Please register to participate in our discussions with 2 million other members - it's free and quick! Some forums can only be seen by registered members. After you create your account, you'll be able to customize options and access all our 15,000 new posts/day with fewer ads.
The city of Atlanta grew (from 505K in April 2022 to 520K in April 2023) and is the highest growth rate in the region at 2.8%, followed by Cherokee County at 2.5%.
The city of Atlanta grew (from 505K in April 2022 to 520K in April 2023) and is the highest growth rate in the region at 2.8%, followed by Cherokee County at 2.5%.
Totally unreliable stats coming from local-yokel planning commissions and other ATL boosters. They're notorious for inflating the numbers to suit local expectations (and local pride). The 2022 US Census estimate was 499,127, up 0.1% from the official 2020 census of 498,715, but shy of 500K. Better to wait till next March for 2023 US Census estimates.
Totally unreliable stats coming from local-yokel planning commissions and other ATL boosters. They're notorious for inflating the numbers to suit local expectations (and local pride). The 2022 US Census estimate was 499,127, up 0.1% from the official 2020 census of 498,715, but shy of 500K. Better to wait till next March for 2023 US Census estimates.
This begs the following question: how accurate has the ARC's population estimate model been in the past? Let's compare the ARC's 2020 population estimates to the findings of the 2020 Census:
Cherokee County:
269,500 ARC 2020
266,620 Census 2020
Result: ARC overestimate of 2,880, 1.1% error
Clayton County:
288,900 ARC 2020
297,595 Census 2020
Result: ARC underestimate of 8,695, 2.9% error
Cobb County:
774,600 ARC 2020
766,149 Census 2020
Result: ARC overestimate of 8,451, 1.1% error
DeKalb County:
758,230 ARC 2020
764,382 Census 2020
Result: ARC underestimate of 6,152, 0.8% error
Douglas County:
147,000 ARC 2020
144,237 Census 2020
Result: ARC overestimate of 2,763, 1.9% error
Fayette County:
119,400 ARC 2020
119,194 Census 2020
Result: ARC overestimate of 206, 0.2% error
Fulton County:
1,049,470 ARC 2020
1,066,710 Census 2020
Result: ARC underestimate of 17,240, 1.6% error
Gwinnett County:
941,300 ARC 2020
957,062 Census 2020
Result: ARC underestimate of 15,762, 1.6% error
Henry County:
246,800 ARC 2020
240,712 Census 2020
Result: ARC overestimate of 6,088, 2.5% error
Rockdale County:
96,800 ARC 2020
93,570 Census 2020
Result: ARC overestimate of 3,230, 3.5% error
City of Atlanta:
478,200 ARC 2020
498,715 Census 2020
Result ARC underestimate of 20,515, 4.1% error
It seems that the ARC's model historically underestimated populations in urban areas and in areas with high ratios of multifamily or rental housing, while slightly overestimating populations in suburban areas and in places with high ratios of single-family or owner-occupied housing. This could indicate that the ARC is actually underestimating the population in the City of Atlanta in 2023.
However, I personally doubt that's the case, as housing trends have changed quite a bit since 2020.
I too am skeptical of the ARC's population estimates, though not for the same reasons you are. I think it's more a case of imperfect methodology (which I find to rely too heavily on building permit data alone), and less a case of "boosterism" or "inflating expectations". If I must guess, the actual numbers lie somewhere between the 2022 Census Estimates and the ARC's 2023 numbers: maybe 510k in the City of Atlanta in 2023, and not 521k... still a record high and a high rate of growth, regardless. Yes, the Census Bureau will weigh in on the matter by next summer, and I suspect their number will be about 510k.
It appears that the ARC has not taken into account the almost-certain change in average household sizes since 2020 -- households have almost certainly become smaller since then for a variety of reasons (COVID, WFH, fewer kids, etc.). In fairness to the ARC... there isn't any good, recent, publicly-available data on household sizes. The Census Bureau found the City of Atlanta to have an average household size of 2.06 during the 2017-2021 period... that number has almost certainly fallen below 2.0 by now.
