Pittsburgh Is New “In” City Of 2012

[Centropolis]

You can play with density numbers but sprawl is sprawl, and it's ugly and dysfunctional no matter if the houses have a large backyard or not. There's simply more sprawl, and simply more people living in it in the places I mentioned. PGH didnt have the regional pop or economics to build huge amounts of 1990s mega sprawl - "dense" or not. Denver did in a huge way, so did KC and StL. I just don't see it any other way.

Yes, western sprawl is denser, but that wasnt to my point nor does it matter in this situation. No reason to perseverate over it in the cause of being a good son of Denver, or whatever.

[Private Dick]

Originally Posted by bunt_q

Even Pittsburgh, Aaron. You just don't see it, probably... it's so spread out, it doesn't look spread out, it looks almost rural. (That's been my observation in PA - the line between suburb and small town and rural is almost impossible to make out.)

So moral of the story, get some of those folks back into the core, Pittsburgh!

(I also looked at some individual townships, and it's the same thing. Pitt is nice and dense, but by the time you hit Penn Hills, McCandless Township, Monroeville, Ross Township, etc., you're into 1,500 - 2,200 ppsm. Whereas even our dreaded Aurora outside Denver manages to hang in at 2,800 ppsm, and Lakewood, Thornton, Arvada and the like are 3,100-3,600 ppsm. For the curious PA folks, Pittsburgh itself was around 6,000 ppsm, so in the state-level numbers the census churns out you get a sizable boost from Philly.)

The little comparison study you've undertaken is terribly uninformative -- as it does nothing to address the history of the region. Pittsburgh certainly has sprawl, but the majority of it is very different in context from what exists in a place like Denver.

Pittsburgh and the numerous other Southwestern Pennsylvania cities and towns developed concurrently and separately. Settlement patterns in SW PA in particular are far different than what has historically been the case in an area like the Denver region. Development did not spread out from a central Pittsburgh core to the surrounding region.

The region consisted (and still in large part consists) of numerous (hundreds literally) nodes... out from which development emanated and joined the fractured, topographically-severe, and geographically-isolated populated areas together... not radially, due to the extreme terrain, but instead hugging the flat lands of the river valleys along transportation routes.

What you observed about PA (above in bold) is a direct result of a very different pattern of development -- one that was dictated centuries ago by factors far different that what drives suburban sprawl in a place like Denver.

[Jelly Roll]

Originally Posted by bunt_q

This is actually very not true, and a common misperception. By and large, the suburbs get denser as you go west.

If they are perceived as more "macro" it can be due to a couple things. (1) There's less older development in the core of western cities to "balance" them psychologically, and we tend to judge a city by its core (rightfully so). (2) I think this is often given too little credit - they are just more visible out west. It's hard to grasp the full scope of a place in trees and hills like most eastern cities have around them. And (3) the newer master-planned stuff you have seen and think of as "macro" is more prevalent in newer cities (west and south), and where it happens out east, they're still somewhat smaller single projects (less land) and so far out, it's easy to miss them. In part, (3) is why the sprawl is worse by-the-numbers in the east - because it's less coordinated. But it's still there.

Lucky us, we have census data to sort this out. I'll stick with Colorado and PA, because I'm familiar with Denver, Aaron knows Pitt, and the two are pretty similar population-wise. Oh, and it's a Pitt thread.

Here are the census urbanized area maps for each, if you've never seen them:
Pitt: http://www2.census.gov/geo/maps/urba...ua69697_00.pdf
Denver: http://www2.census.gov/geo/maps/urba...ua23527_00.pdf

(Could look at the same for others if you wanted... Philly, Boulder, etc. since the best data is compiled by state.) Unfortunately, this level of detail still isn't out for 2010 census data. But we'll go with the most recent we have.

Let's compare state level data.

The easiest comparison is in the census urban/rural data. Within the urban data, they classify people as "in central place" or "not in a central place". That alone is fairly revealing, because many urbanized areas in PA are outside of central places- not something you see out west. (Ignore raw numbers - I am excluding other classifications that are pointless for this discussion, so it isn't going to add up.)

PA population "in urbanized area": 8,210,985
- In central place: 2,732,769 (33.3%)
- Not in central place: 5,478,216 (66.7%) (convenient numbers, eh?)

CO population "in urbanized area": 3,212,849
- In central place: 2,284,260 (71.1%)
- Not in central place: 928,589 (28.9%)

But here's the fun part - population densities (per square mile) within those:

PA in central place: 7,835 ppsm
PA not in central place: 1,765 ppsm

CO in central place: 4,086 ppsm
CO not in central place: 2,060 ppsm

That's very revealing, and matches what we'd expect I think. You see similar patters east-to-west all over the country. Western cities are much more compact, but at a lower, more uniform density. Eastern cities pack people in in the center, but then spread far and wide at drastically lower densities. Pitt does this too.

