The Official 2020 Census Thread

[cinyc]

94.1% total enumerated in today’s stats. The self-response rate remains at 66.0%.

Idaho remains tops at 99.8%; Alabama is last at 86.3%.

[jimrtex]

Interesting slides from the Census Scientific Advisory Committee Meeting yesterday:

https://www2.census.gov/cac/sac/meetings/2020-09/presentation-2020-census-update.pdf

https://www2.census.gov/cac/sac/meetings/2020-09/presentation-post-enumeration-methodology.pdf

https://www2.census.gov/cac/sac/meetings/2020-09/presentation-update-administrative-record-usage.pdf

https://www.census.gov/about/cac/sac/meetings/2020-09-meeting.html

The first was particularly interesting with some of the metrics.

The turnover in enumerators was quite surprising.

1026K Selections (I don't if this is applicants or what), since

676K Invited to training. You're "selected" but not invited to train? Most training could be done on line, but there was a requirement for at least one day of in-person training. Perhaps the Selections were permitted to do the online training, and either dropped out or were dropped before being invited to the in-person training.

426K Hired. Perhaps these were those who showed up for in-person training. I don't know where training occurred, since some ACO are statewide. Travelling to Billings, training for X hours and then returning home may require an overnight stay. But are there enough enumerators to provide satellite training in Great Falls, Missoula, Butte, Havre, etc.?

224K Active. This is pretty odd, that half of those hired have stopped working. Perhaps some were hired in the spring and then didn't return in August. There may have been layoffs in places like Idaho and West Virginia. But it sounds like some started working for a few days and quit. Maybe some are eased out based on under-performance, mistakes, etc. 
I imagine the enumerator in South Carolina who was shot at may have had second thoughts. But this is consistent with recent temporary hires, which has been steady around 270K-280K for several weeks now. The first level supervisors are temps, working from home, and there must be bunches of other temps processing paper forms, working phones, etc.

192K enumerators needed to complete enumeration by September 30. Of course bunches of the 225K are in the wrong places.

69K in training. Quite a bit for the last two weeks, but clearly needed in Alabama, Georgia, and other areas in the Southeast.

Current List of areas needing enumerators

Baldwin, AL and Bay and Walton, FL are on the list ("Applicants must have boat or high clearance vehicle. Sand buggies, and 4-wheel drive a plus."

Holmes, OH ("Applicants must have buggy. German fluency a plus")

Riley and Douglas, KS ("COVID-19 immunity a plus. Opportunity to make extra money since you can fast forward over most of the online lectures. Must be able to text, use Tik Tok and Instagram. Facebook not necessary")

There may be others college-related as well: Albany, WY; Monongalia, WV; and Grand Forks, ND.

Smaller counties may have few enumerators, and more vulnerable to statistical variation.

220K for 330M = 220 for 330K = 2 for 3K. In a  county with 3K, if those 2 aren't willing or able to work 19 hours, or one quits or is ill or injured, or their car is broken, you could be 50% short.

In a county with 330K and 220 enumerators, you're likely to have a smaller share be out of commission, and others may be willing to work a few more hours per week, since the Census Bureau is paying a bonus.

A neighbor had a census pouch and her infant daughter. I asked if her daughter was going to help. She seemed to think it an odd question, and said that she was taking the baby to the babysitter, before she started work. The Census videos show diverse enumerators, I thought it quite normal that there would be mothers with children enumerating.

[Tender Branson]

Trump: "Constitution doesn't require census to be accurate."

https://edition.cnn.com/2020/09/18/politics/census-lawsuit-accuracy/index.html

LOL. What's the purpose of holding one every 10 years then ?

[Tender Branson]

There are massive IT problems in Puerto Rico:

https://eu.usatoday.com/story/news/2020/09/19/puerto-rico-residents-2020-census-undercount/3481896001/

PR now has among the highest enumeration rates, but the IT systems are not saving them as "completed" and assign already-enumerated households for re-interviewing ...

[jimrtex]

ACO's that need to average 2.0% or more daily to reach 100% of the NRFU workload by September 30. Report day September 18, data day September 17. 74.0% or less.

This assumes the data reported on September 18 is as of 0:00:00 (midnight), so that if the projected completion time was 1.75 days , it would be completed at around 6:00 pm on the 19th.

