agreeed, it’s like a freeway around almost every border of Wake county. massive
1 Like
I think that Charlotte’s loop somehow stays fully within Mecklenburg, despite that county being physically smaller than Wake. I’ll have to check out that statement on google maps now, in case I have to eat those words.
I-485 stays in Mecklenburg County.
7 Likes
Yep. As soon as I posted my comment, I ran to Google Maps to verify. Clearly they were very careful to NOT leave the county.
1 Like
Incoming data for all of of data nerds.
No surprise that Raleigh is more densely populated than Charlotte, Nashville, Jacksonville, Oklahoma City, Virginia Beach, and others, but surprising that we are more densely populated than Austin, El Paso and New Orleans.
https://filterbuy.com/resources/most-and-least-densely-populated-cities/
12 Likes
This was interesting BUT!
They start by showing us maps of proportion of population in states that live inside Urbanized Areas/Urban Clusters (UA/UC), but then they use city limits to make their lists of densest and least dense urban areas. You can’t change the definitions halfway through your Results section, that’s bad data science! I expect more from an air filter company!
I was mainly aggravated that they’d say Anchorage is so non-dense, because Alaska doesn’t even acknowledge city limits in the same way that we do in the lower 48, so Anchorage and Juneau appear to be rural areas even though both are pretty densely developed (Juneau very much so). Plus it’s not fair to always call Jacksonville so non-dense. I get it, but most of the land in Duval County isn’t developed and shouldn’t be included. Furthermore, I hate the way Cambridge MA always shows up on these lists, as if it’s not just a small subsector of the greater Boston urban area. If the area within a few blocks around NCSU were a separate city it’d look pretty dense, too, but it still wouldn’t be separate from the greater Raleigh urban area.
Instead, if we’re really comparing density, we should look exclusively at the developed parts of our cities, not the outlying undeveloped acreage that happens to be inside a city’s incorporated boundary or in the outlying counties of a metro area .
So I went and took all the data on UAs and UCs from the 2010 Census (most recent), which includes population and area and therefore density information, and I made a list based on that. Too cold to go outside yet so I might as well do something else, right? Plus I have more important work I want to put off.
Top 5 Giant (over 2.5M in UA)
LA-Anaheim
San-Francisco-Oakland
New York
Miami
San Diego
Bottom 5 Giant
Philly
Minneapolis-St Paul
Boston
Atlanta
Detroit
Top 10 Large (1M-2.5M in UA)
San Jose
Salt Lake City
Sacramento
Denver
Riverside-San Ber
Portland OR
San Antonio
Hampton Roads
Columbus
Austin
Bottom 10 Large
Kansas City
Providence
Memphis
Indianapolis
Cincinnati
Jacksonville
Pittsburgh
Las Vegas
Charlotte
Baltimore
Top 10 Medium (500k-1M in UA)
Honolulu
(Mission Viejo CA)
Fresno
Bakersfield
New Orleans
El Paso
(Concord CA)
Colorado Springs
Albuquerque
Omaha
Ogden UT
Buffalo
Tucson
Bottom 10 Medium
Springfield MA
Akron
Nashville
Raleigh
Baton Rouge
Fort Myers
Columbia
Birmingham
Knoxville
Grand Rapids
Top 10 Small City (100k-500k in UA)
(Santa Maria CA)
Oxnard CA
Stockton
(Simi Valley CA)
Vallejo-Fairfield
Modesto
Salinas-Monterey
El Centro-Calexico
Laredo
Boulder
Santa Barbara
Bottom 10 Small City
Myrtle Beach
(Rock Hill SC)
Johnson City TN
Ann Arbor MI
Asheville
Gainesville GA
Spartanburg SC
Kingsport-Bristol
Barnstable Town MA
Muskegon MI
Hickory-Morganton-Lenoir
The list is… disappointing. Several places we don’t think of as dense show up in the listing as quite dense, and vice versa. I see three reasons for this.
One, since the definition of UA/UC is invariable across geographies, all UAs and UCs have a fairly narrow band of density, with an average around 1700 p/mi2, but a range only from about 500 to 6000. San Francisco city limits has a density over 18000 p/mi2, so we have to acknowledge that the numbers we’re used to seeing in terms of density don’t apply here.
