The Official 2020 Census Thread

[Brittain33]

Can anyone figure out which post broke the thread on the last page and report it?

Jim’s Google link I guess is too big.

Thank you!

[jimrtex]

Can anyone figure out which post broke the thread on the last page and report it?

Jim’s Google link I guess is too big.

Which link?

[jimrtex]

Thanks for your informative reply as always, Jimrtex.

In my letter to James Whitehorne and others in the Redistricting and Voting Rights Office, I had mentioned that the Voting Districts were not really needed for the redistricting itself (and Maine's liaison had mentioned that in her e-mail to me on February 21, 2020 when she told me that it didn't look like her office would be able to add county subdivisions as voting districts for those 10 counties - the Census Bureau seemed to limit the 2nd verification phase of the voting district project (VDP) to areas where the state had made change requests in one or more of the annotation and first verification phases, so even though that was before March 31 it was likely too late).  I mentioned that my main concern was that Dave's Redistricting used voting districts as the indivisible units (in their interface, at least; one can download any plan onto DRA) in what I'd say was the generally used format (I know Dave Bradlee also does a Block Group one, but most Atlas people use the Voting District one and Block Groups as you've noted in the past don't always coincide with municipalities either).  I said, "it can sometimes seem like people feel that if a line's not on Dave's Redistricting, it might as well not exist."

So I was basically asking the RDO to take extraordinary measures so I didn't have to look at people on this forum (and perhaps AAD as well) with not much knowledge of Maine doing redistricting plans on the App and treating it as the be all end all for amateur redistricting of the state.  So I can't be too upset with them for not obliging.  I had written Dave Bradlee last February after I heard from Maine's liaison about the absent (or effectively county-wide) "voting districts" in the 10 counties likely remaining and asked if he could do a hybrid instead.  But in that e-mail I threatened that I would make a post on this forum asking "Is Dave Bradlee mentally retarded?" if I claimed to have added Maine for post-2020 redistricting without my requested revisions, which he replied was very disrespectful of those with disabilities and that my approach "is the worst way of getting me to do something for you."  I then apologized, and he seemed to have forgiven me by last July when I forwarded him the bad news (from my perspective) from James Whitehorne and again apologized for having been a jerk in February, replying that the discussion was was interesting and giving me Kudos for pursuing it with the Census Bureau.  There's still no guarantee that he won't just use the voting districts as defined by the Census Bureau, though.

What would it take to generalize DRA so that if someone wanted to draw electoral districts for Hamburg or Manchester they could?

It's just different data sets. The mechanics of the App are the same.

Have you looked at DistrictBuilder https://www.districtbuilder.org/

It asks for a state, and if it is not on the list (e.g. Maine is not) to make a request. If you export a map, it generates a list of block number, district pairs. It seems that it might be possible to import such a list.

Are you familiar with the history of PL 94-171? Following 'Reynolds v Sims' a federal district court drew a map. Where they split a county they might follow precinct lines. Because the census data was based on enumeration districts, they might have to count houses to estimate the population within a district.

Following the 1970 Census, Alabama was still under the jurisdiction of the court. As now, Alabama held legislative elections every four years, and the next election was 1974. Alabama said we have plenty of time. The court said I am a federal court judge, and drew maps based on enumeration districts.

Enumeration districts are work units. They might not represent a community of interest. If an enumerator can do 200 houses, 140 might be in one area and 60 in the other. The judge said, no matter, and besides with the election not for 3 years, you have plenty of time to draw new election precincts, find polling places, and inform the voters that they will have to find a ride to the polls where they used to walk.

Similar things happened in other states where there was a struggle to maintain population equality between districts. As a result, Congress passed PL 94-171

https://uscode.house.gov/statutes/pl/94/171.pdf

"94" refers to the 94th Congress in 1975. If you read the statute, it is actually about negotiation between the States and the Census Bureau as to producing data usable for redistricting purposes by the states. You can think of the Block Boundary Suggestion Project (BBSP) as a negotiation, what a state might want for its election.

For the 1980 Census, the Census Bureau added Voting Tabulation Districts (VTD) for which the Census Bureau would tabulate data. But the Census Bureau did not permit use of artificial lines, so they weren't always useful, and they were useless for defining small districts such as wards. I think there were census blocks in some areas.

In 1990, the Census Bureau defined census blocks across the United States, and VTD's were no longer necessary for redistricting purposes, and state officials become indifferent to maintaining them, as has happened in Maine.

[jimrtex]

Can anyone figure out which post broke the thread on the last page and report it?

Better?

I wish a moderator could have disabled the post rather than simply removing a link. Presumably I provided the link for a reason other breaking the forum.

But maybe it was a matter of making sure the trains run on time.

[Tender Branson]

A court ruling bans the Census Bureau from releasing the 2020 Census data during the next 21 days.

Which could mean Raimondo is already confirmed as Sec. of Commerce then.

Census Bureau Statement on National Urban League Case 21-Day Stay

Yesterday, an official Order by the Court in the National Urban League case was issued in which the Department of Commerce and U.S. Census Bureau are defendants.  The Order enters a Stipulation for a 21-day stay in the case, and all necessary and appropriate steps must be taken to ensure compliance with the Court’s Stipulation and Order.  Accordingly, all Census Bureau staff are hereby instructed to read the order and ensure compliance with the following:

1) Reports, estimates, or data relating to the July 21, 2020, Presidential Memorandum on Excluding Illegal Aliens from the Apportionment Base following the 2020 Census will not be finalized, reported or publicly disclosed prior to the change of Administration on January 20, 2021. Should such information be finalized after the change of Administration but prior to the end of the proposed stay, Defendants would provide Plaintiffs with 7 days’ detailed notice prior to reporting or publicly disclosing it.

2) Reports, estimates, or data relating to Executive Order 13880, entitled “Collecting Information About Citizenship Status in Connection with the Decennial Census” (July 11, 2019), will not be finalized, reported or publicly disclosed prior to the change of Administration on January 20, 2021. Should such information be finalized after the change of Administration but prior to the end of the stay, Defendants would provide Plaintiffs with 7 days’ detailed notice prior to reporting or publicly disclosing it.

3) Neither the Census Bureau nor the Department of Commerce will report or publicly disclose any population counts or estimates relating to the population as of April 1, 2020, including counts or estimates of the illegal alien/undocumented immigrant population, prior to the change of Administration on January 20, 2021. To the extent such population counts or estimates are developed after the change of Administration but prior to the end of the stay, Defendants would provide Plaintiffs with 7 days’ detailed notice prior to reporting or publicly disclosing them.

https://www.census.gov/newsroom/press-releases/2021/national-urban-league-21-day-stay.html

[Frodo]

Trump's creature at the Census Bureau is out:

Census Bureau director stepping down after outcry over immigrant count
Pressure mounted on Steven Dillingham after an inspector general memo alleged he pressured employees to rush a report on the number of unauthorized immigrants.

