News:

The AARoads Wiki is live! Come check it out!

Main Menu

2018 city population estimates

Started by golden eagle, June 16, 2019, 08:24:21 PM

Previous topic - Next topic

DTComposer

Quote from: bandit957 on June 17, 2019, 10:00:11 AM
A ranking of metropolitan areas might seem more accurate. Cleveland basically has fixed boundaries, but Columbus was able to annex all over the place. Cities in Texas can also annex, but in New England, they can't.

It's less common for cities to deannex, though I think Memphis is deannexing that huge undeveloped area on the southwest.

The problem with metropolitan areas though is that parts of them seem to be not really what people would consider part of the city that they are centered on. Is northern Kentucky part of Cincinnati or not? Should the census bureau just rank places that are city-sized but not necessarily coterminous with actual ctiies (like how they do CDP's for unincorporated areas)?

Metropolitan areas are also tricky, since they're county-based. So you get the San Bernardino County problem, where Needles is part of the Riverside metro area, even though it's 150+ miles of desert away from even the exurbs of the urban area.

On the other hand, using Census-defined Urban Areas are also problematic, since physical geography might separate two areas that are deeply tied together socially/culturally.

In my opinion, all Census-defined areas are problematic, since their primary data point is commuting patterns. This may have worked in 1960, when most households were single-income, and most people commuted to a commercial core for their job. With the rise of multiple-income households (including multiple incomes from one person), the "gig economy," telecommuting and hyper-commuting, this metric is a lot less clear than it used to be. Also, it ignores the rest of people's lives: where they go for shopping, entertainment and leisure, etc.

The Bay Area is a good example of this. The "real world" definition has always been the nine counties that touch San Francisco Bay (and many state and local agencies use this definition), and the region is very close-knit from a social, cultural, economic, and media standpoint. But there is no Census-defined area (or combination of areas) that contains just those nine counties. Therefore, the data presents the Bay Area as much smaller than it really is (using the San Francisco-Oakland Urban Area) or much larger than it really is (using the San Jose-San Francisco-Oakland Combined Statistical Area).


bandit957

Quote from: DTComposer on June 17, 2019, 07:06:08 PM
Metropolitan areas are also tricky, since they're county-based. So you get the San Bernardino County problem, where Needles is part of the Riverside metro area, even though it's 150+ miles of desert away from even the exurbs of the urban area.

On the other hand, using Census-defined Urban Areas are also problematic, since physical geography might separate two areas that are deeply tied together socially/culturally.

It might be more accurate if they come up with metropolitan areas that use townships and MCD's instead of counties. They already do this for New England as an alternative to the county-based areas.
Might as well face it, pooing is cool

kevinb1994

Quote from: bandit957 on June 17, 2019, 07:11:22 PM
Quote from: DTComposer on June 17, 2019, 07:06:08 PM
Metropolitan areas are also tricky, since they're county-based. So you get the San Bernardino County problem, where Needles is part of the Riverside metro area, even though it's 150+ miles of desert away from even the exurbs of the urban area.

On the other hand, using Census-defined Urban Areas are also problematic, since physical geography might separate two areas that are deeply tied together socially/culturally.

It might be more accurate if they come up with metropolitan areas that use townships and MCD's instead of counties. They already do this for New England as an alternative to the county-based areas.

Don't forget CCDs.

Brandon

Quote from: bandit957 on June 17, 2019, 07:11:22 PM
Quote from: DTComposer on June 17, 2019, 07:06:08 PM
Metropolitan areas are also tricky, since they're county-based. So you get the San Bernardino County problem, where Needles is part of the Riverside metro area, even though it's 150+ miles of desert away from even the exurbs of the urban area.

On the other hand, using Census-defined Urban Areas are also problematic, since physical geography might separate two areas that are deeply tied together socially/culturally.

It might be more accurate if they come up with metropolitan areas that use townships and MCD's instead of counties. They already do this for New England as an alternative to the county-based areas.

Not everywhere uses or has townships.
"If you think this has a happy ending, you haven't been paying attention." - Ramsay Bolton

"Symbolic of his struggle against reality." - Reg

NWI_Irish96

Quote from: SP Cook on June 17, 2019, 09:59:30 AM
BTW, the major point the you  usually get out of census estimates is just how wrong they are when the actual count is done in two year.  Much like political polls.

Would love to see the data to support your claim of how wrong the annual estimates are.
Indiana: counties 100%, highways 100%
Illinois: counties 100%, highways 61%
Michigan: counties 100%, highways 56%
Wisconsin: counties 86%, highways 23%

Max Rockatansky

Quote from: cabiness42 on June 18, 2019, 11:50:33 AM
Quote from: SP Cook on June 17, 2019, 09:59:30 AM
BTW, the major point the you  usually get out of census estimates is just how wrong they are when the actual count is done in two year.  Much like political polls.

Would love to see the data to support your claim of how wrong the annual estimates are.