53.0 (1.48% six-day average, projected completion 10/19) Shreveport, LA (western and northern Louisiana)
57.4 (1.53%, 10/15) Window Rock, AZ (Navajo reservation)
58.4 (2.37%, 10/5) Columbus, GA (southern Georgia)
63.4 (1.75%, 10/8) Birmingham, AL (central Alabama)
65.0 (2.52%, 10/1) Jackson, MS (western and northern Mississippi)
67.0 (1.73%, 10/7) Gwinnett County, GA (Gwinnett and points east)
67.9 (1.32%, 10/12) Pensacola, FL (Redneck Riviera)
68.4 (1.28%, 10/12) Mobile, AL (southern Alabama)
69.6 (2.03%, 10/2) Philadelphia-Penn, PA (north Philadelphia)
69.9 (1.77%, 10/5) Colorado Springs, CO (southern Colorado)
70.2 (1.67%, 10/5) Charleston, SC (Low Country, South Carolina)
70.8 (1.43%, 10/8) Douglasville, GA (western Atlanta suburbs and points west.
70.9 (1.68%, 10/5) Durham, NC (north central, includes Greensboro, but not Raleigh or Winston-Salem)
71.0 (3.00%, 9/27) Macon, GA (central Georgia, south-central from Atlanta perspective)
71.1 (2.03%, 10/2) Maricopa West, AZ (Glendale and westward)
71.9 (1.57%, 10/5) Huntsville, AL (northern Alabama)
72.2 (1.68%, 10/4) Greenville, NC (eastern North Carolina)
72.6 (1.58%, 10/5) Billings, MT (Montana)
73.0 (1.23%, 10/9) Columbia, SC (Midlands)
73.1 (1.98%, 10/1) Baton Rouge, LA (Baton Rouge and areas to east and west)
73.2 (1.90%, 10/2) Atlanta, GA (Fulton County)
73.4 (2.17%, 9/30) Brooklyn 1, NY (eastern Brooklyn)
73.5 (1.68%, 10/3) Tampa, FL (Hillsborough County)
74.0 (2.00%, 10/1) Fort Bend County, TX (southern Houston suburbs, and down to Victoria)
74.0 (2.65%, 9/27) Egg Harbor, NJ (far southern New Jersey)

Columbia, SC has fallen into the list, and is actually the poorest recent performer, even worse than Mobile and Pensacola. Fort Bend County and Egg Harbor have nearly escaped.

2.0% per day is clearly doable (see Macon, GA which has averaged 3.0% for 6 days) assuming that there is not a slowdown as an ACO gets closer to 100%.

I think that if they had an extra 15 days they could do a proper cleanup in most areas, and be close in even the slowest areas.

[jimrtex]

Trump: "Constitution doesn't require census to be accurate."

https://edition.cnn.com/2020/09/18/politics/census-lawsuit-accuracy/index.html

LOL. What's the purpose of holding one every 10 years then ?

I think the argument the lawyer was making was that it was not judiciable, as there is no manageable standard,

Congress could provide additional time. The court should order Pelosi to bring the extension bill to the floor.

[jimrtex]

Twenty-nine-day averages (8/17/2010 to 9/15/2010)

Total 2.06 (0.071% per day)
Internet 1.66% (0.057% per day)
Paper+Phone 0.38% (0.013% per day)

At this rate, the USA would reach its 2010 response rate of 66.5% on September 23. The current estimate is that it will take 8.44 days, and there are 16 days available.

Two states reached their 2010 response rate in June: MI(15) and WA(30)
Two states reached their 2010 response rate in July:  NV(20) and KY(27)
Eight states reached their 2010 response rate in August: ID(1), CO(17), OR(21), UT(22), VA(24), NH(25), AZ (29) and OH (31).
Nine states have reached their 2010 response rate in September: MD(3), MN(3), NJ(7), CT(8), ME(8), IN(11), CA(14), NE(14), AL(15)

Alabama reached its 2010 response rate on September 15, becoming the 21st state, and 1st southern state to do so. It appears that there might have been an initial campaign to keep the 7th congressional seat, which led to relatively good performance compared to past censuses.

Illinois is projected to reach its 2010 response rate on September 18 (gap is 0.2%).
Delaware is projected to reach its 2010 response rate on September 19 (gap is 0.4%)
Florida is projected to reach its 2010 response rate on September 20 (gap is 0.4%)
Massachusetts is projected to reach its 2010 response rate on September 21 (gap is 0.5%)
Vermont is projected to reach its 2010 response rate on September 21 (gap is 0.5%)
South Dakota is projected to reach its 2020 response rate on September 25.

No states are projected to reach their 2010 response rate in the remainder of September, resulting in 27 states (53%) reaching their 2010 response rate during the official enumeration period.

The following five states would reach within 28 days, but not by the end of September.

Georgia is projected to reach its 2010 response rate on October 2.
Rhode Island is projected to reach its 2010 response rate on October 2.
New York is projected to reach its 2010 response rate on October 4.
Kansas is projected to reach its 2010 response rate on October 5.
Alaska is projected to reach its 2010 response rate on October 10.