[Kevinstat]

Census Bureau to Release 2020 Census Geographic Products

The U.S. Census Bureau announced today that 2020 Census redistricting (P.L. 94-171) geographic support products will be available for all states, the District of Columbia and Puerto Rico on a flow basis starting Jan. 19 and ending no later than Feb. 28.

These products consist of 2020 Census geography with the newly created 2020 Census blocks and updated block groups, census tracts, voting districts, and current boundaries for legal governments and school districts referenced to Jan. 1, 2020. Each state’s package will include shapefiles, maps, block assignment files and name lookup tables. These products are provided to support redistricting efforts by state and local governments.

In addition, a 2010 to 2020 block crosswalk will be provided with each state release to assist data users in comparing geographic data between censuses. Shapefiles, maps, block assignment files, name lookup tables, and the crosswalk will be provided in the same format as the 2010 Census version of these files.

The Census Bureau uses geography as the basis for conducting the census and tabulating census data. Census geography is a key component of the data products for decision-makers and data users. The Census Bureau collects, defines and maintains boundaries for over 29 different categories of geographic areas for statistical purposes, but this variety of geographic areas are often important to other federal, state and local agencies as well. Many agencies rely on the Census Bureau boundaries for incorporated places, minor civil divisions and counties since they are updated and maintained each year.

https://www.census.gov/newsroom/press-releases/2021/2020-census-geographic-products.html

(emphasis mine)

Were any states' geographic support products released today?  I thought I'd ask since today is the 19th.  But it may follow the 2011 pattern (for the actual data) with announcements on a Thursday as to which states will have the products released in the following week.  I don't know if the Census Bureau has to send these files directly to government officials in the states and wait for acknowledgement of receipt for these geographic files like they will with the data, but they may.

The order regarding the 21-day data release ban relating to the National Urban League lawsuit doesn't affect the release of the geographic support products, does it?

[jimrtex]

Census Bureau to Release 2020 Census Geographic Products

The U.S. Census Bureau announced today that 2020 Census redistricting (P.L. 94-171) geographic support products will be available for all states, the District of Columbia and Puerto Rico on a flow basis starting Jan. 19 and ending no later than Feb. 28.

These products consist of 2020 Census geography with the newly created 2020 Census blocks and updated block groups, census tracts, voting districts, and current boundaries for legal governments and school districts referenced to Jan. 1, 2020. Each state’s package will include shapefiles, maps, block assignment files and name lookup tables. These products are provided to support redistricting efforts by state and local governments.

In addition, a 2010 to 2020 block crosswalk will be provided with each state release to assist data users in comparing geographic data between censuses. Shapefiles, maps, block assignment files, name lookup tables, and the crosswalk will be provided in the same format as the 2010 Census version of these files.

The Census Bureau uses geography as the basis for conducting the census and tabulating census data. Census geography is a key component of the data products for decision-makers and data users. The Census Bureau collects, defines and maintains boundaries for over 29 different categories of geographic areas for statistical purposes, but this variety of geographic areas are often important to other federal, state and local agencies as well. Many agencies rely on the Census Bureau boundaries for incorporated places, minor civil divisions and counties since they are updated and maintained each year.

https://www.census.gov/newsroom/press-releases/2021/2020-census-geographic-products.html

(emphasis mine)

Were any states' geographic support products released today?  I thought I'd ask since today is the 19th.  But it may follow the 2011 pattern (for the actual data) with announcements on a Thursday as to which states will have the products released in the following week.  I don't know if the Census Bureau has to send these files directly to government officials in the states and wait for acknowledgement of receipt for these geographic files like they will with the data, but they may.

The order regarding the 21-day data release ban relating to the National Urban League lawsuit doesn't affect the release of the geographic support products, does it?

No release. They probably said the 19th because the 18th was a holiday, and they might have really meant "this week". There are scheduled news releases every Friday through July so I anticipate that will be when they announce the states for the following week. The period they gave was 41 days vs. 50 days in 2011, so I assume they will pretty much follow the same pattern.

They may have established the system when the data was distributed by physical media. Since the data is specifically for compliance with PL 94-171 which is for the state legislatures, I think they will continue to give notice this week, and then notice for each individual state. In 2011, they would also produce little notices about the state population, etc.

This was the pattern in 2011: Announce states that will be shipped the next week (in this case to LA, MS, NJ, and VA)

https://www.census.gov/newsroom/releases/archives/2010_census/cb11-cn05.html

Announcement that data has been shipped to each state (this is for MS)

https://www.census.gov/newsroom/releases/archives/2010_census/cb11-cn10.html

Announcement that state has received data, and public availability. This notice for MS also included the five most populous cities (Jackson, Gulfport, Southaven, Hattiesburg, and Biloxi), as well as some other data that a newspaper could use for a story. It also a county map showing population data. Since there will no data, they might figure something else. Perhaps a block map for Tupelo, MS.

https://www.census.gov/newsroom/releases/archives/2010_census/cb11-cn14.html

There is 21-day stay of proceedings in the district court case in California. The plaintiffs are asking for 1000s of documents during discovery, and the defendants have asked for a 3-judge panel (i.e. get rid of Koh).

As part of the stay, there was a stipulation that no data be released before January 20, and if there is any release after that it the plaintiffs be given 7-days notice, presumably so they can get an injunction. They probably expect a policy change after today.

States like Alabama might intervene seeking preservation of records. They clearly have a tangible interest in not losing a representative.

There are two sets of data at issue:

(1) Adjustment of apportionment counts for illegal aliens (persons not authorized to establish residence in the United States).

(2) Production of data sets of citizen and CVAP populations for all population levels (down to the block level).

The second could conceivably be tied to the geography if it included templates for the actual data. But they could just leave that out.

[Kevinstat]

2020 Census Geographic Products Available (for Arkansas, Colorado, Maryland, New Jersey and Virginia) (U.S. Census Bureau)

One can view maps or download shapefiles, etc. from the Decennial Census P.L. 94-171 Redistricting Data page.

[jimrtex]

2020 Census Geographic Products Available (for Arkansas, Colorado, Maryland, New Jersey and Virginia) (U.S. Census Bureau)

One can view maps or download shapefiles, etc. from the Decennial Census P.L. 94-171 Redistricting Data page.

These were released on January 19. The next two tranches have also been announced.

States Posted on 01/19/2021: AR, CO, MD, NJ & VA
States Posting on 01/26/2021: CA, IA, IN, LA, MO, NC, NY, OH, & OR
States Posting on 02/02/2021: DE, HI, ME, MS, NE, NV, PA, SD, WA, WI

NJ and VA are the two states holding legislative elections this year, and being first among the geographical products suggests that they may be first with the population data.