Two, as I mentioned in an earlier post, the UA/UC definitions suck in some places. They suck in two different ways, though. First there’s the Hops/Skips/Jumps system the Census Bureau uses to decide whether a chunk of land should be included in a UA/UC or not. So a UA/UC boundary can jump over something like a body of water (of a certain size) to include the dense development on both sides–especially important in the case of San Francisco-Oakland, but relevant to places like St Louis and New Orleans and Evansville, too. This is good. UA/UC definitions can also include areas with no population that are nonetheless densely developed: this makes sense in the case of big industrial/warehouse developments in the middle of urban areas, but it can mean that an office park on the edge of a wildnerness gets included, too, which may or may not be good depending on your point of view. It is also how Hickory is allowed to connect to Morganton and Lenoir, because fast food restaurants and gas stations along highways can be roped into a UA through this definition. This is a problem for the definitions that doesn’t look likely to be changed. Airports can also be included in UA/UC definitions, which means a UA can jump over an airport to get at the development on the other side (good), but the definition now includes the land area of the airport, decreasing the UA’s density without actually changing the perception of density to an observer on the ground (bad). This is especially problematic in a case where a large airport is in the UA of a small city–Kinston, for example.
The other way the definitions suck is, as I noted in an earlier post, sometimes politics has gotten in the way of the definitions, so that Greensboro and High Point are considered separate UAs even though no one in their right mind would perceive them as distinct and separate on the ground from an urban development perspective. (Same for LA and Riverside, San Jose and Frisco, Charlotte and Concord…)
Third issue is the way land is developed based on climate and environment. In the southwest and high plains, cities are surrounded by desert or farmland, so their UAs are compact and don’t include long skips and jumps to collect nearby residential areas. In the east and south, people are more spread out, and the skips and jumps take in lots of marginally urban land, depressing their apparent density. Thus western and plains cities we don’t think of as dense–San Jose, Tucson–are actually denser than we think when we consider that they remain moderately dense to their edges, unlike southern cities that gradually peter out over a longer suburban area.
Anyway if you read this whole mess… thanks I guess? I don’t think I can draw any useful conclusions about what places are denser and what are less so based on this, but maybe? If anyone really wants to geek out about it I’ll send you my dataset. 
8 Likes
If you want more up to date UA data, you an go to Raleigh, NC Urbanized Area - Profile data - Census Reporter
This link uses the ACS survey data from 2019.
While the filterbuy report does start with the Urbanized areas as measured by UAs and Urban Clusters, the rest of the data is based on actual city limits, within which (and not surprisingly) several large land area cities’ “size” stories get exposed as the ruses that they sometimes are. When I say that, I am talking about cities like Jacksonville, FL, Virginia Beach, and Oklahoma City, which greatly benefit from their enormous land areas when ranked by city proper.
By UA, Raleigh actually holds its own considering that practically all of the Triangle’s suburban towns are included in its UA, while Durham’s only incorporates Chapel HIll (not really a suburb in the traditional sense), and HIllsborough.
Some of the South’s legacy cities show their structural roots through their UA data like New Orleans & Hampton Roads, while others like Birmingham, Richmond, Memphis, Charleston, and Louisville don’t.
Among the cities below, Charlotte is interesting for 2 reasons: 1) By the map, it would seem to me that the UAs of both Gastonia and Concord should be consolidated with Charlotte. This would increase Charlotte’s UA population by over half a million people, but…2) doing so would decrease their overall UA density: which is already slightly lower than Raleigh’s.