12 states in remainder of October.
3 states and DC in November
2 states in December
1 state in 2021, West Virginia on January 25

Green-7: Reached (20)
in July (2): MI, WA;
in August (10): NV, KY, ID, CO, OR, UT, VA, NH (25), AZ, OH;
in September (9): MD(3), MN, NJ(7), CT(8), ME, IN, CA, NE, AL
Green-6: Next 14 days (6): IL, DE(19), MA(21), VT(21), FL, SD
Green-5: By September 30 (0):
Red-3: By October 13, 28 days (5): GA, RI(2), KS, AK, NY
Red-4: October (12): PA, TN, NM, TX, HI, WI, OK, AR, WY, LA, MS, NC
Red-5: November (4): DC, MO, IA, SC
Red-6: December (2): ND, MT
Red-7: 2021 (1): WV

[jimrtex]

Thirty-day averages (8/17/2010 to 9/16/2010)

Total 2.14 (0.071% per day)
Internet 1.69% (0.056% per day)
Paper+Phone 0.44% (0.015% per day)

At this rate, the USA would reach its 2010 response rate of 66.5% on September 24. The current estimate is that it will take 8.42 days, and there are 15 days available.

Two states reached their 2010 response rate in June: MI(15) and WA(30)
Two states reached their 2010 response rate in July:  NV(20) and KY(27)
Eight states reached their 2010 response rate in August: ID(1), CO(17), OR(21), UT(22), VA(24), NH(25), AZ (29) and OH (31).
Nine states have reached their 2010 response rate in September: MD(3), MN(3), NJ(7), CT(8), ME(8), IN(11), CA(14), NE(14), AL(15)

Illinois is projected to reach its 2010 response rate on September 17 (gap is 0.1%).
Delaware is projected to reach its 2010 response rate on September 19 (gap is 0.3%)
Florida is projected to reach its 2010 response rate on September 20 (gap is 0.3%)
Massachusetts is projected to reach its 2010 response rate on September 21 (gap is 0.4%)
Vermont is projected to reach its 2010 response rate on September 22 (gap is 0.5%)
South Dakota is projected to reach its 2020 response rate on September 26.

No states are projected to reach their 2010 response rate in the remainder of September, resulting in 27 states (53%) reaching their 2010 response rate during the official enumeration period.

The following seven states would reach within 28 days, but not by the end of September.

Georgia is projected to reach its 2010 response rate on October 1.
Rhode Island is projected to reach its 2010 response rate on October 2.
Kansas is projected to reach its 2010 response rate on October 3.
New York is projected to reach its 2010 response rate on October 4.
Alaska is projected to reach its 2010 response rate on October 10.
Pennsylvania is projected to reach its 2010 response rate on October 12.
Tennessee is projected to reach its 2010 response rate on October 13.

10 states and DC in remainder of October.
3 states in November
2 states in December
1 state in 2021, West Virginia on January 31

Green-7: Reached (21)
in July (2): MI, WA;
in August (10): NV, KY, ID, CO, OR, UT, VA, NH (25), AZ, OH;
in September (9): MD(3), MN, NJ(7), CT(8), ME, IN, CA, NE, AL
Green-6: Next 14 days (6): IL, DE(19), FL, MA(21), VT(22), SD
Green-5: By September 30 (0):
Red-3: By October 13, 28 days (7): GA, RI(2), KS, NY, AK, PA, TN
Red-4: October (11): NM, TX, HI, WI, OK, AR, WY, LA, MS, NC, DC
Red-5: November (3): IA, MO, SC
Red-6: December (2): MT, ND
Red-7: 2021 (1): WV

[cinyc]

94.6% total enumeration rate as of today’s report. The self-response rate ticked up to 66.1%, 0.4 points away from the final 2010 rate.

[jimrtex]

94.6% total enumeration rate as of today’s report. The self-response rate ticked up to 66.1%, 0.4 points away from the final 2010 rate.

It is interesting that they show the self-response rate on the total response page, but take the weekend off on the self-response page itself.

[jimrtex]

ACO's that need to average 2.0% or more daily to reach 100% of the NRFU workload by September 30. Report day September 19, data day September 18. 76.0% or less.

Average rate is for seven days. We'll start to use a rolling average tomorrow.