LA and MS also hold odd-year legislative elections, but they are on a 4-year cycle with the next elections in 2023.


In our discussion of staggered state senate terms did I ever mention that North Dakota truncates the terms of senators and representatives when there has been considerable change in the district.

So if odd-numbered districts were contested in 2020, and an odd-numbered district would be significantly changed in redistricting, the senator's term would be truncated to two years, and an election for a two-year term would be held in 2022.

This would mean that even-numbered districts contested in 2018 would be up in 2022 and there would be no restrictions on redistricting for 2022.

North Dakota also varies the number of districts, so for example, they might add a district in the Williston Basin rather than sliding districts into an area of high growth.


There is clear precedent for not redistricting four year (concurrent) when the first election would be in the -4 year.

Maryland elects legislators on a four year cycle. There was a companion case to Reynolds v Sims, where the SCOTUS explicitly stated that it was OK for Maryland to not be  redistricted until the 1966 (the other states, including Alabama were required to redistrict for 1964).

Alabama was (and is) on a four-year cycle, and was not redistricted until 1974 even though it was still under jurisdiction of the district court.


Am I misremembering, but wasn't Maine forced to do its congressional redistricting for the 2012 election, even though in the past it had waited for the -4 elections.

[Biden his time]

2020 Census Geographic Products Available (for Arkansas, Colorado, Maryland, New Jersey and Virginia) (U.S. Census Bureau)

One can view maps or download shapefiles, etc. from the Decennial Census P.L. 94-171 Redistricting Data page.

These were released on January 19. The next two tranches have also been announced.

States Posted on 01/19/2021: AR, CO, MD, NJ & VA
States Posting on 01/26/2021: CA, IA, IN, LA, MO, NC, NY, OH, & OR
States Posting on 02/02/2021: DE, HI, ME, MS, NE, NV, PA, SD, WA, WI

NJ and VA are the two states holding legislative elections this year, and being first among the geographical products suggests that they may be first with the population data.

LA and MS also hold odd-year legislative elections, but they are on a 4-year cycle with the next elections in 2023.


In our discussion of staggered state senate terms did I ever mention that North Dakota truncates the terms of senators and representatives when there has been considerable change in the district.

So if odd-numbered districts were contested in 2020, and an odd-numbered district would be significantly changed in redistricting, the senator's term would be truncated to two years, and an election for a two-year term would be held in 2022.

This would mean that even-numbered districts contested in 2018 would be up in 2022 and there would be no restrictions on redistricting for 2022.

North Dakota also varies the number of districts, so for example, they might add a district in the Williston Basin rather than sliding districts into an area of high growth.


There is clear precedent for not redistricting four year (concurrent) when the first election would be in the -4 year.

Maryland elects legislators on a four year cycle. There was a companion case to Reynolds v Sims, where the SCOTUS explicitly stated that it was OK for Maryland to not be  redistricted until the 1966 (the other states, including Alabama were required to redistrict for 1964).

Alabama was (and is) on a four-year cycle, and was not redistricted until 1974 even though it was still under jurisdiction of the district court.


Am I misremembering, but wasn't Maine forced to do its congressional redistricting for the 2012 election, even though in the past it had waited for the -4 elections.

So we won't receive the reapportionment data all in one file, but it'll be made public state-by-state on a flow basis?

[jimrtex]

2020 Census Geographic Products Available (for Arkansas, Colorado, Maryland, New Jersey and Virginia) (U.S. Census Bureau)

One can view maps or download shapefiles, etc. from the Decennial Census P.L. 94-171 Redistricting Data page.

This is a reference map for Arkansas House District 1, which appears to be relatively coincident with Texarkana, AR.

https://www2.census.gov/geo/maps/DC2020/PL20/st05_ar/legislativedistrict_maps/lowerchamber/sldl05001/PL20SLD_L05001.pdf

[jimrtex]

2020 Census Geographic Products Available (for Arkansas, Colorado, Maryland, New Jersey and Virginia) (U.S. Census Bureau)

One can view maps or download shapefiles, etc. from the Decennial Census P.L. 94-171 Redistricting Data page.

These were released on January 19. The next two tranches have also been announced.

States Posted on 01/19/2021: AR, CO, MD, NJ & VA
States Posting on 01/26/2021: CA, IA, IN, LA, MO, NC, NY, OH, & OR
States Posting on 02/02/2021: DE, HI, ME, MS, NE, NV, PA, SD, WA, WI

NJ and VA are the two states holding legislative elections this year, and being first among the geographical products suggests that they may be first with the population data.

LA and MS also hold odd-year legislative elections, but they are on a 4-year cycle with the next elections in 2023.


In our discussion of staggered state senate terms did I ever mention that North Dakota truncates the terms of senators and representatives when there has been considerable change in the district.

So if odd-numbered districts were contested in 2020, and an odd-numbered district would be significantly changed in redistricting, the senator's term would be truncated to two years, and an election for a two-year term would be held in 2022.

This would mean that even-numbered districts contested in 2018 would be up in 2022 and there would be no restrictions on redistricting for 2022.

North Dakota also varies the number of districts, so for example, they might add a district in the Williston Basin rather than sliding districts into an area of high growth.


There is clear precedent for not redistricting four year (concurrent) when the first election would be in the -4 year.

Maryland elects legislators on a four year cycle. There was a companion case to Reynolds v Sims, where the SCOTUS explicitly stated that it was OK for Maryland to not be  redistricted until the 1966 (the other states, including Alabama were required to redistrict for 1964).

Alabama was (and is) on a four-year cycle, and was not redistricted until 1974 even though it was still under jurisdiction of the district court.


Am I misremembering, but wasn't Maine forced to do its congressional redistricting for the 2012 election, even though in the past it had waited for the -4 elections.

So we won't receive the reapportionment data all in one file, but it'll be made public state-by-state on a flow basis?

The state populations for reapportionment will be issued all at once along with the new apportionments.

What is being released now is just the geography. You wouldn't want all the states in one file. Arkansas for example is 1.6 GB.

[Biden his time]

2020 Census Geographic Products Available (for Arkansas, Colorado, Maryland, New Jersey and Virginia) (U.S. Census Bureau)

One can view maps or download shapefiles, etc. from the Decennial Census P.L. 94-171 Redistricting Data page.

These were released on January 19. The next two tranches have also been announced.

States Posted on 01/19/2021: AR, CO, MD, NJ & VA
States Posting on 01/26/2021: CA, IA, IN, LA, MO, NC, NY, OH, & OR
States Posting on 02/02/2021: DE, HI, ME, MS, NE, NV, PA, SD, WA, WI

NJ and VA are the two states holding legislative elections this year, and being first among the geographical products suggests that they may be first with the population data.

LA and MS also hold odd-year legislative elections, but they are on a 4-year cycle with the next elections in 2023.