One real world example is Phoenix.  The population estimates had the city passing Philadelphia but came way down by the time the official census was done in 2010.  I would attribute that instance to a lot of people not claiming Phoenix as their primary residence (Snow Birds).   

NWI_Irish96

Quote from: Max Rockatansky on June 18, 2019, 01:11:29 PM
Quote from: cabiness42 on June 18, 2019, 11:50:33 AM
Quote from: SP Cook on June 17, 2019, 09:59:30 AM
BTW, the major point the you  usually get out of census estimates is just how wrong they are when the actual count is done in two year.  Much like political polls.

Would love to see the data to support your claim of how wrong the annual estimates are.

One real world example is Phoenix.  The population estimates had the city passing Philadelphia but came way down by the time the official census was done in 2010.  I would attribute that instance to a lot of people not claiming Phoenix as their primary residence (Snow Birds).   

That doesn't make the annual estimates wrong.  The annual estimates are based on ACS data, which uses different residency criteria (currently staying at address 2+ months) than the census (where you live most of the year).  A snowbird who lives in Phoenix between 2 and 6 months would correctly be counted if they received an ACS form while in Phoenix but correctly not counted if they received their census form while in Phoenix. 

If you are looking purely at the data for data's sake, it creates an inaccuracy. If you are looking at the data from an urban planning standpoint, getting the data both ways is helpful for Phoenix to understand the number of snowbirds.
Indiana: counties 100%, highways 100%
Illinois: counties 100%, highways 61%
Michigan: counties 100%, highways 56%
Wisconsin: counties 86%, highways 23%

ftballfan

Quote from: Brandon on June 18, 2019, 07:09:01 AM
Quote from: bandit957 on June 17, 2019, 07:11:22 PM
Quote from: DTComposer on June 17, 2019, 07:06:08 PM
Metropolitan areas are also tricky, since they're county-based. So you get the San Bernardino County problem, where Needles is part of the Riverside metro area, even though it's 150+ miles of desert away from even the exurbs of the urban area.

On the other hand, using Census-defined Urban Areas are also problematic, since physical geography might separate two areas that are deeply tied together socially/culturally.

It might be more accurate if they come up with metropolitan areas that use townships and MCD's instead of counties. They already do this for New England as an alternative to the county-based areas.

Not everywhere uses or has townships.
Most states that use townships are in the Northeast and Midwest

kalvado

Quote from: cabiness42 on June 18, 2019, 11:50:33 AM
Quote from: SP Cook on June 17, 2019, 09:59:30 AM
BTW, the major point the you  usually get out of census estimates is just how wrong they are when the actual count is done in two year.  Much like political polls.

Would love to see the data to support your claim of how wrong the annual estimates are.
If you google "syracuse NY population", google comes up with the graph of few upstate cities (you may click "explore more" under the graph and add more cities). There are noticable bumps around 2000 and 2010 for upstate cities estimates, either up or down. Not huge, but noticable.
I assume you can do the same for cities near you.

webny99

#34
Quote from: kalvado on June 19, 2019, 10:55:03 AM
Quote from: cabiness42 on June 18, 2019, 11:50:33 AM
Quote from: SP Cook on June 17, 2019, 09:59:30 AM
BTW, the major point the you  usually get out of census estimates is just how wrong they are when the actual count is done in two year.  Much like political polls.
Would love to see the data to support your claim of how wrong the annual estimates are.
If you google "syracuse NY population", google comes up with the graph of few upstate cities (you may click "explore more" under the graph and add more cities). There are noticable bumps around 2000 and 2010 for upstate cities estimates, either up or down. Not huge, but noticable.
I assume you can do the same for cities near you.

Yes, the changes in the census year are almost always way bigger than the other year-on-year changes, sometimes erasing or reversing the gain or loss for the entire ten year period.
Using Rochester as an example, every year between 2000 and 2010 showed a small loss. Then all of a sudden in 2010 there was a gain of around 2,000 people. Same with the 1990's; losses posted every year but then a gain of around 5,000 in 2000.
I have seen this across many cities in many areas of the country.

DTComposer

Quote from: cabiness42 on June 18, 2019, 11:50:33 AM
Quote from: SP Cook on June 17, 2019, 09:59:30 AM
BTW, the major point the you  usually get out of census estimates is just how wrong they are when the actual count is done in two year.  Much like political polls.

Would love to see the data to support your claim of how wrong the annual estimates are.

From my earlier post:

Quote from: DTComposer on June 17, 2019, 06:48:27 PM
The original 2017 estimate for New York was 8,622,698. With the release of the 2018 data, they revised the 2017 number down to 8,438,271. So without explanation, they decided they were off by nearly 200,000 people last year.

So either the 2017 number was way off, or the 2018 number will prove to be way off, or perhaps both.

Flint1979

In Michigan it's either City or Township. All of the villages are a part of a Township.



Opinions expressed here on belong solely to the poster and do not represent or reflect the opinions or beliefs of AARoads, its creators and/or associates.