Below is a quick look at UA densities among UAs with over a million in the Southeast. For this, I am not including south of Jacksonville, FL, Texas/Oklahoma, or D.C. northward. From most dense to least, and grouped by similar densities, they are:
New Orleans: 3807.6 ppl/m2
Hampton Roads: 2882.7 ppl/m2
Jacksonville: 2194.7 ppl/m2
Memphis: 2156.5 ppl/m2
Louisville: 2106.7 ppl/m2
Richmond: 2080.7 ppl/m2
Raleigh: 2072.2 ppl/m2
Charlotte: 2059.3 ppl/m2
Atlanta: 1947.1 ppl/m2
Nashville: 1919.1 ppl/m2
Other densities of notable Southern/NC UAs below are:
Charleston (SC): 2200.6 ppl/m2
Durham: 2152.5 ppl/m2
Wilmington: 1946.7 ppl/m2
Greensboro: 1853.7 ppl/m2
Fayetteville: 1682.7 ppl/m2
Baton Rouge: 1650.5 ppl/m2
Columbia: 1604.1 ppl/m2
Birmingham: 1422.6 ppl/m2
Knoxville: 1377.8 ppl/m2
Winston-Salem: 1290.7 ppl/m2
Asheville: 1180.23 ppl/m2
7 Likes
Thank you for this detailed explanation of UAs. The nuance about airports, bodies of water, and commercial areas is helpful and also sheds light on the density rankings.
New Orleans as a city is indeed much denser than many Southeast cities. The density of the UA benefits greatly from compactness related to the bodies of water surrounding the city.
2 Likes
Interim Census Bureau director sent a memo out last night. Apportionment data–state-level only–is set for release on 30 April (which is four months late but pandemic/politics/etc). No update on when finer-grained data will be out, which is required for actual redistricting, BUT, per my friend who works there, they won’t release the apportionment numbers if they aren’t satisfied with their accuracy/cleanliness, so releasing county/city/census block data should occur fairly soon thereafter, probably before 4th of July.
7 Likes
Si we won’t know Raleigh pop before July 4th, or if metro areas will be recombine before then correct!!!
I’m not sure metro redefining is done anytime soon. 2023 maybe?
2 Likes
Correct me if I am wrong everyone, but the cycle is usually 5 years, and the next opportunity for reunification is 2023.
You’re right. Census will release new Urban Area/Cluster definitions probably late 2022, and OMB will release new Metro/Micro/Combined area definitions probably May 2023.
All of this could go out the window if we end up with a Census redo. But the county and city population/housing counts should come this summer regardless.
2 Likes
I think will be at a solid 500,000for Raleigh which means more federal funding!!! But yeah I have given census my complaints.
I wouldn’t hold your breath for 500,000, but I hope that I’m wrong. I’ll be happy with 485K, but that might be high as well.
Good explanation of the long process for the 2020 Census
5 Likes
This evening (buried in the Friday night news cycle) the Census Bureau finally broke down and admitted what we all knew: they’re not going to get the data out in time.
The new official (not constitutional) deadline for the apportionment data itself–that is, just how many seats each state will be entitled to–is now 30 April.
The expected delivery date (the Bureau says “deadline” but they invented it so it’s not really a deadline; Congress can’t impose one, either) for all other redistricting data–that is to say, population and housing counts with race data (and I assume age and sex, but I’m not actually sure) down to the census block level–is now 30 September.
I’ve never known the Census Bureau to release anything prior to their deadline, so don’t focus on that word. It’s a scheduled release. So that’s how long we have to wait to find out whether Raleigh officially hit 500k or not.
In prior years the Census has released redistricting results piecemeal, a few states at a time over several weeks. The plan is to release all data in a single gigantic chunk on 30 September. There are lots of issues here, from some states having constitutional mandates that redistricting must be completed by the end of the summer, to the obvious indication that the Bureau is aware of exactly how bad their data is and are likely building a case for a re-do in maybe 2022 or 2023. But anyway, this is the latest update from the horse’s mouth.
4 Likes
Raleigh was under 475,000 in the last Census estimate for 2019. Unless those estimates were woefully short over the remainder of the decade, I just don’t see how the city hits 500,000 with this Census.
Estimates are calculated differently than the census. When Atlanta famously saw a discrepancy of 120,000 people between its 2009 estimates and 2010 census, it wasn’t because all those people left in a year. It was because they never existed and the estimates starting circa 2004 got out-of-control due to faulty assumptions.
Raleigh’s census population could be about +/- 10% from the 2019 estimates and no one should bat an eye as that’s probably reasonable. Atlanta’s 22% overestimate is the outlier.
4 Likes
If I remember correctly, Raleigh’s Census number in 2010 fell slightly short of its 2009 estimate. I am just trying to temper unchecked enthusiasm, so that we aren’t overly deflated. I’d love for Raleigh to pass 500K in the Census count, but I’m not expecting it at all.
2 Likes