In our discussion of staggered state senate terms did I ever mention that North Dakota truncates the terms of senators and representatives when there has been considerable change in the district.

So if odd-numbered districts were contested in 2020, and an odd-numbered district would be significantly changed in redistricting, the senator's term would be truncated to two years, and an election for a two-year term would be held in 2022.

This would mean that even-numbered districts contested in 2018 would be up in 2022 and there would be no restrictions on redistricting for 2022.

North Dakota also varies the number of districts, so for example, they might add a district in the Williston Basin rather than sliding districts into an area of high growth.


There is clear precedent for not redistricting four year (concurrent) when the first election would be in the -4 year.

Maryland elects legislators on a four year cycle. There was a companion case to Reynolds v Sims, where the SCOTUS explicitly stated that it was OK for Maryland to not be  redistricted until the 1966 (the other states, including Alabama were required to redistrict for 1964).

Alabama was (and is) on a four-year cycle, and was not redistricted until 1974 even though it was still under jurisdiction of the district court.


Am I misremembering, but wasn't Maine forced to do its congressional redistricting for the 2012 election, even though in the past it had waited for the -4 elections.

So we won't receive the reapportionment data all in one file, but it'll be made public state-by-state on a flow basis?

The state populations for reapportionment will be issued all at once along with the new apportionments.

What is being released now is just the geography. You wouldn't want all the states in one file. Arkansas for example is 1.6 GB.

Yes, the state populations both in total and up to the census tract level are what I meant, whether those would be released as well in the same manner as the geographies. Thank you for the quick answer.

[Kevinstat]

2020 Census Geographic Products Available (for Arkansas, Colorado, Maryland, New Jersey and Virginia) (U.S. Census Bureau)

One can view maps or download shapefiles, etc. from the Decennial Census P.L. 94-171 Redistricting Data page.

These were released on January 19.

So they say, and probably so.  But when I checked Tuesday afternoon, and I'm pretty sure it was after 5 p.m., I didn't see that data online.  But I could have been looking in the wrong place.

The next two tranches have also been announced.

States Posted on 01/19/2021: AR, CO, MD, NJ & VA
States Posting on 01/26/2021: CA, IA, IN, LA, MO, NC, NY, OH, & OR
States Posting on 02/02/2021: DE, HI, ME, MS, NE, NV, PA, SD, WA, WI

NJ and VA are the two states holding legislative elections this year, and being first among the geographical products suggests that they may be first with the population data.

LA and MS also hold odd-year legislative elections, but they are on a 4-year cycle with the next elections in 2023.

So I get to look at Hudson (NY) census geography next week to tide me over for the Maine census geography a week later.

The next two batches are larger than the 2011 batches (for the actual data) were from what I recall.  The batches for the actual data might be smaller than that this year, but I imagine New Jersey and Virginia will be among the first.  Five states seems roughly in line with what I recall from 2011.  Four or five.  So maybe the batches including New York and Maine will each be divided into two batches for the actual data.

In our discussion of staggered state senate terms did I ever mention that North Dakota truncates the terms of senators and representatives when there has been considerable change in the district.

So if odd-numbered districts were contested in 2020, and an odd-numbered district would be significantly changed in redistricting, the senator's term would be truncated to two years, and an election for a two-year term would be held in 2022.

This would mean that even-numbered districts contested in 2018 would be up in 2022 and there would be no restrictions on redistricting for 2022.

North Dakota also varies the number of districts, so for example, they might add a district in the Williston Basin rather than sliding districts into an area of high growth.

I'm not sure if you had mentioned that about North Dakota.  Something to that effect in some state rings a bell.

It's interesting about the flexible number of legislators in North Dakota.  Maine allows for either 31, 33 or 35 state senators, and at the beginning the period for which that was the range we had 33, but we've had 35 since 1984 and I don't think anyone expects that to change, even if the Supreme Judicial Court (the official name for the state's top court) has to draw the lines.

There is clear precedent for not redistricting four year (concurrent) when the first election would be in the -4 year.

Maryland elects legislators on a four year cycle. There was a companion case to Reynolds v Sims, where the SCOTUS explicitly stated that it was OK for Maryland to not be  redistricted until the 1966 (the other states, including Alabama were required to redistrict for 1964).

Alabama was (and is) on a four-year cycle, and was not redistricted until 1974 even though it was still under jurisdiction of the district court.

Are you saying there's clear precedent for letting Legislators serve out the terms they are elected to?  (At least when all terms are expiring at the same time, and it seems like the courts are fine with that even when it results in "deferral".)  I was aware of that, and it's not surprising.  I had thought it might be different for a state like Mississippi with '1' year elections that sometimes doesn't get redistricting done in the '1' year, because of that happened in the 1990s (the terms elected in 1991 were actually commuted to 1 year, with Legislators elected for 3-year terms from new districts in 1992), but you informed me that the Legislators who were elected in 2011 from the old districts were able to complete their 4-year terms.

Am I misremembering, but wasn't Maine forced to do its congressional redistricting for the 2012 election, even though in the past it had waited for the -4 elections.

You are not misremembering.  Here is the thread on Maine congressional redistricting for the 2010s that I started after that lawsuit was launched.  There was no lawsuit against the then-existing Legislative districts being used in 2012, but in 2011 the Maine Constitution was amended so that starting after the 2020 Census, practically all* redistricting in Maine would happen in time for the '2' year elections, mostly in the '1' year**.

*The timetable for dividing Knox County into Budget Committee districts isn't specified in the Maine Constitution (they're one of two counties with directly elected finance/budget committees, and the other one (Aroostook) uses subdivisions of county commissioner districts so those districts are presumably redrawn whenever the CC districts are redrawn).

**The legislative part of the 2011 state constitutional amendment on redistricting was pretty much written by yours truly.  The congressional and county commissioner parts were added in a floor amendment and the county commissioner part was done somewhat sloppily, mostly just taking the then-existing statutes and adding 38 years (effectively subtracting 2 years "modulo 10"), so the county redistricting could happen as late as mid-April 2022 if the Clerk of the House took the maximum time allotted to prepare a legislative document and the Legislature failed to enact a plan so the Court had to do it.  The filing deadline for candidates of parties that qualify to hold primaries is (the first business day on or after) March 15 of even years.  Congressional and legislative redistricting in Maine have to occur in the '1' year now.

[○∙◄☻¥tπ[╪AV┼cVê└]

The state level numbers are currently scheduled for release next month.

[Tender Branson]

The U.S. Census Bureau is implementing the President’s Executive Order “Ensuring a Lawful and Accurate Enumeration and Apportionment Pursuant to the Decennial Census,” released January 20, 2021.

The 2020 Census data products including the P.L. 94-171 Redistricting Summary Data File will not include information on citizenship or immigration status.

