Building Dense Does Not Have To Be Dense
November 4, 2014

This post was actually a comment I wrote on the Strong Town's article A Case For Height Restrictions. The comments broke into a discussion over density, building heights, property values, and the issue of people holding onto vacant lots. My response was long enough to warrant its own blog post.

Our built environment is largely a physical manifestation of the economy. If your city goes from vacant lot to block-sized high-rise monoliths - what does it say about your economy?

It's not a uniquely American problem. The Gold Coast in north-eastern Australia has a similar problem. From a distance it looks like an impressive skyline;

Up close it's vacant lots, single family homes, 2 story motels, the odd strip mall, next door to massive high-rise condos;

Houses can go for millions just based on their location because the owners are waiting for either a condo developer to buy them out or for a millionaire to build their dream vacation home. The center of city is quite walkable and tourist friendly (but only takes up a 1/4 mile x 1/4 mile) but the vast majority of the Gold Coast is auto-dependent and sprawled - despite being highrise and "dense" by most planning definitions.

When you combine density with the auto-oriented pattern (you can't walk anywhere - except maybe down to the beach) everyone still needs to drive - so you end up with bad traffic. Fortunately, most people in these holiday towns are either holidaying or retired - they're not having to commute during rush hour - but could you imagine the traffic problems if you had 100 cars coming out each skyscraper at 8:00am?

Some people call this pattern "towers in a park" or "skyscrapers in the suburbs", it's also known as the Radiant City pattern;

We know it mostly fails today because it's expensive (public - handling all of the traffic, private - large amount of capital required to build in this environment) - not to mention unwalkable and rather hostile and bland on foot (too many cars, everything is very spaced out) and it's very high risk (if I build a $25m building, will I sell all of the apartments?)

The Swiss-French architect Le Corbusier thought the Radiant City was the ultimate solution to our problems and suggested we chase our tails (wider roads, taller buildings) that we'd end up with something like this (notice the freeway interchange at every intersection);

Not a world I want to live in.

"Hey Chuck, we believed you when you told us our suburban pattern isn't generating a ROI. But now you're saying dense condos are bad! You're sending me mixed messages man!"

When urbanists talk about density, the main argument is that density will reduce the scale between destinations. By reducing the scale between destinations - walking and biking become more viable (don't have to accommodate all of those cars), we don't need to provide as much parking, there's more street life and other street activities become more viable (vendors, street performers, etc.), it's more fine grained.

The problem with building skyscrapers in suburbia is that it's still single-use and auto-dependent. By increasing density but not bringing destinations closer together, we're just increasing then number of people trying to get from A to B, without making A and B closer together.

If 1 family per lot in the auto-dependent suburban format is unsustainable, then how will 25 families per lot in auto-dependent suburbia be sustainable? Sure, we've scaled up the number of families per lot by 25, but we also have to scale up our auto-oriented infrastructure by 25. We're still inheriting all the problems of auto-suburbia, just scaling up the numbers (both in tax revenue, and in expenses.)

Is it possible to build dense without building up? Tokyo is one of the densest cities in the world;

Yet, most people live single family homes. How does that work?

Let's do the math: Average lots in suburban Tokyo are around 900 sqft. You can fit a 2 story, 1200 sqft home in 600 sqft, leaving 300 sqft for a courtyard or garage. If we assume 1000 sqft per property (when taking into account streets) - that's about 27,878 properties per square mile. Assuming 3 people per house - 83,634 people per square mile. It's possible because of basic pattern of narrow streets and small lots.

Dense, low-rise, and fine grained. But it's not for everyone. This is only one possibility. Of course, Tokyo is very expensive - but that's because it's also very huge (13m in the city, 35.6m in the metro) so there's huge demand. What if a small American town adopted this format instead?

Provincetown, MA has adopted the format of narrow streets and small lots. Here's their main street;

Here are some dense residential streets in Provincetown, MA.