55.1 (1.57%, 10/18) Shreveport, LA (western and northern Louisiana)
59.1 (1.56%, 10/16) Window Rock, AZ (Navajo reservation)
60.7 (2.36%, 10/6) Columbus, GA (southern Georgia)
65.4 (1.79%, 10/9) Birmingham, AL (central Alabama)
67.5 (2.51%, 10/2) Jackson, MS (western and northern Mississippi)
68.9 (1.76%, 10/7) Gwinnett County, GA (Gwinnett and points east)
69.8 (1.30%, 10/13) Mobile, AL (southern Alabama)
69.8 (1.40%, 10/11) Pensacola, FL (Redneck Riviera)
71.3 (1.99%, 10/4) Philadelphia-Penn, PA (north Philadelphia)
71.6 (1.76%, 10/6) Colorado Springs, CO (southern Colorado)
71.8 (1.37%, 10/10) Douglasville, GA (western Atlanta suburbs and points west.
71.8 (1.66%, 10/6) Charleston, SC (Low Country, South Carolina)
72.4 (1.66%, 10/6) Durham, NC (north central, includes Greensboro, but not Raleigh or Winston-Salem)
73.1 (2.03%, 10/3) Maricopa West, AZ (Glendale and westward)
73.5 (1.57%, 10/6) Huntsville, AL (northern Alabama)
73.8 (1.67%, 10/5) Greenville, NC (eastern North Carolina)
73.9 (2.99%, 9/28) Macon, GA (central Georgia, south-central from Atlanta perspective)
74.2 (1.59%, 10/6) Billings, MT (Montana)
74.4 (1.26%, 10/10) Columbia, SC (Midlands)
75.0 (1.97%, 10/2) Baton Rouge, LA (Baton Rouge and areas to east and west)
75.1 (1.90%, 10/3) Atlanta, GA (Fulton County)
75.2 (1.69%, 10/4) Tampa, FL (Hillsborough County)
75.2 (1.80%, 10/3) Lakeland, FL (south central peninsula)
75.4 (2.14%, 10/1) Brooklyn 1, NY (eastern Brooklyn)
75.6 (1.66%, 10/4) Philadelphia-Franklin, PA (south and west Philadelphia)
75.9 (1.54%, 10/5) Tulsa, OK (northeast Oklahoma)
75.9 (1.99%, 10/2) Fort Bend County, TX (southern Houston suburbs, and down to Victoria)

Three new ACO make the list. Polk and Osceola are in the northern part of Lakeland, FL ACO with generally smaller counties to the south, though it does include Indian River and St. Lucie on the Atlantic Coast. Both Philadelphia ACO are now on the list. Fort Bend is going to be inundated all next week as Hurricane Roberta Frances stalls along the coast, so it could slide further into the list.

It will be interesting whether Saturday is peak NRFU time or not, as those with 5x8 jobs are home, and little to do.

The 97 ACO not in the 85% band will need to average 1.25%/day to avoid the NRFU Curfew.

NRFU completion

[Tender Branson]

At this point it's pretty clear that the 30 or so ACOs at risk of not finishing work by Sept. 30th will need more time until Oct. 31st.

Congress or a Court must act now !

The CB can still start quality checks and tabulations for the rest of the country in October, while finishing the count elsewhere.

Census Bureau to complete final stretch of 2020 count despite ‘challenging’ conditions

“The Census Bureau’s leadership will continue to adapt, and our operations and decision-making will be flexible and dynamic, as much of this work in ongoing. As circumstances change, so will our plans as we adapt,” Dillingham said.

Ten states have less than 90% of their enumeration complete, so the bureau has sent 5,000 of its top-performing enumerators from states where the work is almost complete to areas that need more help.

https://federalnewsnetwork.com/agency-oversight/2020/09/census-bureau-to-complete-final-stretch-of-2020-count-despite-challenging-conditions

Even if they sent 5.000 additional census-takers to AL and elsewhere, it's not showing up in the daily numbers so far ...

They probably need to send 50.000 to those lagging states.

[Tender Branson]

95.0% (+0.4)

5 states remain below 90%:

AL, MS, LA, SC and MT.

[jimrtex]

ACO's that need to average 2.0% or more daily to reach 100% of the NRFU workload by September 30. Report day September 20, data day September 19. 78.0% or less.

Average rate is for seven days (rolling average).