The Census Redistricting and Voting Rights Data Office will reengage the Department of Justice’s Voting Section to confirm that Citizen Voting Age Population (CVAP) data produced from the American Community Survey (ACS) continue to meet its statistical needs.

The Census Bureau has released this product annually since 2011.

All work on the immigration status of the population of the United States regarding the 2020 Census was suspended on January 12, 2021.

https://www.census.gov/newsroom/press-releases/2021/executive-order-pl94-171-cvap.html

[jimrtex]

Yes, the state populations both in total and up to the census tract level are what I meant, whether those would be released as well in the same manner as the geographies. Thank you for the quick answer.

I'm guessing that the population data will be released the same way as in 2011, though I could be wrong.

They announced that they had released the geographic data for 5 states on Tuesday, and also announced 9 and 10 states for the next two weeks.

In 2011, they announced each Thursday what would be released the following week. They then announced each state individually on Monday, Tuesday, or Wednesday, and then posted the data publicly after getting confirmation that the state legislature had received the data. I'd guess that the states may be prickly about this privilege.

[Tender Branson]

The state level numbers are currently scheduled for release next month.

Says who ?

Last I heard was early-to-mid-March ...

[jimrtex]

The next two tranches have also been announced.

States Posted on 01/19/2021: AR, CO, MD, NJ & VA
States Posting on 01/26/2021: CA, IA, IN, LA, MO, NC, NY, OH, & OR
States Posting on 02/02/2021: DE, HI, ME, MS, NE, NV, PA, SD, WA, WI

NJ and VA are the two states holding legislative elections this year, and being first among the geographical products suggests that they may be first with the population data.

LA and MS also hold odd-year legislative elections, but they are on a 4-year cycle with the next elections in 2023.

So I get to look at Hudson (NY) census geography next week to tide me over for the Maine census geography a week later.

The next two batches are larger than the 2011 batches (for the actual data) were from what I recall.  The batches for the actual data might be smaller than that this year, but I imagine New Jersey and Virginia will be among the first.  Five states seems roughly in line with what I recall from 2011.  Four or five.  So maybe the batches including New York and Maine will each be divided into two batches for the actual data.

So it's all about ME?

You may want to look at BAF's, NLT's, and crosswalks.

The releases for 2011 has notices of releases.

https://www.census.gov/newsroom/releases/archives/2011.html

The size of the weekly groups at the end of 2011 were similar what has been announced so far, but it appears that there was more of a ramp up in 2011. The order is similar to that for 2011.

Jan 28, 2011: 4 (4/4 in first 3 releases 2021).
Feb 2-3, 2011: 5 (4/5)
Feb 9, 2011: 4 (1/4)
Feb 17, 2011: 8 (6/8)
Feb 24, 2011: 5 (3/5)
Mar 2, 2011: 7 (4/7)
Mar 10, 2011: 9 (0/9)
Mar 17, 2011: 10, including DC and PR (2/10)

It is possible that they are paying more attention to when the data can be used, rather than when the redistricting will need to be done.

IL and TX with the two earliest congressional primaries, and therefore filing deadlines, are not in the first three groups for 2021, but were in the third group for 2011. But having the districts set eight months before the filing deadline so would be candidates can begin fund-raising and hiring the best campaign consultants is not necessarily a concern of the Census Bureau.

States that were in the first 5 groups in 2011, which are not in the first three groups for 2021: Week 2: VT; Week 3: IL, OK, TX; Week 4: AL, UT; Week 5: KS, WY.

There were 26 states in the first 5 weeks in 2011, 24 states in the first three weeks in 2021. 18 of the 26 earlier states in 2011 are in the early three weeks in 2021.

States that were in the last 3 weeks in 2011, which have been advanced to the first 3 weeks in 2021: Week 6: CA, OH, PA, WI; Week 7: (none); Week 8: ME, NY.

There were 26 states (and territories) in the last 3 weeks in 2011. 20 of those 26 have not been announced for 2021.

You might be credited for moving ME up. In 2011 it was in the last group.

In our discussion of staggered state senate terms did I ever mention that North Dakota truncates the terms of senators and representatives when there has been considerable change in the district.

So if odd-numbered districts were contested in 2020, and an odd-numbered district would be significantly changed in redistricting, the senator's term would be truncated to two years, and an election for a two-year term would be held in 2022.

This would mean that even-numbered districts contested in 2018 would be up in 2022 and there would be no restrictions on redistricting for 2022.

North Dakota also varies the number of districts, so for example, they might add a district in the Williston Basin rather than sliding districts into an area of high growth.

I'm not sure if you had mentioned that about North Dakota.  Something to that effect in some state rings a bell.

It's interesting about the flexible number of legislators in North Dakota.  Maine allows for either 31, 33 or 35 state senators, and at the beginning the period for which that was the range we had 33, but we've had 35 since 1984 and I don't think anyone expects that to change, even if the Supreme Judicial Court (the official name for the state's top court) has to draw the lines.

It was fairly ordinary for state constitutions to provide for a range of legislators. This would permit the legislature to grow as new areas were filled in, and avoiding removing districts from slower growing areas. Sometimes when the maximum was reached, the constitution was amended to set a fixed number. This has happened in Texas, where at one time the range of the size of the House was 90 to 150. When other parts of the apportionment provisions were cleaned up, the count was simplified to 150.

North Dakota grew rapidly from the 1880s until about 1920, and then stagnated. Its 2010 population was less than 1930. While the overall population was static, there was a change in distribution as Fargo, Grand Forks, and Bismarck grew, while rural areas declined. Cass (Fargo) tripled from 1930-2010; Grand Forks (Grand Forks) doubled; Burleigh (Bismarck) nearly quadrupled; and Ward (Minot) doubled.

Changing the number of legislative districts (North Dakota elects representatives and senators from the same districts for four year terms with the same stagger) reduces the amount of change to many districts. North Dakota uses an odd-even system for the stagger, and can truncate terms, or elect to a 2-year term to get new districts back on schedule.

There is clear precedent for not redistricting four year (concurrent) when the first election would be in the -4 year.

Maryland elects legislators on a four year cycle. There was a companion case to Reynolds v Sims, where the SCOTUS explicitly stated that it was OK for Maryland to not be  redistricted until the 1966 (the other states, including Alabama were required to redistrict for 1964).

Alabama was (and is) on a four-year cycle, and was not redistricted until 1974 even though it was still under jurisdiction of the district court.

Are you saying there's clear precedent for letting Legislators serve out the terms they are elected to?  (At least when all terms are expiring at the same time, and it seems like the courts are fine with that even when it results in "deferral".)  I was aware of that, and it's not surprising.  I had thought it might be different for a state like Mississippi with '1' year elections that sometimes doesn't get redistricting done in the '1' year, because of that happened in the 1990s (the terms elected in 1991 were actually commuted to 1 year, with Legislators elected for 3-year terms from new districts in 1992), but you informed me that the Legislators who were elected in 2011 from the old districts were able to complete their 4-year terms.