Based on some rough Google Maps measurements, those are approximately 2,300 sqft lots and the street is 25ft wide (measuring the space between front doors). Small lots and narrow streets. It's dense, walkable, there's street life, and has all of the urban characteristics, yet it still has small-town, rustic charm despite being one of the most dense streets in America (see - density is not such a bad word!) Is this not a more desirable outcome than skyscrapers divided by freeways?

Well, maybe you like tall buildings, wide roads, and lots of traffic. I can't change that.

"That's great, but what about the that person that refuses to sell until he's offered $2.5 million?"

Every city has the story of the property owner with the vacant lot that won't sell until he gets the right offer. They're exercising their right in the free-market to sell or not. Maybe you can force them (even if it's morally or legally wrong) out. But that's not the point.

1) They're asking too much.

You could try putting an upper limit on how much they expect to get out of their property. One way is through price controls (but that's anti-free-market.) Another way is through Chuck's suggestion of using a height limit so they know that nobody is realistically going to build a 15 story condo on their property unless the surrounding area seriously intensifies first.

2) They're not doing anything with their lot to keep their taxes down.

Try switching to a land value tax instead of a property tax, so they don't have an incentive to not repair or improve their property. We could also have a law that if a property is not used productively (as a home, business, or under active construction) for 6 months it's declared abandoned and the city will auction it off.

Alternatively we could:

3) Minimize the impact.

Some stubborn people just aren't going to sell because they're a) stubborn and/or b) delusional about what they can get. Ultimately, it's going to be their loss. If we chase after density (as I mentioned above with fine-grained lots) - instead of them holding onto suburban-scale 120 ft of street frontage - what if they're only holding on to a 30 foot wide lot? By making lots finer-grained, we're minimizing the impact of 'holes' in the urban streetscape.

My solution would be a combination of the above, depending on the context of the town. But in general, more Provincetown and less Le Corbusier.

(I've left out talking about zoning - but the message for zoning and planning is implied - are we planning for Provincetown or are we planning for Le Corbusier?)



Optional link


What is 7 + 1 ?

Osric • 12.27.2014 • 06:55 AM (MST)
Interesting to see you use my home town as an example of small-scale density. The streets you show were built out long before zoning ordinances (over 100 years ago) and you can't build like that in town any more. Everything has to have significant lot size and setbacks, and there's a 2 1/2 story height limit. The current affordable housing crisis may cause some of that to change, and the streets you show are protected to some degree by historic district regulations.
Jacob Lynn • 11.22.2014 • 21:44 PM (MST)
Agreed in principle, and high-density compact streets with low-scale buildings sing to my soul. But the problem with the height restrictions that it's basically inconsistent with real-world incrementalism. If you buy a lot with a single-family home already on it, tearing it down and replacing it with a duplex or triplex will barely be worth your while as a developer because of the high upfront costs. The marginal cost of each new housing unit is lower if you can spread out the fixed land/demolition costs over more units. Alon Levy has a good post on this: What you are identifying is that we need to reduce barriers to pleasant, pedestrian-scale urbanism -- not that height limits are a good idea.
George Carty • 11.20.2014 • 06:58 AM (MST)
If you read the Urban Kchoze blog, you'll find out that the big advantage of the Japanese model is that it builds to a high-density where it is justified (ie in the city centre, and near to train stations), while the rest of the city is lower-density. This may be because Japanese people have far less expectation that the nature of their neighborhood will not change over the year -- perhaps it's because they do not view homes (as opposed to their sites) as investments as Westerners tend to do. Western cities do not make good use of their transit infrastructure, which means it generally needs heavy subsidies -- unlike Japanese suburban and underground railways which make a profit. Also, London (to give an example) has lots of people travelling by bus (usually over distances of 1 mile or less -- symptomatic of the "last mile" problem) because there isn't enough density near suburban Tube stations, while almost no-one travels by bus in Japanese cities.
Paul Fritz • 11.07.2014 • 18:24 PM (MST)
This is a great explanation of the difference between 'good' and 'bad' density. So many people have such a visceral reaction to density. They imagine Manhattan skyscrapers. But it's important to recognize that single-family neighborhoods can be dense and create really great walkable places as shown in your examples. Thanks.