57.3 (1.69% average, 10/17) Shreveport, LA (western and northern Louisiana)
60.6 (1.61%, 10/16) Window Rock, AZ (Navajo reservation)
63.4 (2.36%, 10/7) Columbus, GA (southern Georgia)
67.6 (1.86%, 10/9) Birmingham, AL (central Alabama)
70.0 (2.51%, 10/3) Jackson, MS (western and northern Mississippi)
71.0 (1.30%, 10/14) Pensacola, FL (Redneck Riviera)
71.0 (1.83%, 10/8) Gwinnett County, GA (Gwinnett and points east)
71.2 (1.23%, 10/15) Mobile, AL (southern Alabama)
73.1 (1.74%, 10/7) Colorado Springs, CO (southern Colorado)
73.3 (1.94%, 10/5) Philadelphia-Penn, PA (north Philadelphia)
73.4 (1.37%, 10/11) Douglasville, GA (western Atlanta suburbs and points west.
73.4 (1.67%, 10/7) Charleston, SC (Low Country, South Carolina)
73.7 (1.59%, 10/8) Durham, NC (north central, includes Greensboro, but not Raleigh or Winston-Salem)
74.9 (2.06%, 10/4) Maricopa West, AZ (Glendale and westward)
75.1 (1.56%, 10/8) Huntsville, AL (northern Alabama)
75.2 (1.60%, 10/7) Greenville, NC (eastern North Carolina)
75.6 (1.57%, 10/7) Billings, MT (Montana)
75.8 (1.27%, 10/11) Columbia, SC (Midlands)
76.6 (1.87%, 10/4) Atlanta, GA (Fulton County)
76.7 (1.69%, 10/5) Tampa, FL (Hillsborough County)
76.8 (1.97%, 10/3) Baton Rouge, LA (Baton Rouge and areas to east and west)
76.9 (2.94%, 9/29) Macon, GA (central Georgia, south-central from Atlanta perspective)
77.0 (1.61%, 10/6) Philadelphia-Franklin, PA (south and west Philadelphia)
77.1 (1.80%, 10/4) Lakeland, FL (south central peninsula)
77.1 (2.07%, 10/2) Brooklyn 1, NY (eastern Brooklyn)
77.2 (1.51%, 10/6) Tulsa, OK (northeast Oklahoma)
77.6 (1.94%, 10/3) Fort Bend County, TX (southern Houston suburbs, and down to Victoria)

Half full version. Eight days ago (September 12, the average for these ACO was 58.6%; it is now 72.9%).

The 85 ACO not in the 85% band will need to average 1.36%/day to avoid the NRFU Curfew.

NRFU completion

[jimrtex]

At this point it's pretty clear that the 30 or so ACOs at risk of not finishing work by Sept. 30th will need more time until Oct. 31st.

Congress or a Court must act now !

The CB can still start quality checks and tabulations for the rest of the country in October, while finishing the count elsewhere.

Census Bureau to complete final stretch of 2020 count despite ‘challenging’ conditions

“The Census Bureau’s leadership will continue to adapt, and our operations and decision-making will be flexible and dynamic, as much of this work in ongoing. As circumstances change, so will our plans as we adapt,” Dillingham said.

Ten states have less than 90% of their enumeration complete, so the bureau has sent 5,000 of its top-performing enumerators from states where the work is almost complete to areas that need more help.

https://federalnewsnetwork.com/agency-oversight/2020/09/census-bureau-to-complete-final-stretch-of-2020-count-despite-challenging-conditions

Even if they sent 5.000 additional census-takers to AL and elsewhere, it's not showing up in the daily numbers so far ...

They probably need to send 50.000 to those lagging states.

There are a few high performers such as Macon, GA; Columbus, GA; and Jackson, MS.

Perhaps they sent the shock troops to those areas.

A problem with outsiders is that they may speak with an accent, or speak too fast.

The Texas legislature uses students at the LBJ School of Public Affairs at UT-Austin as interns. One time an intern was briefing a committee on a bill. Perhaps nervous, she spoke fast. At the end, one of the senators, commented on the speed, and that he didn't comprehend the comments, and then suggested that the intern might be from New York City.

With shock that anyone would suggest such a dastardly thing, she exclaimed, "Ahm a Taix'n!!!!

[Tender Branson]

Sept. 20 total enumeration rates by state (rounded):

US: 95.4%
PR: 98.0%

Sept. 10 total enumeration rates by state (rounded):

US: 90%
PR: 90%

August end-of-month total enumeration rates by state (rounded):

US: 83%
PR: 76%

I will update this map 3x in the final month: on Sept. 11, on Sept. 21 and on Oct. 1

[Tender Branson]

If they are deploying additional census takers on the ground in the lagging states in the final 10 days, most states should finish out nicely at around 98-100%.

Only AL (and to a lesser extent LA and SC) continue to have a significant problem ...

[jimrtex]

Thirty-one-day averages (8/17/2010 to 9/17/2010)

Total 2.19 (0.071% per day)
Internet 1.73% (0.056% per day)
Paper+Phone 0.46% (0.015% per day)

At this rate, the USA would reach its 2010 response rate of 66.5% on September 24. The current estimate is that it will take 7.09 days, and there are 14 days available.