Yes, the City of Atlanta has permitted far more housing units than other Metro Area jurisdictions, but the overwhelming majority of those units (85%) have been multifamily units with a unit mix skewed heavily towards studios and 1-bedrooms versus 2-bedrooms and 3-bedrooms... an increased share of new multifamily units are occupied by singles instead of couples, roommates, or families. Additionally, multifamily occupancy rates have fallen just a little bit compared to last year, region-wide (from ~96% to ~93%).
Average household sizes for single-family homes have probably decreased a bit too for similar reasons, though likely to a smaller degree than for multifamily homes: smaller households are occupying larger single-family homes (due to home offices, etc.), and there are fewer children (due to lower birth rates), but there has also been an increase in multigenerational living in suburban single-family homes: young adult children are increasingly living at their parents' large suburban homes, and likewise, older/retired adults are a bit more likely to reside with their children/grandchildren than previously.
To be clear, these phenomena are not exclusively impacting the City of Atlanta... all jurisdictions are impacted. However, given the City of Atlanta's multifamily-heavy development, it is also the most likely to have an overshoot in the population estimates using the ARC's building-permit-based population change model. A slight majority of new Cobb County and DeKalb County housing units have been multifamily too (and surprisingly, Forsyth County has been at nearly 50% multifamily), so those places are likely to have high estimates too (though not to the degree of the City of Atlanta). Meanwhile, a place like Cherokee County -- where 83% of new housing units have been single-family -- is likely to have more accurate population estimates using the ARC's model.
This begs the following question: how accurate has the ARC's population estimate model been in the past?
City of Atlanta:
478,200 ARC 2020
498,715 Census 2020
Result ARC underestimate of 20,515,
This could indicate that the ARC is actually underestimating the population in the City of Atlanta in 2023.
"Underestimating" compared to the official 2020 US Census, yes. But the apples-to-apples comparison -- what ARC estimated for the City of Atlanta in 2020 versus its new numbers in 2023 -- stands out as blatant feel-good boostered stats. ARC said there were 478,200 Atlantans in 2020 but now estimates over 520,000 in 2023, a whopping gain of 42,000 in just 3 years. (A whopper of a tale might be more accurate.) The OP's link shows an ARC gain of 14,300 (+2.8%) in the City of Atlanta in just the last year, 2022-2023. That's pure ATL Fantasyland compared to US Census data, which show just over 1,500 new Atlanta residents (+0.1%) over TWO full years (2020 census and 2022 estimate). I'd take ARC's City of Atlanta numbers with a giant grain of salt.
Last edited by masonbauknight; 08-10-2023 at 05:55 PM..
Mason, as an outsider your not seeing many local trends.
ARC is more correct, than wrong on the trajectory. Whata happening on the ground level is the steel housing costs happening nationwide has been acutely bad in the Atlanta region.
The large change that is impacting the city is there has overnight been a large shift to people having more roommates or having roommates when they didn't before. This is because the extreme shift in housing costs over the past 2 years reverses the rate at which singles in their 30's and even 40's get rentals on their own. It also causes many in more expensive areas to stay with parents longer.
This prompts dwelling to have a higher average number of people per dwelling. Thus happens more in the city as there are more rentals and more singles
"Underestimating" compared to the official 2020 US Census, yes. But the apples-to-apples comparison -- what ARC estimated for the City of Atlanta in 2020 versus its new numbers in 2023 -- stands out as blatant feel-good boostered stats. ARC said there were 478,200 Atlantans in 2020 but now estimates over 520,000 in 2023, a whopping gain of 42,000 in just 3 years. (A whopper of a tale might be more accurate.) The OP's link shows an ARC gain of 14,300 (+2.8%) in the City of Atlanta in just the last year, 2022-2023. That's pure ATL Fantasyland compared to US Census data, which show just over 1,500 new Atlanta residents (+0.1%) over TWO full years (2020 census and 2022 estimate). I'd take ARC's City of Atlanta numbers with a giant grain of salt.
