The US Census Bureau thread

[wwmiv]

Originally Posted by dimondpark

This is just estimates, not official census.

Yes. I am aware.

I am sure others, however, appreciate your clarification.

Would it have made it more obvious to you that I understood this basic fact if I had placed my entire post in quotation marks since they “are just estimates” and not “relevant” since “the counts” (as if they are not prone to other sources of error, particularly this go-round) are forthcoming?

Last I checked this was a general post vis-a-vis the Census Bureau, not a “2020 Census Count” thread, so I think they’re relevant numbers to post here regardless of how anyone feels about their merit.

[wwmiv]

Originally Posted by Crawford

These estimates are completely meaningless. The estimates only have value in years where there's no enumerated count, which we now have.

Census shouldn't even publish decennial year estimates. It's like posting an election poll after an election.

Actually, there are good reasons for doing so:

The release of Vintage 2020 estimates provides the opportunity to make comparisons with the 2020 Census to evaluate the accuracy of the estimates.

https://www.census.gov/programs-surv...estimates.html

Each decade, the Population Estimates Program has the opportunity to make comparisons between the final series of estimates for the decade and the latest decennial census to assess the accuracy of the estimates. Differences between the estimates and Census counts are typically interpreted as error in the estimates and are used to inform research and methodological improvements over the decade.

https://www.census.gov/programs-surv...estimates.html

They use the data for the same purpose many people here are going to do and enjoy doing: comparing the estimates to the actual count allows them to refine estimation techniques for better future accuracy. In other words, it is in iterative process.

[craigs]

The Census Bureau estimates San Francisco declined YOY from 881,549 residents to 866,606--a loss of 14,943 residents.

The city is still more populous than it was in 2010, but considering all the churn in employment and the residential outflow, it will be interesting to see if/when San Francisco can surpass the 2019 estimate.

[ChiSoxRox]

Originally Posted by Steely Dan

So the 2020 bureau estimate has Chicago down 17,955 (-0.66%) from it's official 2010 figure of 2,695,598.

Given that the 2020 estimate for IL was low by 225K, if even just 10% of that underestimate was in Chicago, then the city might not have lost any people at all. And considering the city of Chicago is ~21% of IL's population, and considering the CB's awful track record of estimating big messy urban cities, that's not a far-fetched proposition.

Here's to hoping!

Another hopeful undercount is St. Louis. The 2020 estimate is 297,645 -- it would be nice to see the city actually remain above 300k.

[Steely Dan]

Originally Posted by ChiSoxRox

Another hopeful undercount is St. Louis. The 2020 estimate is 297,645 -- it would be nice to see the city actually remain above 300k.

it would certainly be a nice psychological boost to see STL stay above 300K.

i looked up MO 2020 estimate vs. actual and is was one of the closest states in the nation. the CB only underestimated the state by 3,365 people, so let's hope all of the "missing" people are in st. louis!


looking at milwaukee, the CB 2020 estimate is 589,067.

WI as a state was underestimated by about 61K, so maybe an outside chance milwaukee got over 600K?

[dimondpark]

Originally Posted by craigs

The Census Bureau estimates San Francisco declined YOY from 881,549 residents to 866,606--a loss of 14,943 residents.

The city is still more populous than it was in 2010, but considering all the churn in employment and the residential outflow, it will be interesting to see if/when San Francisco can surpass the 2019 estimate.

Well, California will be adjusted and 200,000+ more people will be added. Let's see how many SF gets.

[SIGSEGV]

Originally Posted by wwmiv

Actually, there are good reasons for doing so:



https://www.census.gov/programs-surv...estimates.html



https://www.census.gov/programs-surv...estimates.html

They use the data for the same purpose many people here are going to do and enjoy doing: comparing the estimates to the actual count allows them to refine estimation techniques for better future accuracy. In other words, it is in iterative process.

there is academic value to them, and they should be released, but not publicized in a way that makes them sound legitimate.

[tech12]

Originally Posted by RST500

For the Bay Area Asians are rapidly growing with Latinos declining but this is prior to the remote work exodus:

"Comparing the newly released 2019 Census Household Numbers vs 2017 Census on:
Race of households in the 5-county Bay Area.

White-only Households Downwards 19,200
Black-only Households Upwards 1,300
Latino-only Households Downwards5,700
Asian-only Households Upwards 29,600
Multiracial Households: Downwards 4,800"

There are 9 counties in the Bay Area:

San Francisco
Alameda
Contra Costa
San Mateo
Santa Clara
Solano
Sonoma
Marin
Napa

By "5-county Bay Area" I assume you mean the SF-Oakland MSA.

[dimondpark]

Originally Posted by tech12

There are 9 counties in the Bay Area:

San Francisco
Alameda
Contra Costa
San Mateo
Santa Clara
Solano
Sonoma
Marin
Napa

By "5-county Bay Area" I assume you mean the SF-Oakland MSA.