If I go now and look at this by MSAs, it gets more odd, because of the way MSAs are defined. There is a large number of "rural" population within MSAs, which is actually listed when you dive into the breakdowns, so you can pull it out. Within MSAs you have this hierarchy:

MSA:
>In central city
>Not in central city
-->Urban
---->In urbanized area
---->In urban cluster
-->Rural

If we just compare the "In central city" population density numbers versus the remaining "urban" numbers outside the central city, it looks like this:

PA in central city: 7,562 ppsm (25.8% of total PA metropolitan population)
PA urban (but not in central city): 1,747 ppsm (57.7% of total CO metropolitan population)
(the remaining 16+% of PA population inside MSAs doesn't meet urbanized density standards, and so is listed as rural)

CO in central city: 2,618 ppsm (41% of total CO metropolitan population)
CO urban (but not in central city): 2,630 ppsm (51.5% of total CO metropolitan population)

Same pattern. This shouldn't come as a surprise to us. Those folks out east who live in cities live in much denser cities than urban-dwellers out west. But in percentage terms, fewer folks out east actually live in those areas, and the rest are spread out farther and at lower densities. (MSAs still get distorted a little by land inside county boundaries, but this is pretty decent for analysis purposes.)

Even Pittsburgh, Aaron. You just don't see it, probably... it's so spread out, it doesn't look spread out, it looks almost rural. (That's been my observation in PA - the line between suburb and small town and rural is almost impossible to make out.)

So moral of the story, get some of those folks back into the core, Pittsburgh!

(I also looked at some individual townships, and it's the same thing. Pitt is nice and dense, but by the time you hit Penn Hills, McCandless Township, Monroeville, Ross Township, etc., you're into 1,500 - 2,200 ppsm. Whereas even our dreaded Aurora outside Denver manages to hang in at 2,800 ppsm, and Lakewood, Thornton, Arvada and the like are 3,100-3,600 ppsm. For the curious PA folks, Pittsburgh itself was around 6,000 ppsm, so in the state-level numbers the census churns out you get a sizable boost from Philly.)

I realize that this is a thread about Pittsburg but is 3,600 pppsm really any different then 1,500-2,200 ppsm range? Both are too low to properly support public transit. So what do you gain by having more people living closer together still having to drive themselves everywhere?

[bunt_q]

Originally Posted by Jelly Roll

I realize that this is a thread about Pittsburg but is 3,600 pppsm really any different then 1,500-2,200 ppsm range? Both are too low to properly support public transit. So what do you gain by having more people living closer together still having to drive themselves everywhere?

You use up half the land, they drive half as far. Yes, "traditional" 1980s suburban development is very different in character from later exurban development at half the density.

Sprawl is not all the same. New urbanists figured that out a decade ago. More and more planners are starting to figure it out (see books like "retrofitting suburbia"). But it's still very important and drastically under-studied. When 3/4-plus of our metro areas live in the areas you describe as "too low to properly support public transit," it might be worth your effort to figure out how to change/improve that in-place. Those burbs aren't going away.

Encouraging infill and moving people back into core cities is only half the battle; maybe even less than half the battle. The rest - improving the burbs - doesn't get enough attention on this forum, that's for sure. Denying or misrepresenting sprawl in cities isn't going to help anything. This isn't a big dick contest, it's about understanding what's actually there (not what we think is there, or what we want to be there), and then figuring out how to improve it.

It doesn't necessarily matter how those small towns developed initially. Although you misunderstand how midwest cities grew up, I think. Many suburbs out west were also small standalone towns that got swallowed up. It's the pattern of the "filling in" between these towns that's the problem. That's true in Pittsburgh too. And that also happened mostly post-WWII, even out east.

So when you guys say "just fewer people live in sprawl out east" that is objectively wrong. Pittsburgh proper is, what, 300,000, in a metro of 2 million+? That's about the same as everywhere else probably, outside of Sun Belt cities that annex everything in sight. So you'd better be looking at the suburbs, and looking at them closely, if you want to understand how that urban system is working. And until you understand it, you can't address it. The age old metric of "growth in metropolitan land area"/"growth in metropolitan population" is one you guys need to look at more closely. Just because a center city is growing/improving/thriving doesn't mean you're not still losing the regional battle.

I know the character of the surrounding suburbs is different in SW PA - they are older, have good cores of their own, etc. So if they still only have densities in the 2,000 ppsm range, you should be asking yourselves why. I would expect suburbs of that character to be denser than newer crap built last week, so if they're not... why? And maybe, if you have good old small towns around Pitt, you should be looking and building on that. A multi-polar metropolitan area makes more sense there than it does in post-WWII suburban edge-city hell.