I spot a few misunderstandings here.
Quote:
Originally Posted by masonbauknight
what ARC estimated for the City of Atlanta in 2020 versus its new numbers in 2023 -- stands out as blatant feel-good boostered stats. ARC said there were 478,200 Atlantans in 2020 but now estimates over 520,000 in 2023, a whopping gain of 42,000 in just 3 years.
The ARC didn't increase their population estimate by an unreasonable, illogical amount (+42,000 in 3 years) out of thin air. After the results of the 2020 Census were released, the ARC began to use the new Census numbers as a baseline for future population modeling. This is standard practice in population modeling -- think of this as a "reset" of the model. This is addressed on slide 10 of the presentation located near the bottom of this webpage. From the "Methodology" section of this slide:
Quote:
The foundation for these population estimates is the 2020 United States Census.
Previous ARC estimates -- including their 2020 estimates, which were released before the 2020 Census results were published -- were developed using the 2010 Census as a baseline. So, the reason for the massive jump -- from a 2020 estimate of 478,200 to well over 500,000 a year or two later -- is because of the new 2020 Census baseline of 498,715, which was 20,515 people more than the ARC had previously estimated. Accordingly, the ARC adjusted their model upward by 20,515 to reflect the new Census baseline data. That leaves an estimated actual population increase of 22,600 since then (521,315 estimated in 2023 minus 498,715 counted by the Census in 2020). 22,600 more people over 3 years averages to 7,533 more people annually, which does not seem at all far-fetched or unrealistic (in fact, that's right in line with the average annual increase during the 2010s decade), though as I explained at length in my earlier response, I too find this number to be a bit high.
Quote:
Originally Posted by masonbauknight
That's pure ATL Fantasyland compared to US Census data, which show just over 1,500 new Atlanta residents (+0.1%) over TWO full years (2020 census and 2022 estimate).
Consider the following Census Bureau population figures for the City of Atlanta:
2020 Census: 498,715 2021 estimates: 496,461 (-2,254 from previous year) 2022 estimates: 499,127 (+2,666 from previous year)
These census numbers indicate a population drop between 2020 and 2021 -- reflective of COVID, school closures, remote working, low mortgage rates, crime increases, etc. -- and then a population increase between 2021 and 2022 -- reflective of an increase in housing deliveries, increased in-migration from more expensive regions, increased international immigration, school reopenings, and so on.
So to state that the population had essentially been stagnant between 2020 and 2022 -- while correct -- does not explain what's going on "under the hood". Given that the population actually increased over the prior year, and given the recent intensification of the trends propelling that population growth over the prior year -- one might expect for the population to more intensely increase in 2023.
For what it's worth, the ARC pegged the City of Atlanta's population at 507,015 in 2022, versus the Census estimate of 499,715 during the same year (2022), so the ARC's number was likely about 7,300 too high in 2022. Extrapolating that error from the ARC's 2022 estimate into 2023, one might expect for the 2023 ARC estimate of 521,315 to in reality reflect an actual population closer to 514,015.
Quote:
Originally Posted by masonbauknight
stands out as blatant feel-good boostered stats
A whopper of a tale might be more accurate
That's pure ATL Fantasyland
I, truly, do not see any evidence of boosterism, or evidence of some kind of pro-City of Atlanta conspiracy -- nor can I think of any real incentive for such behavior.
The previous two years of ARC estimates showed a higher rate of growth in suburban and exurban jurisdictions -- I don't think anybody was complaining of some anti-City of Atlanta agenda from the ARC last year or two years ago.
What I do see is a regional organization doing its best to help local governments make informed decisions with the most current data available -- but with a potentially flawed, potentially outdated methodology.
Mason, as an outsider your not seeing many local trends.
