Results of the 2010 Census

[jimrtex]

After the 2010 Census, when will we know the results?  When will the new Congressional Districts be effective?

The State apportionment population totals were delivered to the President on December 28, 2000.

The data necessary to do redistricting, which included population counts by race and Hispanicity down to the block level was delivered to the States during March 2001.  The mapping data had been delivered earlier.   Generally, an attempt was made to deliver the data to States that had earlier deadlines, but all data was distributed over a roughly 3-week period. Census 2000 Archive of Product Release Dates

In Texas, the legislature only meets in odd years and has a limited session length that only extends into May.  Further, the filing period begins in December with the deadline in early January, so it was toward the beginning of the list, even though it is a more populous State.

The first lawsuits were filed in December 2000 after the apportionment numbers were released, alleging that Texas did not have 32 congressional districts, that the existing districts did not have 1/32 of the population, and that they were based on 1990 census data.  The legislature did not pass any districting plans, and the Legislative Redistricting Board drew the legislative districts during the summer of 2001.  Because the Democrats were intent on having the congressional districts drawn by the courts, no special sessions were called, and the district court drew the districts in late 2001 in time for representatives to file.

[jimrtex]

So, let me see if I have this correct?

The results will start coming in on April 1, 2010 with state numbers coming in April, May, and June with the new CD's and EV's being effective for the 2012 Presidential Cycle.

So, I guess I should see my Census form in the mail in early January 2010, correct?

No.  The Census is based on April 1, 2010.

In 2000, census forms were mailed starting on March 13.  Some early operations in the Alaska interior started as early as January.

The Census bureau reported that 61% of households had sent their completed form in by April 11.  Ultimately this would reach 66%,  Near the end of April, census workers started contacting the 42 million households that had not been heard from.  This stage was reported as completed at the end of June.  They then begin on their quality control, which could include followups on forms that made no sense.

The State counts and apportionment were issued at the very end of 2000.  No breakdowns on individual counties, cities, blocks, etc. were released until March 2001, at which time legislatures could begin redistricting.

Legally, any redistricting that was not done in time for 2010 could be legally challenged as a violation of one man/one vote.  Conceivably there could be challenges in States like Louisiana, Mississippi, Kentucky, and Virginia that hold odd year legislative elections in 2011.  But they could probably argue that they didn't have time to redistrict in time for the districts to be reviewed by the DOJ, the courts, voting precincts adjusted, candidates filing, and primaries.  So they will probably won't be in effect until 2013.  In other states, and congressional districts will be for 2012.  But Maine waits until 2014.

[jimrtex]

My understanding is that Maine's late re-districting is more a tradition, and that if anyone wanted to challenge it in court they would win.  Any challenge would be based on equal protection grounds, which would override anything in the State constitution or law.

[jimrtex]

It's a tradition that only began in the 1970s for the State House - after the bulk of the earth-shattering "Apportionment Decisions" had been handed down - and in the 1980s for the State Senate.

Has the Supreme Court explicitly ruled (for Congress, the Legislature, or both) that redistricting has to be done by the first election after the numbers from the latest census come in, rather than merely once every decade (which they might have assumed states would understand to mean right after the census)?  If so I'd like to hear what the case or cases are that established that rule.  That could be useful in getting Legislative action on moving our redistricting ahead (earlier) two years next year (the last sensible time to try to get that done for the next post-census redistricting, as there would be hardly enough time to redistrict in 2012 after the necessary Constitutional amendment referendum was passed in 2011 - although a state Constitutional amendment wouldn't be necessary to change our congressional redistricting cycle which is in statute only).

I'd start with the US Code mandating election from districts.  It supersedes 2 USC §2a (c), which has a grandfather provision for States that have had no change in representation, permitting them to continue to use existing districts until redistricting was done.

You must be logged in to read this quote.

Since Congress changed the existing law, it is clear that their intent was that the districts be established subsequent to the census, even when there was no change in representation.  While popularly known as a "reapportionment", the apportionment of representatives is done without regard to any previous apportionment.  Maine is apportioned 2 representatives on the basis of the 2000 Census.  That it previously had 2 representatives based on the 1990 Census is of no significance.