Yeah, there is no 5-county Bay Area, it's the 9-county Bay Area to locals and the 14 county SJ-SF-O Combined Statistical Area. I remember when people had a hard time accepting that Santa Cruz was part of the CSA, hell now we have to reconcile having Merced and Stanislaus

This is why it's important to not look at just the SF MSA when it comes to demographic shift in the region because data shows most movement out of the 5-county SF MSA is to neighboring MSAs--especially when it comes to the movement of Blacks in the past 20 years.

[wwmiv]

Originally Posted by SIGSEGV

there is academic value to them, and they should be released, but not publicized in a way that makes them sound legitimate.

Can you be constructive instead of just tearing me down? In what way did I make them sound more legitimate than they are? And what could I have done and so in the future instead to make you happier with the way I talk?

I already made reference to the fact that they are estimates, multiple times actually. I’ve made reference to their academic value and their limitations. Others have as well. I’m sorry you feel that I am “legitimizing” the numbers... I don’t know what more I can do. This is why I’ve mostly stopped participating here, one of the only places I used to feel useful and included and at home... every direct response to me is just plain a tear down and there’s nothing constructive at all. It’s just rude.

[SIGSEGV]

Originally Posted by wwmiv

Can you be constructive instead of just tearing me down? In what way did I make them sound more legitimate than they are? And what could I have done and so in the future instead to make you happier with the way I talk?

I'm not trying to tear you down. Sorry if it sounded like that. I'm trying to tear down the Census Bureau for poor messaging.

[wwmiv]

Originally Posted by SIGSEGV

I'm not trying to tear you down. Sorry if it sounded like that. I'm trying to tear down the Census Bureau for poor messaging.

Well when you quote someone and don’t actually say that, it usually just comes across as directed at the person you’re quoting.

[SIGSEGV]

Originally Posted by wwmiv

Well when you quote someone and don’t actually say that, it usually just comes across as directed at the person you’re quoting.

you're right, I should have been more clear! wasn't trying to shoot the messenger!

[galleyfox]

Originally Posted by wwmiv

Can you be constructive instead of just tearing me down? In what way did I make them sound more legitimate than they are? And what could I have done and so in the future instead to make you happier with the way I talk?
.

Sorry that you got caught up in our Northerner angst.

Especially for Chicagoans, it’s been a decade of the media trumpeting Chicago’s population loss-which suddenly looks like modeling error.

But thanks for the link.

Some pretty interesting stuff in the data even accounting for the Census whiff.

Illinois 2010 Census: 12,830,632
Illinois 2020 Census: 12,812,508
Net Population: -18,124 // -0.14%

Illinois 2010 Census: 12,830,632
Illinois 2020 Est: 12,587,530
Net Population: -243,102 // -1.89%

Cook County 2010: 5,194,675
Cook County 2020 est: 5,108,284
Net Population: -86,391 // -1.66%

Cook County wo Chicago 2010: 2,499,077
Cook County wo Chicago 2020: 2,430,641
Net Population: -68,436 // -2.74%

Chicago 2010: 2,695,598
Chicago 2020 est: 2,677,643
Net Population: -17,955 // -0.67%

[ChiSoxRox]

Intriguingly, Washington DC was *over*estimated.

2020 Census - 689,545

2020 Estimates - 712,816

[Ant131531]

Originally Posted by wwmiv

https://www.census.gov/programs-surv...estimates.html

1. New York City: 8,253,213 / MSA: #1. 19,124,359
2. Los Angeles: 3,970,219 / MSA: #2. 13,109,903
3. Chicago: 2,677,643 / MSA: #3. 9,406,638
4. Houston: 2,316,120 / MSA: #5. 7,154,478
5. Phoenix: 1,708,127 / MSA: #10. 5,059,909
6. Philadelphia: 1,578,487 / MSA: #8. 6,107,906
7. San Antonio: 1,567,118
8. San Diego: 1,422,420
9. Dallas: 1,393,266 / MSA: #4. 7,694,138 / MSA Division: 5,171,934
10. San Jose: 1,013,616 / CSA: 9,608,006

That is the second YOY decline in estimates from 2018 peak of 1,034,877 for San Jose.

... 11. Austin: 995,484
... 12. Fort Worth: 927,720 / MSA: #4. see above / MSA Division: 2,522,204
... 13. Jacksonville: 920,570

The metropolitan areas for the others are significantly smaller at 3,332,427 (San Diego, #17), 2,590,732 (San Antonio, #24), 2,295,303 (Austin, #29), and 1,971,161 (San Jose, #35). All these are the same rank as prior. A notable shift is that Vegas metro has surpassed Pittsburgh metro, switching numbers 27 and 28. Austin is just behind that, while Nashville metro is poised to overtake San Jose. Jacksonville metro brings up the rear at #39, outgrowing Milwaukee metro: 1,587,892.

As far as city propers, Jacksonville has been outgrown by Fort Worth in the last year. Ha. So many jokes. Austin is poised to overtake San Jose as the most populated tech city, Dallas is poised to leapfrog San Diego, and San Antonio is poised to leapfrog Philadelphia. How long until Houston overtakes Chicago? Until Fort Worth overtakes San Jose? That’s some major advancement for a selection of cities all in one state at the same time.