ARC is more correct, than wrong on the trajectory. Whata happening on the ground level is the steel housing costs happening nationwide has been acutely bad in the Atlanta region.
The large change that is impacting the city is there has overnight been a large shift to people having more roommates or having roommates when they didn't before. This is because the extreme shift in housing costs over the past 2 years reverses the rate at which singles in their 30's and even 40's get rentals on their own. It also causes many in more expensive areas to stay with parents longer.
This prompts dwelling to have a higher average number of people per dwelling. Thus happens more in the city as there are more rentals and more singles
I'm an old Georgia boy. That probably makes me more skeptical about ATL boosterism (it's always there), but those ARC numbers foretell at least 100,000 more people in the City of Atlanta within 10 years, 2020-2030. Not likely, I'd say. In fact, it's most unlikely. Look, we all want Atlanta to be successful and to prosper -- and its metro area will show solid growth in the coming decade. But all the roommate movements of Gen Z in the world won't make Atlanta pass 600,000, 625,000 by 2030. That's full-blown Atlanta-love bordering on self-promotion. I don't buy it -- or rather, I don't buy ARC's stats. Their City of Atlanta numbers are wishful thinking.
The ARC didn't increase their population estimate by an unreasonable, illogical amount (+42,000 in 3 years) out of thin air. After the results of the 2020 Census were released, the ARC began to use the new Census numbers as a baseline for future population modeling. This is standard practice in population modeling -- think of this as a "reset" of the model. This is addressed on slide 10 of the presentation located near the bottom of this webpage. From the "Methodology" section of this slide:
Previous ARC estimates -- including their 2020 estimates, which were released before the 2020 Census results were published -- were developed using the 2010 Census as a baseline. So, the reason for the massive jump -- from a 2020 estimate of 478,200 to well over 500,000 a year or two later -- is because of the new 2020 Census baseline of 498,715, which was 20,515 people more than the ARC had previously estimated. Accordingly, the ARC adjusted their model upward by 20,515 to reflect the new Census baseline data. That leaves an estimated actual population increase of 22,600 since then (521,315 estimated in 2023 minus 498,715 counted by the Census in 2020). 22,600 more people over 3 years averages to 7,533 more people annually, which does not seem at all far-fetched or unrealistic (in fact, that's right in line with the average annual increase during the 2010s decade), though as I explained at length in my earlier response, I too find this number to be a bit high.
Consider the following Census Bureau population figures for the City of Atlanta:
2020 Census: 498,715 2021 estimates: 496,461 (-2,254 from previous year) 2022 estimates: 499,127 (+2,666 from previous year)
These census numbers indicate a population drop between 2020 and 2021 -- reflective of COVID, school closures, remote working, low mortgage rates, crime increases, etc. -- and then a population increase between 2021 and 2022 -- reflective of an increase in housing deliveries, increased in-migration from more expensive regions, increased international immigration, school reopenings, and so on.
So to state that the population had essentially been stagnant between 2020 and 2022 -- while correct -- does not explain what's going on "under the hood". Given that the population actually increased over the prior year, and given the recent intensification of the trends propelling that population growth over the prior year -- one might expect for the population to more intensely increase in 2023.
For what it's worth, the ARC pegged the City of Atlanta's population at 507,015 in 2022, versus the Census estimate of 499,715 during the same year (2022), so the ARC's number was likely about 7,300 too high in 2022. Extrapolating that error from the ARC's 2022 estimate into 2023, one might expect for the 2023 ARC estimate of 521,315 to in reality reflect an actual population closer to 514,015.
I, truly, do not see any evidence of boosterism, or evidence of some kind of pro-City of Atlanta conspiracy -- nor can I think of any real incentive for such behavior.
The previous two years of ARC estimates showed a higher rate of growth in suburban and exurban jurisdictions -- I don't think anybody was complaining of some anti-City of Atlanta agenda from the ARC last year or two years ago.