Two states reached their 2010 response rate in June: MI(15) and WA(30)
Two states reached their 2010 response rate in July:  NV(20) and KY(27)
Eight states reached their 2010 response rate in August: ID(1), CO(17), OR(21), UT(22), VA(24), NH(25), AZ (29) and OH (31).
Nine states have reached their 2010 response rate in September: MD(3), MN(3), NJ(7), CT(8), ME(8), IN(11), CA(14), NE(14), AL(15)

Illinois is projected to reach its 2010 response rate on September 18 (gap is 0.1%).
Delaware is projected to reach its 2010 response rate on September 19 (gap is 0.2%)
Florida is projected to reach its 2010 response rate on September 19 (gap is 0.3%)
Massachusetts is projected to reach its 2010 response rate on September 20 (gap is 0.3%)
Vermont is projected to reach its 2010 response rate on September 21 (gap is 0.4%)
South Dakota is projected to reach its 2020 response rate on September 25 (gap is 0.5%).

No states are projected to reach their 2010 response rate in the remainder of September, resulting in 27 states (53%) reaching their 2010 response rate during the official enumeration period.

The following seven states would reach within 28 days, but not by the end of September.

Georgia is projected to reach its 2010 response rate on October 1.
Rhode Island is projected to reach its 2010 response rate on October 2.
Kansas is projected to reach its 2010 response rate on October 4.
New York is projected to reach its 2010 response rate on October 4.
Alaska is projected to reach its 2010 response rate on October 11.
Tennessee is projected to reach its 2010 response rate on October 12.
Pennsylvania is projected to reach its 2010 response rate on October 14.

10 states in remainder of October.
3 states + DC in November
2 states in December
1 state in 2021, West Virginia on February 5

Green-7: Reached (21)
in July (2): MI, WA;
in August (10): NV, KY, ID, CO, OR, UT, VA, NH (25), AZ, OH;
in September (9): MD(3), MN, NJ(7), CT(8), ME, IN, CA, NE, AL
Green-6: By September 30 (6): IL, DE(19), FL, MA(20), VT(21), SD
Red-3: By October 14, 28 days (7): GA, RI(2), NY, KS, AK, PA, TN
Red-4: October (10): NM, TX, HI, NC,WI, OK, AR, WY, LA, IA
Red-5: November (3): MO, MS, DC, SC
Red-6: December (2): MT, ND
Red-7: 2021 (1): WV

[jimrtex]

If they are deploying additional census takers on the ground in the lagging states in the final 10 days, most states should finish out nicely at around 98-100%.

Only AL (and to a lesser extent LA and SC) continue to have a significant problem ...

I think it will end up like the case where a blizzard closes the school, and you have a few extra days to complete the assignment. A few more days in October will do wonders.

Sunday was a slow day, at 0.5% or less in ACO around 90%.

In the low rated ACO, it was more like a normal day, with completions near 2%. I suspect they are paying bounties to work more hours, perhaps up to 60 per week.

[jimrtex]

Thirty-two-day averages (8/17/2010 to 9/18/2010)

Total 2.23 (0.070% per day)
Internet 1.75% (0.055% per day)
Paper+Phone 0.48% (0.015% per day)

At this rate, the USA would reach its 2010 response rate of 66.5% on September 25. The current estimate is that it will take 7.16 days, and there are 13 days available.

Two states reached their 2010 response rate in June: MI(15) and WA(30)
Two states reached their 2010 response rate in July:  NV(20) and KY(27)
Eight states reached their 2010 response rate in August: ID(1), CO(17), OR(21), UT(22), VA(24), NH(25), AZ (29) and OH (31).
Ten states have reached their 2010 response rate in September: MD(3), MN(3), NJ(7), CT(8), ME(8), IN(11), CA(14), NE(14), AL(15), IL(18)

Illinois reached its 2010 response rate on September 18, becoming the 22nd state to do so.

Delaware is projected to reach its 2010 response rate on September 20 (gap is 0.2%)
Florida is projected to reach its 2010 response rate on September 20 (gap is 0.2%)
Massachusetts is projected to reach its 2010 response rate on September 20 (gap is 0.3%)
Vermont is projected to reach its 2010 response rate on September 23 (gap is 0.4%)
South Dakota is projected to reach its 2020 response rate on September 26 (gap is 0.5%).

No states are projected to reach their 2010 response rate in the remainder of September, resulting in 27 states (53%) reaching their 2010 response rate during the official enumeration period.

The following seven states would reach within 28 days.

