Here is an interesting chart (source) showing housing starts in Canada, by type, between 2000 and 2023:

As recent as 2000, single-family houses accounted for 61% of total starts and multi-family housing accounted for 39%. This flipped somewhere around the financial crisis and, last year in 2023, the percentages were 23% and 77%, respectively. This is a meaningful inversion which has helped our cities become more vibrant and more conducive to non-car modes of transport.

But in this recent article about Canadian housing, Donald Wright more or less argues: so what? We've been densifying our cities for all these years, but it hasn't helped our affordability problem. Supply must not be the answer to our housing crisis.

I'm not exactly sure what he believes to be the solution, but I don't think this problem is as simple as "we've built some housing, we made our cities denser, and yet housing is still expensive -- more supply must not be the answer. Let's move on."

Among many other things, it's important to understand what kind of density we've been building. Because up until very recently, we've basically taken the position that single-family neighborhoods should never be touched, and that density should only go in very specific areas -- and only after a lengthy and expensive rezoning process has been completed.

We've designed new housing to be expensive.

But attitudes are changing all across North America. We are now starting to do two very important things: (1) we are opening up more of our cities to intensification and (2) we are now allowing more multi-family housing on an as-of-right basis. Meaning, no lengthy rezoning exercises and no risk of community opposition.

These are two fundamental changes that should alter the kind of density that gets built. And in my view, it's going to be a positive thing for Canadian cities.

When we build next to transit, we often call this transit-oriented development. 

What’s interesting about this moniker is that it implies we’re doing something a little special — something out of the ordinary. And I guess that makes sense because, in many cities, it is often out of the ordinary. 

That’s why you don’t hear people at real estate conferences saying, “check out this new cutting edge car-oriented development that our firm is developing.” That doesn’t need to be specified.

But at the end of the day, I’m not sure how special transit-oriented development really is; it’s basically just urban development. Meaning, you put density on top of and next to transit stations and then more people take transit. That’s how this works.

On that note, here is an interesting study from the School of Cities that looked at Toronto’s transit network and how the populations around each station have changed (or not changed) between 1996 and 2021 (census data). 

If you look at the various transit lines, you’ll see that, in some cases, like downtown, we have added a lot of new transit-oriented development. This is good. Populations increased. 

But in many/most other cases, populations remained flat; or worse, they declined. This is a serious problem, and it shows how land use restrictions are forcing us to underutilize our existing transit assets.

Maybe what we need to do is stop thinking about transit-oriented development as something special, and instead remind ourselves that this is standard operating procedure. It’s just what you do next to transit.

Thanks to Sam Kulendran for sharing the above study with me.

I tweeted this out yesterday:

What I was getting at is that there's lots of available room within our existing boundaries for infill housing. We are nowhere near full, despite what some people will tell you. In fact, most areas are not dense enough to properly support modes of transport that aren't the car.

Of course, there are a number of ways that one could be offended by a statement like this.

One, you could argue that more density would make the city unlivable. Two, you could get into the chicken-and-egg game of whether a more expansive transit system is needed before allowing more density. Three, you could say that we already have enough zoned and unbuilt housing supply -- so why do we need more? And I'm sure that there are many others that I'm not mentioning here.

Density can be a counterintuitive feature for cities. It can actually make a place more livable by encouraging more amenities adjacent to where people live and work, and it can also reduce traffic congestion by empowering alternative forms of mobility. If the only reasonable way to get around is by car, then of course most people will drive.

We also need to avoid the chicken-and-egg mental trap when it comes to mobility infrastructure. Land use and transportation always work hand in hand and need to be thought of and executed on simultaneously.

Finally, the objection of already having lots of sites zoned for new housing is an enticing one. But zoned and delivered are two vastly different things. And the unfortunate reality is that there are a lot of zoned sites that won't be able to develop in the short and medium terms because the market isn't there. But that doesn't mean that other housing typologies couldn't be built.

At the same time, we need move away from "cruise ships of urbanity." Broadly speaking, Paris -- to cite just one of many examples-- is at least and on average about 4x denser than Toronto. And somehow, people still like living and visiting there.