I am saying that in states with four year terms and no stagger (all elected at once), redistricting can occur on a 12, 8, 12, 8, year cycle, taking effect at the normal time of election.

However, when Alabama was forced to redistrict in 1971, for an election that occurred in 1974, the federal district court was premature in its interference. It should not have intervened until it was certain that the Alabama would not draw new districts in time for the 1974 elections.

[Kevinstat]

So it's all about ME?

You may want to look at BAF's, NLT's, and crosswalks.

What are BAFs and NLTs?  I think BAFs (presumably "Block Assignment Files") might be the relationships between 2010 census blocks and 2020 census blocks, but I'm not sure about that and have no idea what NLT stands for, although I could look it up, so I may know the answer by the time you reply.  But your explanations on these kinds of things are cool so I won't answer my own question here.

You might be credited for moving ME up. In 2011 it was in the last group.

I'm sure the main reason for Maine being moved up is that we no longer delay the use of new districts to 2014.  In 2011 the Census Bureau was probably unaware that Republican lawyers had lined up plaintiffs from the seemingly faster-growing first congressional district (they might have lined up plaintiffs from the second district too, I don't know) in anticipation of a lawsuit forcing an early redraw, while Republicans had the trifecta.

Of course, as I noted, I was involved in getting the Maine Constitution amended to have all redistricting go into effect for the '2' year elections, so that may be why you were saying I might be credited for moving Maine up, rather than my more recent pestering of the Census Bureau.

It was fairly ordinary for state constitutions to provide for a range of legislators. This would permit the legislature to grow as new areas were filled in, and avoiding removing districts from slower growing areas. Sometimes when the maximum was reached, the constitution was amended to set a fixed number. This has happened in Texas, where at one time the range of the size of the House was 90 to 150. When other parts of the apportionment provisions were cleaned up, the count was simplified to 150.

The Maine House also.  Started at 141, with an established "pseudo-maximum" of 200 (when that size was reached, there would a popular vote every 10 years on increasing or reducing the size of the House (although 200 may have been a hard maximum, as the potential increasing could have been after some reductions following prior decennial votes; the minimum was only 100).  Went to 18_ (I forget the exact number) in the early 1830s and hit the maximum of 2000 after the 1840 census, and in 1841 (the last year of the 18_ House) the Legislature proposed and the voters approved (presumably at the same time they were electing a 200-member House) a Constitutional amendment setting the size of the House at 151, beginning with the Legislature to be elected in 1842.  So Maine had a 200-member State House for one year.

I am saying that in states with four year terms and no stagger (all elected at once), redistricting can occur on a 12, 8, 12, 8, year cycle, taking effect at the normal time of election.

Understood.  Also something I've been aware of, although some things I had written may have suggested otherwise, and it's good to have that on record for others here who may not have known that.

Members of the State and Local Government committee in Maine got tripped up on that in 2019 when they had to deal with a constitutional amendment resolution to have four-year concurrent and non-staggered terms for both houses of the Legislature.  There was concern about someone being drawn out of her/his district in a '1' year that was the first year of her/his term (you could still do the redistricting in the '1' year even if there was no '2' year election) being ineligible to continue serving, which seems silly when I write it that way but it seems to be the crux of the issue (and prior 4-year term bills had been written with 12-year redistricting cycles, which have obvious issues).  After the bill had died, with even the sponsor voting against the report of those on the committee who didn't vote straight ONTP ("Ought Not to Pass"), I sent an email to committee members and others involved in the discussion that you could simply have had something in the bill stating that the redistricting would take effect in 2022 and every 20 years thereafter and in 2034 and every 20 years thereafter (so alternating 12- and 8-year lives of a plan; elections were to be in Gubernatorial election years in that bill).  The committee had inserted staggering within a decade (with the Illinois Senate style 2-4-4, 4-4-2 and 4-2-4 cycles), which the sponsor of the bill was adamantly against.  That could have just been what the committee wanted to do, but they seemed to have been led to believe by government officials that the initial bill with 10-year redistricting cycles was problematic (I would have just assumed that that meant alternating 8- and 12-year lives of district lines).

Interestingly, in 2017 there had been a bill for four-year terms, with no staggering within either chamber but with State Representatives and State Senators being elected 2 years apart (Representatives elected in Gubernatorial election years and Senators elected in Presidential election years).  The bill as initially drafted had a 12-year redistricting cycle, but with that bill the Deputy Secretary of State who heads the Bureau of Corporations, Elections and Commissions suggested a 10-year cycle (which seemed to correspond with what you are calling a "12, 8, 12, 8, year cycle").  There it seemed to be assumed that people could serve out the term they were elected to even as members in the other chamber were elected from new districts (and the new districts would likely have been drawn for the first chamber as well, and there might be some pairs of incumbents from some districts in that chamber).  Maybe it was the fact that at least some people were to be elected in the year following redistricting that made things easier to grasp, while having a '1' year redistricting with elections not until the '4' year every other decade seemed to snag in a lot of people's minds.

[jimrtex]

You may want to look at BAF's, NLT's, and crosswalks.

What are BAFs and NLTs?  I think BAFs (presumably "Block Assignment Files") might be the relationships between 2010 census blocks and 2020 census blocks, but I'm not sure about that and have no idea what NLT stands for, although I could look it up, so I may know the answer by the time you reply.  But your explanations on these kinds of things are cool so I won't answer my own question here.

They are linked to from here.

https://www.census.gov/programs-surveys/decennial-census/about/rdo/summary-files.html#P2

BAF are block assignment files. For Arkansas there are BAF for Congressional Districts, Legislature Districts Lower and Upper, Unified School Districts, Incorporated Place, and Voting Districts (VTD). These are text files with block number-area ID pairs. For entities that do not cover the state, such as incorporated places, there will not be a PlaceID for blocks outside any incorporated place. So a BAF has as many rows as there are census blocks in a state.

The entities vary by state. For New Jersey there are Minor County Subdivision (MCD), Elementary and HS School Districts, and AIANNH areas.

NLT are Name Lookup Tables. They are pairs of IDs and names. So for Arkansas there is an entry.

05|41000|Little Rock|Little Rock city

Arkansas is "05". The ID (within AR) for Little Rock is 41000. There is both a name for use on maps, and a formal name.

Then in the BAF one line is:

051190005001000|41000

05 is Arkansas
119 is Pulaski County.
000500 is Census Tract 5 (remember the last two digits of a tract are a suffix, with .00 suppressed in visual displays as well as leading zeros).
1000 is block 1000 of Census Tract 5.

41000 says that this census block is part of Little Rock.