Rhode Island is projected to reach its 2010 response rate on October 2.
Georgia is projected to reach its 2010 response rate on October 2.
Kansas is projected to reach its 2010 response rate on October 6.
New York is projected to reach its 2010 response rate on October 6.
Alaska is projected to reach its 2010 response rate on October 11.
Pennsylvania is projected to reach its 2010 response rate on October 13.
Tennessee is projected to reach its 2010 response rate on October 14.

10 states in remainder of October.
3 states + DC in November
2 states in December
1 state in 2021, West Virginia on February 11

Green-7: Reached (22)
in July (2): MI, WA;
in August (10): NV, KY, ID, CO, OR, UT, VA, NH (25), AZ, OH;
in September (10): MD(3), MN, NJ(7), CT(8), ME, IN, CA, NE, AL, IL
Green-6: By September 30 (5): DE(20), FL, MA(22), VT(23), SD
Red-3: By October 15, 28 days (7): RI(2), GA, NY, KS, AK, PA, TN
Red-4: October (10): NM, HI, TX, NC,WI, OK, AR, WY, LA, MS
Red-5: November (3): DC, IA, MO, SC
Red-6: December (2): MT, ND
Red-7: 2021 (1): WV

[jimrtex]

ACO's that need to average 2.0% or more daily to reach 100% of the NRFU workload by September 30. Report day September 21, data day September 20. 80.0% or less.

Average rate is for seven days (rolling average).

59.3 (1.79%, 10/13) Shreveport, LA (western and northern Louisiana)
62.1 (1.66%, 10/13) Window Rock, AZ (Navajo reservation)
65.4 (2.33%, 10/5) Columbus, GA (southern Georgia)
69.2 (1.90%, 10/7) Birmingham, AL (central Alabama)
71.8 (2.46%, 10/2) Jackson, MS (western and northern Mississippi)
72.5 (1.17%, 10/14) Mobile, AL (southern Alabama)
72.5 (1.29%, 10/12) Pensacola, FL (Redneck Riviera)
72.8 (1.87%, 10/5) Gwinnett County, GA (Gwinnett and points east)
74.4 (1.67%, 10/6) Colorado Springs, CO (southern Colorado)
74.6 (1.63%, 10/6) Charleston, SC (Low Country, South Carolina)
74.8 (1.54%, 10/7) Durham, NC (north central, includes Greensboro, but not Raleigh or Winston-Salem)
75.1 (1.40%, 10/8) Douglasville, GA (western Atlanta suburbs and points west.
75.3 (1.94%, 10/3) Philadelphia-Penn, PA (north Philadelphia)
76.2 (1.50%, 10/6) Huntsville, AL (northern Alabama)
76.6 (1.61%, 10/5) Greenville, NC (eastern North Carolina)
76.6 (2.06%, 10/2) Maricopa West, AZ (Glendale and westward)
76.8 (1.26%, 10/9) Columbia, SC (Midlands)
76.9 (1.54%, 10/5) Billings, MT (Montana)
77.9 (1.86%, 10/2) Atlanta, GA (Fulton County)
78.0 (1.69%, 10/4) Tampa, FL (Hillsborough County)
78.2 (1.71%, 10/3) Lakeland, FL (south central peninsula)
78.4 (1.47%, 10/5) Tulsa, OK (northeast Oklahoma)
78.4 (1.60%, 10/4) Philadelphia-Franklin, PA (south and west Philadelphia)
78.4 (1.94%, 10/2) Baton Rouge, LA (Baton Rouge and areas to east and west)
78.6 (1.81%, 10/2) Fort Bend County, TX (southern Houston suburbs, and down to Victoria)
78.8 (2.01%, 10/1) Brooklyn 1, NY (eastern Brooklyn)
79.0 (1.59%, 10/4) Seminole County, FL (Seminole, Volusia, Brevard counties)
79.2 (2.87%, 9/28) Macon, GA (central Georgia, south-central from Atlanta perspective)
79.6 (1.64%, 10/3) Jacksonville, FL (Duval, Nassau, Baker counties)
79.6 (1.69%, 10/3) Winston-Salem, NC (northwestern North Carolina)
79.8 (1.71%, 10/2) Baltimore, MD (Baltimore City)

I messed up yesterday's projected completion dates, so the apparent improvement is not real. 4 ACO's have slid into the behind zone. They are not quite the backwaters of some of the other areas. There appears to be a surge effort in some ACO (Macon, GA; Columbus, GA; Jackson, MS). My guess is that these are areas that have received extra workers.