5 of the 13 American municipalities which are already at or could plausibly soon reach one million are in Texas alone and another 3 are in California (with a smattering of more Texan and Californian cities scattered throughout the list of largest municipalities just beneath that). When metro areas and cities are viewed in composite like above, both California and Texas stand out even more.

Here are the populations of some other major metropolitan areas (up to 13, like I did with cities):

#6. Washington: 6,324,629 / City: #20. 712,816
#7. Miami: 6,173,008 / City: #42. 471,525
#9. Atlanta: 6,087,762 / City: #37. 512,550
#11. Boston: 4,878,211 / City: #21. 691,531
#12. San Francisco: 4,696,902 / City: #17. 866,606 / CSA: 9,608,006
#13. Inland Empire: 4,678,371

These metro areas retain the same rank, but there is some churn for their anchor cities: Indianapolis (#16, 887,756) has grown larger than San Francisco but not Charlotte (#15, 900,350). Atlanta is nipping on the heels of Sacramento (#36, 512,818). No one city in Inland Empire predominates, although Riverside is first among equals (#58, 330,786).

Feel free to add more detail. Seattle, Detroit, Twin Cities, etc.

Cool so according to their estimates, Atlanta city proper still grew about 6k. Not bad considering Covid.

[iheartthed]

I looked at some numbers based on the state population growth to make a prediction about where some cities will land once the actual census numbers are released. I'm not going to put actual numbers to many of them to avoid that type of conversation, but here are my general guesses.

The Good

NYC looks in good shape. I'd be surprised if the actual population is below 8.5M. My official guess is 8.71M for NYC.

Texas - Houston, Dallas, and San Antonio will probably come in close to the ACS estimates. However, I wouldn't be surprised if Austin is a slight miss. I think it's certainly not going to hit 1m this census.

Miami - Prospects for growth are good. Like Austin, the ACS estimates might be a little too rosy, but growth seems certain.

Possible surprises in the older midsize cities - Baltimore, Cleveland, St. Louis, and Milwaukee might actually pulled out of their decades long population declines.

Pacific Northwest - The major cities in the Northwest, Seattle and Portland, both appear to continue to grow.

The Bad

I think the slower than usual growth in the Sun Belt is going to introduce a few surprises at the city level. North Carolina's cities (Raleigh and Charlotte) probably didn't grow as fast as projected. Ditto for Nashville. And, I wouldn't be surprised to see a small population drop in Atlanta.

The Ugly

Chicago and California. Especially L.A. and Chicago. L.A. seems on track for a first ever population loss. I see a small loss for San Francisco as well, but it's close enough that it could also be a small increase.

Chicago will also probably show a loss, which shouldn't be too controversial considering that Illinois was the only state to lose population. But Chicago's population losses might be as bad as what happened last decade. Or worse. And Houston might have overtaken it as America's third largest city.

[SIGSEGV]

Originally Posted by iheartthed

I looked at some numbers based on the state population growth to make a prediction about where some cities will land once the actual census numbers are released. I'm not going to put actual numbers to many of them to avoid that type of conversation, but here are my general guesses.

The Good

NYC looks in good shape. I'd be surprised if the actual population is below 8.5M. My official guess is 8.71M for NYC.

Texas - Houston, Dallas, and San Antonio will probably come in close to the ACS estimates. However, I wouldn't be surprised if Austin is a slight miss. I think it's certainly not going to hit 1m this census.

Miami - Prospects for growth are good. Like Austin, the ACS estimates might be a little too rosy, but growth seems certain.

Possible surprises in the older midsize cities - Baltimore, Cleveland, St. Louis, and Milwaukee might actually pulled out of their decades long population declines.

Pacific Northwest - The major cities in the Northwest, Seattle and Portland, both appear to continue to grow.

The Bad

I think the slower than usual growth in the Sun Belt is going to introduce a few surprises at the city level. North Carolina's cities (Raleigh and Charlotte) probably didn't grow as fast as projected. Ditto for Nashville. And, I wouldn't be surprised to see a small population drop in Atlanta.

The Ugly

Chicago and California. Especially L.A. and Chicago. L.A. seems on track for a first ever population loss. I see a small loss for San Francisco as well, but it's close enough that it could also be a small increase.

Chicago will also probably show a loss, which shouldn't be too controversial considering that Illinois was the only state to lose population. But Chicago's population losses might be as bad as what happened last decade. Or worse. And Houston might have overtaken it as America's third largest city.

Not at all obvious to me that Chicago will show a loss. It could, but I think downstate places like Danville or Decatur are more likely to account for a larger share of the Illinois population loss. The 2020 estimate has both of them 9% down in population from 2010, while Chicago is "only" 0.7% down, so relatively speaking, they ought to do worse. I think most likely Chicago will be effectively flat, maybe gaining or losing a few tenths of a percent.

Keep in mind that the 2020 estimate has Illinois losing 243,000 people while in actuality only 18,000 people were lost in the actual 2020 census. Given how wrong the estimate was, I'm not sure we can trust the relative numbers in the estimate too much either.

[Austin55]

Hopefully, this is legible...


https://docs.google.com/spreadsheets...it?usp=sharing

[subterranean]

Shouldn’t those be ranked by 2020 population?