So perhaps you could create BAF for Maine voting districts and feed them to DRA. But there also are shapefiles for VTD, so maybe you could produce an alternate set of shapefiles.

You can think of DRA as manipulating Redistricting Atoms (RDA) such as VTD or Census Tracts. The RDA can be combined into districts. Associated with an RDA are its geographical outline and demographic and possibly election data. When you combine RDA, you combine the areas and the demographic and election data.

The RDA hide the underlying component census blocks, just as an atom hides details of the protons and neutrons. I think DRA lets you work at a county level, which is an aggregation of RDA, equivalent to a molecule. It is a way to let you select the groups of RDA.

It might be possible to use an alternate set of RDA with DRA.

Block cross-walk tables show the relationship between 2010 blocks and 2020 block.

There can be simple transformations: that are 1:1, 1:N, and N:1.

A 1:1 transformation is the block is unchanged, with perhaps a renumbering. So that Block A (2010) is Block B (2020).

A 1:N transformation is a division of 2010 block into 2 or more blocks. This might occur in when an area is subdivided, and new streets are identified.

Block A (2010) (part) is Block B (2020)
Block A (2010) (part) is Block C (2020)
Block A (2010) (part) is Block D (2020)
etc.

A N:1 transformation is a consolidation of 2010 blocks into a single 2020 block. In 2010, people went crazy converting driveways, etc. into block boundaries, creating bunches of unpopulated blocks (e.g. see Hudson cemetery). So you might have:

Block A (2010) is Block Z (2020) (part)
Block B (2010) is Block Z (2020) (part)
Block C (2010) is Block Z (2020) (part)
etc.

And there may be complex changes, such as when two adjacent blocks are transformed into three blocks, with the third block being formed from part of both, and the remnants of the 2010 blocks becoming 2020 blocks.

If nothing else, it might be a quick way to identify changes in the 2020 geography in Hudson or Augusta.

You might be credited for moving ME up. In 2011 it was in the last group.

I'm sure the main reason for Maine being moved up is that we no longer delay the use of new districts to 2014.  In 2011 the Census Bureau was probably unaware that Republican lawyers had lined up plaintiffs from the seemingly faster-growing first congressional district (they might have lined up plaintiffs from the second district too, I don't know) in anticipation of a lawsuit forcing an early redraw, while Republicans had the trifecta.

Of course, as I noted, I was involved in getting the Maine Constitution amended to have all redistricting go into effect for the '2' year elections, so that may be why you were saying I might be credited for moving Maine up, rather than my more recent pestering of the Census Bureau.

The PL 94-171 data is used for legislative and local redistricting in addition to congressional redistricting.

The order of data release appears to be based on need. In 2011, ME apparently did not need the data until 2013. It also appears that a consideration in 2011 was when the new districts would first be used (for filing and primaries). It may be that for 2021 that when the state can start drawing districts is more significant.

The order between 2011 and 2020 is highly correlated. ME and CA are outliers in moving up. IL and TX are outliers in moving down. Note that CA has moved their primary three  months earlier - but it might also be a consideration that the redistricting commission has moved up its schedule. Incidentally, a former Census Director was a member of the California Redistricting Commission in 2011+.

The California redistricting law permits the commission to recommend changes in the law for the next commission. One recommendation was to start earlier so that COI could be identified earlier. If a COI does in fact exist, it exists for purposes other than drawing districts. Instead of districts creating COI, they should recognize COI.

In 2011, the commission was advised to accept the boundaries proposed for advocates for the community. But you might not be able to distinguish between Sleazy Partisan Political Hacks and Community Advocates.

It was fairly ordinary for state constitutions to provide for a range of legislators. This would permit the legislature to grow as new areas were filled in, and avoiding removing districts from slower growing areas. Sometimes when the maximum was reached, the constitution was amended to set a fixed number. This has happened in Texas, where at one time the range of the size of the House was 90 to 150. When other parts of the apportionment provisions were cleaned up, the count was simplified to 150.

The Maine House also.  Started at 141, with an established "pseudo-maximum" of 200 (when that size was reached, there would a popular vote every 10 years on increasing or reducing the size of the House (although 200 may have been a hard maximum, as the potential increasing could have been after some reductions following prior decennial votes; the minimum was only 100).  Went to 18_ (I forget the exact number) in the early 1830s and hit the maximum of 2000 after the 1840 census, and in 1841 (the last year of the 18_ House) the Legislature proposed and the voters approved (presumably at the same time they were electing a 200-member House) a Constitutional amendment setting the size of the House at 151, beginning with the Legislature to be elected in 1842.  So Maine had a 200-member State House for one year.

Once upon a time, Massachusetts apportioned as many representatives for a town as the town was willing to financially support. This might have been prior to 1820. It is unlikely that towns in Maine could afford to support multiple representatives in Boston.

I am saying that in states with four year terms and no stagger (all elected at once), redistricting can occur on a 12, 8, 12, 8, year cycle, taking effect at the normal time of election.

Understood.  Also something I've been aware of, although some things I had written may have suggested otherwise, and it's good to have that on record for others here who may not have known that.

Members of the State and Local Government committee in Maine got tripped up on that in 2019 when they had to deal with a constitutional amendment resolution to have four-year concurrent and non-staggered terms for both houses of the Legislature.  There was concern about someone being drawn out of her/his district in a '1' year that was the first year of her/his term (you could still do the redistricting in the '1' year even if there was no '2' year election) being ineligible to continue serving, which seems silly when I write it that way but it seems to be the crux of the issue (and prior 4-year term bills had been written with 12-year redistricting cycles, which have obvious issues).  After the bill had died, with even the sponsor voting against the report of those on the committee who didn't vote straight ONTP ("Ought Not to Pass"), I sent an email to committee members and others involved in the discussion that you could simply have had something in the bill stating that the redistricting would take effect in 2022 and every 20 years thereafter and in 2034 and every 20 years thereafter (so alternating 12- and 8-year lives of a plan; elections were to be in Gubernatorial election years in that bill).  The committee had inserted staggering within a decade (with the Illinois Senate style 2-4-4, 4-4-2 and 4-2-4 cycles), which the sponsor of the bill was adamantly against.  That could have just been what the committee wanted to do, but they seemed to have been led to believe by government officials that the initial bill with 10-year redistricting cycles was problematic (I would have just assumed that that meant alternating 8- and 12-year lives of district lines).

Interestingly, in 2017 there had been a bill for four-year terms, with no staggering within either chamber but with State Representatives and State Senators being elected 2 years apart (Representatives elected in Gubernatorial election years and Senators elected in Presidential election years).  The bill as initially drafted had a 12-year redistricting cycle, but with that bill the Deputy Secretary of State who heads the Bureau of Corporations, Elections and Commissions suggested a 10-year cycle (which seemed to correspond with what you are calling a "12, 8, 12, 8, year cycle").  There it seemed to be assumed that people could serve out the term they were elected to even as members in the other chamber were elected from new districts (and the new districts would likely have been drawn for the first chamber as well, and there might be some pairs of incumbents from some districts in that chamber).  Maybe it was the fact that at least some people were to be elected in the year following redistricting that made things easier to grasp, while having a '1' year redistricting with elections not until the '4' year every other decade seemed to snag in a lot of people's minds.