Shreveport, LA had a good Sunday, so it may also have received surge workers, or perhaps it has recovered from Laura. Mobile and Pensacola had the worst weeks following Sally. Fort Bend County which includes Fort Bend County which includes Fort Bend, Brazoria, and Galveston and down to Victoria is being deluged by Roberta Frances, though the heaviest impact appears to be in Houston so far (up to 13 inches) and expected to continue for a day or two more.

Half full version. Nine days ago (September 12, the average for these ACO was 59.4%; it is now 75.1%).

[jimrtex]

This is a secondary list of ACO that need to average 1.5% per day through September 30 to reach 98.0% completion. (80.0% to 82.9% through September 20). Setting a lower target recognizes that it will not be possible to reach 100% NRFU completion. 98% NRFU completion would result in 99% total completion.

Imputation of the remaining 1% will not result in wholesale errors, though it may be decisive in the apportionment of the final representatives. During imputation, a nearby housing unit is substituted for the housing unit with an unknown number of residents. Over 1000s of housing units the effect of random luck will average out.

Let's take an ultra-simplistic model where 6000 housing units in an ACO need to be imputed (there are around 600,000 housing units in an ACO, 6,000 = 1% of that), and every household has either 2 or 3 persons (average 2.5) for a total of about 15,000. There is a 99% probability that the imputed value will between 14,900 and 15,100.

If we assume that the other 99% of the housing units are accurately enumerated, then imputation would introduce an error of about 100, in 1.347M persons, or 0.007%.

Our model was simple, since household size has larger variance, but 90% are 4 or fewer, and 2 and 1 are most common. So maybe we triple our error (90% of households are +/- 1.5 of 2.5 persons).

There may be systemic biases. Perhaps the unknown households are largely single persons who are always on the road as salesmen. Or filled with 10 undocumented aliens who resist being counted. But we are still talking about error less than 0.05%.

80.2 (1.60%, 10/3) Cleveland, OH (Cuyahoga, excluding outer suburbs)
80.6 (1.44%, 10/4) Fayetteville, NC (southeastern North Carolina)
80.6 (1.86%, 10/1) West Palm Beach, FL (Palm Beach and Martin counties)
80.7 (1.54%, 10/3) Albuquerque, NM (northern New Mexico, except Navajo Reservation)
81.0 (2.09%, 9/30) Akron, OH (northeastern Ohio)
81.4 (1.35%, 10/4) Fredericksburg, VA (northern Virginia, excluding Loudoun, Fairfax, Arlington counties, Alexandria)
81.5 (1.56%, 10/2) Lexington, KY (eastern Kentucky)
81.5 (2.23%, 9/29) Tucson, AZ (southern Arizona)
81.5 (2.60%, 9/28) Egg Harbor, NJ (far southern New Jersey)
81.6 (2.31%, 9/28) Maricopa Central, AZ (Glendale and points west)
81.7 (1.96%, 9/30) DeKalb County, GA (DeKalb, Clayton counties)
81.8 (1.33%, 10/4) Memphis, TN (far western Tennessee)
81.8 (1.99%, 9/30) Des Moines, IA (Iowa)
81.9 (1.29%, 10/5) Louisville, KY (western Kentucky)
81.9 (1.79%, 10/1) Fort Myers, FL (Manatee, Sarasota, Charlotte, Lee)
82.2 (1.59%, 10/2) Ocala, FL (north central Peninsula)
82.3 (1.89%, 9/30) Detroit, MI (Wayne County)
82.5 (1.44%, 10/3) Gainesville, GA (nothern Georgia)
82.5 (2.03%, 9/29) Laredo, TX (south Texas, except Lower Rio Grande Valley)
82.7 (1.55%, 10/2) Lansing, MI (southern Michigan)
82.9 (1.23%, 10/4) Chico, CA (Inland Northern California)

[cinyc]

95.8% Total Enumeration and 66.2% Self-Response in today’s report. The total response increase has slowed down to about 0.4 points per day.

[Tender Branson]

The result of Austria's annual register-based mini-census has been released today.

The quality criteria for the annual mini-census are stricter than for the quarterly/annual population estimates, but not as strict as for the 10-year census (the next one is on Oct. 31, 2021).

The population on Oct. 31, 2019 was 8.894.380, up by 42.963 from a year ago (+0.5%).

Austria would just fall short of 9 million at the 2021 census, if current growth rates continue (8.99 million), but the increase would be almost 600.000 since the 2011 census (8.402 million).

Or up by 7%, which would be the highest growth rate since 1900-1910, when it was 11%.

[Tender Branson]

96.2% counted US-wide (+0.4)

The bottom 10 states are not going to make it until September 30th ...

They are averaging just 0.7% per day and need 1.0 to 1.2 per day now.