What I do see is a regional organization doing its best to help local governments make informed decisions with the most current data available -- but with a potentially flawed, potentially outdated methodology.
Then ARC is shifting its statistical model, not just its stats, saying, "Man, did we ever botch the 2020 City of Atlanta numbers! Let's play catch-up with the US Census Bureau's 2020 figure of 498,000 and go off a statistical cliff." Then they cross the magic psychological barrier of 500,000, and add 14,300 people in just one year, 2022-2023. That's exactly what ARC did in my opinion. I'm all for putting numbers out there to "help metro ATL plan" -- that's great. But to take these City of Atlanta numbers seriously at all is stepping into old-time Atlanta boosterism.
Then ARC is shifting its statistical model, not just its stats, saying, "Man, did we ever botch the 2020 City of Atlanta numbers! Let's play catch-up with the US Census Bureau's 2020 figure of 498,000 and go off a statistical cliff." Then they cross the magic psychological barrier of 500,000, and add 14,300 people in just one year, 2022-2023. That's exactly what ARC did in my opinion. I'm all for putting numbers out there to "help metro ATL plan" -- that's great. But to take these City of Atlanta numbers seriously at all is stepping into old-time Atlanta boosterism.
The Census Bureau -- the gold standard for population modeling and estimation -- uses the exact same process on itself! On its own annual population estimates!
You can read more about the methodology for the Census Bureau's annual population estimates here, though I suspect you won't as you have thus far displayed a lack of reading comprehension, a flawed understanding of population modeling, and willful ignorance.
Every 10 years, the Census Bureau's American Community Survey (which is used to develop population estimates) "resets" their model to incorporate more accurate Decennial United States Census population data.
For both the Census Bureau and the ARC, it's not a case of "oh, our model sucked last time so let's fudge the numbers to make ourselves look more competent", but rather a case of "we acknowledge that baseline inputs become less relevant and less accurate with time, so let's update them regularly to ensure an accurate model". Do remember that the ARC overestimated the populations in a few suburban/exurban counties last time around, and accordingly, the ARC adjusted their model downward for those places.
masonbauknight, you are allowed to believe the numbers are wrong, obviously. I do too, to a small extent. But I encourage you to come up with a better explanation than "old-time Atlanta boosterism" and "the ARC's data guy made up the numbers".
The Census Bureau -- the gold standard for population modeling and estimation -- uses the exact same process on itself! On its own annual population estimates!
You can read more about the methodology for the Census Bureau's annual population estimates here, though I suspect you won't as you have thus far displayed a lack of reading comprehension, a flawed understanding of population modeling, and willful ignorance.
Every 10 years, the Census Bureau's American Community Survey (which is used to develop population estimates) "resets" their model to ...
masonbauknight, you are allowed to believe the numbers are wrong, obviously. I do too, to a small extent. But I encourage you to come up with a better explanation than "old-time Atlanta boosterism" and "the ARC's data guy made up the numbers".
With these City of Atlanta stats, we might be "resetting our model" and incorporating new urban conditions, but we're still going off the deep end with the numbers. Sorry, a population gain of +2.8% over just one year (ARC) versus +0.1 over two years (US Census) is a huge chasm in demographics and for an older city of 500K. One of the sources has to be way off. And for the more reliable source, I must go with the bad old US Census Bureau. They are more to be trusted. I'm just a messenger here -- giving ARC a reality check. And I think certain folks are giving far too much credit to the yokels.
Last edited by masonbauknight; 08-10-2023 at 09:22 PM..
Please register to post and access all features of our very popular forum. It is free and quick. Over $68,000 in prizes has already been given out to active posters on our forum. Additional giveaways are planned.
Detailed information about all U.S. cities, counties, and zip codes on our site: City-data.com.
Please register to participate in our discussions with 2 million other members - it's free and quick! Some forums can only be seen by registered members. After you create your account, you'll be able to customize options and access all our 15,000 new posts/day with fewer ads.