[Centropolis]

I said there is more of a mass of sprawl (nothing about proportions, made no assumptions about how people move about in outer mega suburb Denver) in greater Denver than PGH. And of course I'll agree that it's tighter knit, since you are really interested in talking about that.

This is more than a petty regional-patriot war, there seems to be a concensus.

[bunt_q]

Not a regional war... just a comparison of two similarly sized areas with different growth patterns. I'm not passing judgment (although if you want me to, Denver's suburbs are "better" and Pittsburgh's core is "better" ...because I view denser as better, wherever that density happens to be).

This is the part I wasn't sure was true (and I'm still not).

Originally Posted by Centropolis

I said there is more of a mass of sprawl ... in greater Denver than PGH.

I suppose, if we figure that Pitt has ~2.1 million burbians at an average density of 2,400 ppsm... and Denver has ~2.5 million burbians at an average density of 3,200 ppsm... I guess that would tell us who has more mass of sprawl. Not that total mass is particularly relevant to anything. I'm just a stickler for specificity, that's all.


EDIT: I spent a great many hours back in the day trying to delineate urbanized area boundaries/densities from global remote sensing data for a climate change model we were part of... so I probably dwell on this crap more than the average person.

I am actually very excited to see the new urbanized area definitions from the 2010 census.

I have to go be productive, but I checked quick... on the question of "mass"... the census shows the urbanized area of Pittsburgh as 852 square miles (2,056 ppsm overall), per the last census. Denver's is 499 square miles (3,979 ppsm). Portland 474 square miles (3,340 ppsm), St. Louis 829 square miles (2,506 ppsm)... etc. etc. Tampa/St. Petes (802 sq mi, 2,570 ppsm) and St. Louis are almost identical. I think it's fun to look at this stuff, but it does not mean I am passing judgment on the core cities at all. Regional governance structures... yeah, that I might be passing judgment on. Holy nuts... New York is 3,352 square miles. LA 1,668. Monsters they are. Okay I'm done.

[Centropolis]

I say sprawl, I mean those massive vistas of 1990s boxes when you sit up on the higher topography to the east of metro Denver that's seems to go all the way to Ft Collins.

That's a fundamentally different experience than quasi exurban 1930s bungalows in the woods.

[Kingofthehill]

How easy is it to get around Pittsburgh without a car?

[bunt_q]

What if those 1990s boxes, as awful as they are, are built at equivalent densities to the 1930s bungalows? That's not actually impossible, or even uncommon.

Then it becomes an issue of connectivity/mix of land uses I'd say. But all other things being equal, you're still using half the land building at twice the density.

We don't have woods, sorry.

If you were building me the identical neighborhood, I'd choose the 1990s home amenities over 1930s any day (that's why new urbanist developments have been so successful). But that's not what you're building in townships with <3,000 ppsm, nor are those dominated by 1930s bungalows. Anchored maybe, but not dominated. If they are... if bungalows were ever really built at those low densities, hmm... I'd like to see that.

[Centropolis]

I'm sure that those 1990s boxes are built at a higher density than a theoretical 1930s bungalow in the woods. Those kinds of proto exurban properties are much more common in the east . I'm looking at all this from as much of an economic viewpoint as from that of a built environment. I'm looking at this from the viewpoint of how globalism - in a way - really influenced new non urban construction in the 90s, and how that's manifested in what you see when you travel down the street. The scale of it all.

My argument is that PGH doesn't have as much of that fat tree ring of that era of development.

[mhays]

A lot of Pittsburgh and its surrounding older centers is townhouses on very small sites. Back to my point about hillsides and ravines (and river bottoms, etc.), you might have a neighboring town with a core that was originally 15,000/sm and is now 8,000/sm due to reduced household sizes and some abandonment or reuse, but it's next to a forested hill that has scattered houses and a subdivision or two. An area of two miles square might have 10,000 people in it, but most live in one or two 8,000/sm zones within that, and much of the land is relatively natural, to the extent that's possible amidst so many people, mines, etc.

That's the weighted density point discussed in other threads. Sometimes that's more relevant than average density. The small town I'm describing could be fairly cohesive and give people decent access to a bus or train, etc.

[bunt_q]

Originally Posted by mhays

That's the weighted density point discussed in other threads. Sometimes that's more relevant than average density. The small town I'm describing could be fairly cohesive and give people decent access to a bus or train, etc.

Right, but even at pretty small scales, those burbs are lower density than that. All I'm saying is, there's a lot more cul-de-sac modern sprawl around there than folks are indicating. It's not all small lot townhouses on small sites separated by empty hills; they built new subdivisions in those hills.

