« Graph Theory | Main | Monorail Stopped In Its Tracks »

September 22, 2005

The Manhattan Project?

This article -- which, in large measure, holds up Vancouver as a model for the redevelopment of lower Manhattan -- makes this arresting claim:

[D]owntown Vancouver has recently eclipsed Manhattan as North America’s highest density residential area.

This claim came as quite a bit of a surprise to me -- I thought I was a Vancouver statistics geek, but I'd never heard that before.  And after fiddling on the internet for a bit, I now realize why:  as far as I can tell, it's simply wrong. 

According to the 2001 Canadian Census, Vancouver's downtown peninsula -- the densest part of the city -- had just over 70,000 residents in a little over 2 square miles, for an average population density of just under 50 people per acre.  It's probably a bit higher now, but not all that much. 

Meanwhile, Manhattan squeezes more than 1.5 million people into just 23 square miles, which means that the average acre in Manhattan has 100 residents.  (That, of course, includes Central Park, which occupies well over a square mile in the center of the city.  Excluding that, as is done for Stanley Park in Vancouver, pushes Manhattan densities a bit higher.)

That makes Manhattan, on average, about twice as dense as the most heavily populated neighborhood in Vancouver.

I'm not trying to take anything away from Vancouver.  Its record in creating livable urban neighborhoods is truly remarkable.  Lower Manhattan can learn a lot from what Vancouver's accomplished over the past few decades.  Still, there's no purpose to be served in propagating bad information:  that can only lead to distorted expectations and, ultimately, bad decisions.

Update:  Apparently, Vancouver's downtown peninsula isn't even as dense as Brooklyn, let alone Manhattan.  Jeez.

Posted by ClarkWD | Permalink


TrackBack URL for this entry:

Listed below are links to weblogs that reference The Manhattan Project?:


All you gotta do is stand by a subway stop in any residential neighborhood in NYC to know this is nonsense.

I bet there are plenty of neighborhoods in Chicago and Washington DC that also exceed Vancouver.

Posted by: Kent | Sep 22, 2005 2:34:06 PM

Yep, this chart of the density of Chicago neighborhoods shows the near north side with a density of 48,500 per square mile.


There are at least half a dozen Chicago neighborhoods with densities that exceed 30,000 square mile.

Posted by: Kent | Sep 22, 2005 2:46:51 PM

I've heard that statistic many times since moving to Vancouver, but not usually about Vancouver as a whole, or downtown as you've calculated.
It's said that the West End is the highest density neighbourhood in North America. That's not the entire downtown peninsula, but rather the western portion of it. Generally considered west of Burrard, for those of you familiar with Vancouver.
It's full of high-rises like the rest of downtown, but those are ALL residential rather than commercial. It's pretty intense....and not for me and my family.

Posted by: Jeremy Brown | Sep 22, 2005 2:48:03 PM

Thanks, Jeremy. That makes a little more sense. Still, even the stat about the West End seems mistaken, unless I'm missing something.

The population density of the West end was about 234 people per hectare, or 95 people per acre, in 2001. See:

So that certainly makes the West End dense. But it's *still* not as dense as Manhattan is on average. And Manhattan's got business and commercial districts that are largely devoid of housing. So the actual Manhattan neighborhoods where most people live are even denser than the West End.

I'm not sure my family would be up for the West End's level of density, either. But when I was single I definitely was, and -- provided it's safe and pleasant enough -- I'd be very interested in it again when/if I'm an empty-nester.

Posted by: Clark Williams-Derry | Sep 22, 2005 4:38:57 PM

Perhaps the correct statistic is that the West End is the densest neighborhood on the *west coast* of North America. That seems more plausible, though the West End probably has some competition from bits of SF. In the end, it may come down to how a "neighborhood" is defined.

Posted by: Clark Williams-Derry | Sep 22, 2005 4:44:51 PM

Looking at demographia.com (as per Kent's suggestion) I find this:

It's a selection of neighborhood and city densities from around the world, and at various times over the last century or so.

Bottom line -- even the West End of Vancouver, as dense as it may be, ain't even close to being the densest part of North America. In 1990 at least, Manhattan's "Community District #8" fit 210,000 people into 2 square miles -- giving the neighborhood a density of ~166 people per acre.

Posted by: Clark Williams-Derry | Sep 22, 2005 8:05:13 PM

Clark, How about "...the West End is among the densest urban neighborhoods north of the Columbia River and west of the Rockies?"

Posted by: David Sucher | Sep 22, 2005 10:57:49 PM

Clark, How about "...the West End is among the densest urban neighborhoods north of the Columbia River and west of the Rockies?"

Posted by: David Sucher | Sep 22, 2005 10:58:34 PM

David -
That's more like it.

Posted by: Clark Williams-Derry | Sep 23, 2005 10:53:31 AM

What I had heard is "densest area in Canada," which certainly seems plausible depending on the geography. However, Statistics Canada won't give me easy access to that info. Vancouver's planning department says 31,360 per square mile for the entire downtown peninsula, including the West End; Montréal's planning department says one census tract there (can't tell which) nears 113,000, and the overwhelmingly low-rise Plateau arrondissement 33,918 per square mile.

The highest density Manhattan census tracts pack over 200,000 residents onto each square mile. The most dense census tract in Chicago tops out at 91,000 per square mile. Those numbers are lower for the larger geographies (Community Board districts in NYC, Community Areas in Chicago, ZIP codes, city limits, etc.) since a single census tract can be as small as a few entirely residential blocks, excluding any non-residential uses on adjacent blocks.

What makes me further doubt the claim is that Vancouver's towers are tiny compared to those in Chicago, much less Manhattan. Their planning department cites new developments there as averaging a floor-to-area ratio [FAR, they say FSR] of 2-4; in downtown Chicago, home of a famously lenient City Hall, many new towers reach 20 FAR, and we don't even get the parks or schools or social housing that Vancouver demands of its developers. Sure, our apartments are larger, but not ten times bigger!

Posted by: payton chung | Jan 5, 2006 9:30:00 PM