In Washington, senators and representatives are elected from the same legislative districts (the constitution requires nesting, it does not preclude drawing subdistricts for representatives, but the modern practice is to not do so).

Senators are elected by halves for four-year terms, while representatives are elected for two year terms.

So in 2022, there will be some senator holdovers in the 2010 legislative districts, while representative for the same district will be elected from the 2020 legislative districts. If I am not mistaken, any senate special elections would be held using the old boundaries.

Of course, Maine wouldn't have the problem with nesting.

This can result in a difference in procedure. In Texas, the Texas Secretary of State presides over the House of Representatives until the House can organized itself.

Meanwhile, in the Senate the holdover senators answer a roll call, and then the new members are sworn in.

[jimrtex]

The Census Bureau has posted the geographical data for the second group on 1/25 (though there doesn't seem to be a news release). Instead it is on their PL 94-171 web page.

https://www.census.gov/programs-surveys/decennial-census/about/rdo/summary-files.html#P2

In addition, they announced ten states that will be released in the fourth tranche.

States Posted on 01/19/2021: AR, CO, MD, NJ, VA
States Posted on 01/26/2021: CA, IA, IN, LA, MO, NC, NY, OH, OR
States Posting on 02/02/2021: DE, HI, ME, MS, NE, NV, PA, SD, WA, WI
States Posting on 02/09/2021: AL, AK, AZ, CT, IL, MA, MN, MT, OK, TX

This will bring the total for the four groups to 34 states, which will likely mean completion in 6 weeks. In 2011, the data AND geographical support was released over 8 weeks.

So far, the release order is correlated with the 2011 order:

2021 Release 1: Average Week 2.0 in 2011.
2021 Release 2: Average Week 4.2 in 2011.
2021 Release 3: Average Week 4.6 in 2011
2021 Release 4: Average Week 5.4 in 2011.
Unannounced for 2021: Average Week 6.7 in 2011.

2011 Release Week 1: 1, 1, 2, 3 in 2021
2011 Release Week 2: 1, 1, 2, 2 in 2021 (all but VT)
2011 Release Week 3: 3, 4, 4,4
2011 Release Week 4: 1, 2, 2, 3, 3, 3, 4 (all but UT)
2011 Release Week 5: 2, 3, 3 (all but KS and WY)
2011 Release Week 6: 2, 2, 3, 3, 4, 4 (all but ID)
2011 Release Week 7: 4, 4, 4 (FL, GA, KY, NM, ND, TN not announced)
2011 Release Week 8: 2, 3, 4 (MI, NH, RI, SC, WV, DC, PR not announced)

[Rep Jessica]

When will we know the official population of the united states for 2020? When will we know that for each state?

[Kevinstat]

The Census Bureau has posted the geographical data for the second group on 1/25 (though there doesn't seem to be a news release). Instead it is on their PL 94-171 web page.

https://www.census.gov/programs-surveys/decennial-census/about/rdo/summary-files.html#P2

In addition, they announced ten states that will be released in the fourth tranche.

States Posted on 01/19/2021: AR, CO, MD, NJ, VA
States Posted on 01/26/2021: CA, IA, IN, LA, MO, NC, NY, OH, OR
States Posting on 02/02/2021: DE, HI, ME, MS, NE, NV, PA, SD, WA, WI
States Posting on 02/09/2021: AL, AK, AZ, CT, IL, MA, MN, MT, OK, TX

This will bring the total for the four groups to 34 states, which will likely mean completion in 6 weeks. In 2011, the data AND geographical support was released over 8 weeks.

So far, the release order is correlated with the 2011 order:

2021 Release 1: Average Week 2.0 in 2011.
2021 Release 2: Average Week 4.2 in 2011.
2021 Release 3: Average Week 4.6 in 2011
2021 Release 4: Average Week 5.4 in 2011.
Unannounced for 2021: Average Week 6.7 in 2011.

2011 Release Week 1: 1, 1, 2, 3 in 2021
2011 Release Week 2: 1, 1, 2, 2 in 2021 (all but VT)
2011 Release Week 3: 3, 4, 4,4
2011 Release Week 4: 1, 2, 2, 3, 3, 3, 4 (all but UT)
2011 Release Week 5: 2, 3, 3 (all but KS and WY)
2011 Release Week 6: 2, 2, 3, 3, 4, 4 (all but ID)
2011 Release Week 7: 4, 4, 4 (FL, GA, KY, NM, ND, TN not announced)
2011 Release Week 8: 2, 3, 4 (MI, NH, RI, SC, WV, DC, PR not announced)

Since it wasn't specifically posted here yet, I figured I'd post all the states in each week in 2011.  I'm going by week rather than date as with the dates I wonder whether I should use the date the coming data for the states was announced (that's what jimrtex used in his post on Saturday morning), the date the data is delivered to the states or the date it's made public.

2011 Release Week 1: LA, MS, NJ, VA
2011 Release Week 2: AR, IA, IN, MD, VT
2011 Release Week 3: IL, OK, SD, TX
2011 Release Week 4: AL, CO, HI, MO, NV, OR, UT, WA
2011 Release Week 5: DE, KS, NC, NE, WY
2011 Release Week 6: AZ, CA, CT, ID, OH, PA, WI
2011 Release Week 7: AK, FL, GA, KY, MN, MT, ND, NM, TN
2011 Release Week 8: MA, ME, MI, NH, NY, RI, SC, WV, DC, PR

Also, assuming the 2021 geographical release is 6 weeks, here is the reverse of the averaging jimrtex did regarding the 2011 and "2021 Geo" correlation (with ranges where applicable), rounded to the nearest 0.1 except if it's exactly at an 0.05 interval:

2011 Release Week 1: Average Week 1.75 for 2021 Geo
2011 Release Week 2: Average Week 2.2 or 2.4 for 2021 Geo
2011 Release Week 3: Average Week 3.75 for 2021 Geo
2011 Release Week 4: Average Week 2.9 or 3.0 for 2021 Geo
2011 Release Week 5: Average Week 3.6 to 4.0 for 2021 Geo
2011 Release Week 6: Average Week 3.3 or 3.4 for 2021 Geo
2011 Release Week 7: Average Week 4.7 to 5.3 for 2021 Geo
2011 Release Week 8: Average Week 4.4 to 5.1 for 2021 Geo

So not quite as well correlated going in that direction.  Still a decent correlation.