This is no different from what we've built anywhere else (except lower density than any but the wealthiest of subdivisions I've seen) -

http://maps.google.com/?ll=40.533829...h&z=14&vpsrc=6

http://maps.google.com/?ll=40.286466...h&z=13&vpsrc=6

http://maps.google.com/?ll=40.453479...h&z=13&vpsrc=6

(Compare that first link above to the crap we build at the same scale - http://maps.google.com/maps?q=aurora...orado&t=h&z=14)

...

On the other point, globalization or not... local jurisdictions have always had it in their power to require different (non-sprawling) development forms. A choice we made all over, unfortunately.

[fflint]

bunt q has laid out a compelling argument refuting the notion greater Pittsburgh has little low-density suburban sprawl relative to the rest of the country.

[novawolverine]

It's not just Pittsburgh. Everyone knows that the development that we call suburban sprawl is denser in the West than in the East and Midwest. Pittsburgh is quite sprawly when you look at how large its sphere of influence is. The region does have a lot of older nodes spread out all over the place, and that is different compared to many places in the country. Topography plays a role in the development, but the sprawl itself is not much different than what you would find in New Jersey or Philly or whatever. To me, dense sprawl is only better if you can get around easily w/o a car. Otherwise, it's the same crap, IMO.

[Austinlee]

Originally Posted by bunt_q

Western cities are much more compact, but at a lower, more uniform density. Eastern cities pack people in in the center, but then spread far and wide at drastically lower densities.

Definitely. It pretty much comes down to cities that boomed pre-war versus post-war which approximately was the beginning of auto-centric development.

Originally Posted by mhays

I haven't seen much of suburban Pittsburgh, but I agree that hills make for a different dynamic. It's extremely hilly, not with big, orderly, shallow hills but with countless ravines and extremely steep slopes. It's like Seattle but much more so. The result is a lot of hillsides and ravines that aren't built on...I don't have numbers, but it could mean 20% or 30% of the land overall in sizeable swaths. You can build responsible densities on any given site, but still have average densities far lower.

This is very true too. Pittsburgh development has been hugely shaped by the topography.

Although, if it were more flat we would probably still have a differernt level of suburban density than western cities. So both statements are true.

[Jelly Roll]

Originally Posted by novawolverine

To me, dense sprawl is only better if you can get around easily w/o a car. Otherwise, it's the same crap, IMO.

I agree with this.

[Centropolis]

My argument was fairly esoteric, I admit. Huge, contiguous, dense(er) suburbs tend to take on a life of their own and I've always felt Pittsburgh was free(er) of that sort of "big suburb" scaled up soccer mom thing that was really birthed in the 80s and 90s than your average 2-3 million person metro area. I'm way more familiar with Denver than Pittsburgh, admittedly. I've only been to the 'Burgh twice. Cities that had OK economic engines during this time really got down with these kinds of big suburbs, I just assumed that Pittsburgh missed out on some degree of that "corporate-sprawl" (culdesacs, apartment complexes, best buy, six lane arterial ) stamp tool after the 70s. You know what I mean? I'd be interested to know how many 50,XXX + suburbs Pittsburgh has that don't have a pre-war central node. Whether or not metro Pittsburgh or city X is "dense," well I'm not interested in that argument. I could have guessed the answer to that one.

I'm sure that Pittsburgh has lots of stuff from the 30s-? that is quasi-exurban.

[mhays]

Those large burb swaths have been partially a function of metro population growth. Pittsburgh hasn't grown much.

[DBR96A]

In the Pittsburgh area are large swathes of undevelopable land that skew the population density down in the region. And even areas that are developable can still be very difficult.

[glowrock]

Originally Posted by bunt_q

Right, but even at pretty small scales, those burbs are lower density than that. All I'm saying is, there's a lot more cul-de-sac modern sprawl around there than folks are indicating. It's not all small lot townhouses on small sites separated by empty hills; they built new subdivisions in those hills.

This is no different from what we've built anywhere else (except lower density than any but the wealthiest of subdivisions I've seen) -

http://maps.google.com/?ll=40.533829...h&z=14&vpsrc=6

http://maps.google.com/?ll=40.286466...h&z=13&vpsrc=6

http://maps.google.com/?ll=40.453479...h&z=13&vpsrc=6

(Compare that first link above to the crap we build at the same scale - http://maps.google.com/maps?q=aurora...orado&t=h&z=14)

...

On the other point, globalization or not... local jurisdictions have always had it in their power to require different (non-sprawling) development forms. A choice we made all over, unfortunately.

Just figured I'd let you know that much of the land you see in those Google Map images that isn't developed is literally undevelopable, Brent. You know, steep hillsides, ravines, etc...?

Come out here, I'll show you around..

But anyhow, enough of this one. This is getting pretty old, quite frankly.

Aaron (Glowrock)