parking, of wider uncluttered sidewalks, of narrower lanes, is that walking and biking in Sapporo is quite comfortable. Moreso than similarly built North American cities. I had written a bit earlier on the importance of providing adequate pedestrian space in cities, but after being to Sapporo, I must admit omitting at the time the impact of objects put in the way of pedestrians, reducing their walking corridor.

Japan zoning: density and mixed use

My post on Japanese zoning is by far the most read on my blog, and is likely to remain this way. I've talked about how Japanese zoning includes mixed uses by default and how it's much friendlier to a mix of different housing types and greater density. Sapporo being a very recent city, it bears the mark of Japanese zoning more than most other Japanese cities. To summarize a bit the principal points of Japanese zoning:
  1. Maximum use zones rather than single use zones, residential still allowed in commercial zones
  2. Height limits depend on street width
  3. To build malls or offices, you need to allow dense residential uses also (use bundling)
So since Sapporo's streets are wide, it means that tall buildings are common in a lot of the city, typically near subway or train stations. Likewise, as there is residential density in those places, shops and restaurants seek these places out to be within proximity of potential customers. Which creates nice hot spots of density all around the city.

In North American cities with our very strict planning, we tend to adopt an idea that buildings' sizes should be inversely proportional to their distance to the downtown area: skyscrapers in the downtown area then heights falling as you go farther until you hit single-family housing. That's not how Sapporo plays it, at all. If you look at the city from vantage points, you can see "ribbons" of mid-rise and high-rise buildings extending from the downtown areas, grouped around the subway lines.

Click on pictures for bigger versions.

The result of this is that the dense areas of Sapporo go much further away from the downtown area than most North American cities and there is a strong tendency to mix uses, so that dense residential areas tend to be within walking and biking distances from corner stores, supermarkets and malls. What helps this proximity is the reduced parking lots in dense areas, as parking minimums are lower in Japan, plus stores will do underground parking and even parking on the roof of malls to have a more compact form.
Mall in Sapporo with some parking in front...
...but see the ramp in the center of the picture? It takes cars to a parking lot built over the mall

The only parking lot in front of this suburban mall is a bike parking lot
Cars must take the ramp to their own parking upstairs
The market has provided Sapporo with enough parking where it is required, in a form that takes little space, not hurting the walkability of the city:

This parking garage is in the middle of the huge entertainment/red light district: Susukino
Note that there are some problems with the mode of development of Sapporo. In places around subway stations farther from the downtown area, places aren't fully built out, leaving 10-story apartment blocks standing in the middle of currently vacant lots, creating a weird effect. As the areas get built out, the urban fabric will be better and it doesn't hurt the walkability of what is currently there, but many neighborhoods around farther subway stations look a bit like smiles with teeth missing, or even stonehenge.

Some would also say that this makes Sapporo's skyline look pretty poor and mundane as it doesn't look much like a mountain but more like a long hill with few distinct buildings in the mix. They may have a point, but skylines in my eyes are pretty irrelevant to how nice a city is to live in. And anyway, they just have to wait for the night to fall to compensate.
Sapporo at night from Mount Moiwa
In closing, I think Sapporo is a city worth studying for its modernity and for its roots so similar to many North American cities. It shows what could have happened had we not put such strict zoning straitjackets on mid 19th century cities, preventing their downtown areas to spread around transit lines. It shows that even hypertrophic cities can do a lot to make walking, biking and transit viable alternatives to cars. Though cities founded in the mid-19th century may never look like Florence, Tokyo or Paris, it is yet possible for them to develop in a more sustainable, walking-friendly fashion, as Sapporo demonstrates.