In cases dealing with population equality, the Supreme Court has developed a standard that districts have population as equal as "practicable", which it has then interpreted as being able put into practice.  Given block-level census data and computer-aided districting software, it is practicable to have a deviation of 1 person between districts. 

Maine could probably avoid having to split towns as long as it chose towns along the boundary that would provide the best equality, even if this made the boundary more irregular.  Maine could argue that this simplified administration of elections, reduced voter confusion etc., and had no material inequality effect.  But it doubtful that a court would accept district boundaries for use in 2012, that were based on the 2002 Census. 

In the Supreme Court case dealing with the 2003 redistricting of Texas congressional districts, LULAC v Perry, one challenge was on the basis that the legislature had used population data from the 2000 Census.  It is understood that the use of census data is a legal fiction - that everyone knows that the data wasn't accurate to begin with, and that it had changed since the time of the census - yet it is still valid for determining population equality and compliance with the 14th Amendment.  And further, when the courts or a legislature have redrawn district boundaries to comply with Voting Rights Act, or other legal requirement, they have used census data, even if it were 8 years out of date.  Because the courts or the legislature had to draw a lawful plan, they were forced to use the only data available to them. 

But then the argument continued, that when a legislature voluntarily redistricted, that they could no longer rely on census data.  If a legislature redistricted in 2003, they would have to demonstrate that in 2003 that the districts had equal population.  If this meant the legislature conducting its own census, so be it.  The Supreme Court rejected this argument, but this does not mean that a State could continue to use districts based on a previous census, once a new census had been conducted.

If the congressional district boundaries in Maine were challenged, it would not be on the basis that the legislature had not redistricted - but rather that Maine was going to use districts in 2012 that were demonstrably not equal in population, and despite the availability of better data on which to base them on, and to comply with federal law.  An argument that it was impracticable, because of a provision in State law isn't going go hold water. 

New districts in Maine will not go into effect until 2015.  If there is a special election up until January 3, 2015, it would be conducted using the current boundaries.  So Maine would have to justify using 15-year old data, when 5-year old data was available.  At the time of 2000 Census, its two districts were off by 4.6% from the average, over 9% between the two.  If we assume a continued linear shift, the difference was almost 15% by the time new 2000-based districts came into effect, and the new districts may already have been different by almost 5% when they were first used in 2005.

Maine could argue that they wanted stable district boundaries, and if they changed districts for the 2012 election, it would be after only 4 elections with the old boundaries.  But that is a superficial difference, and is a one-time transition.

Their best argument would probably be that they want to make the legislative and congressional districts at the same time to avoid administrative and voter confusion.  But it is unlikely that changing the congressional district boundaries would impose any burden other than a few town clerks having to prepare the ballot for a different congressional district than they had the previous election.  While changes for the legislative districts are more extensive, changing the congressional district of perhaps 5% of the people is not going to add or subtract from the confusion resulting from all the legislative district boundaries.  Because of the small size of the House districts, there is probably going to be wholesale reshuffling.  It is simply not feasible to make small changes when varying growth rates throughout the state will mean complete elimination of districts in some areas, and creation of new districts in other areas.

[jimrtex]

I'd start with the US Code mandating election from districts.  It supersedes 2 USC §2a (c), which has a grandfather provision for States that have had no change in representation, permitting them to continue to use existing districts until redistricting was done.

How can anything other than a constitutional amendment (or a new or previously unrevealed Supreme Court interpretation of the U.S. Constitution including its amendments) supersede the U.S. Constitution?  Perhaps you didn't mean what you wrote above exactly the way you wrote it.  Or perhaps 2 USC §2a (c) merely makes it clear that there is not a Constitutional requirement for states to redraw congressional districts when there is no change in representation, but doesn't make it unconstitutional for federal law (which trumps state constitutions and laws) to require such a redrawing of congressional districts.  That would make sense.

I'm saying that one section of US Code, 2 USC §2c,  takes precedence over another section of US Code, 2 USC §2a (c).

Both 2 USC §2c  and 2 USC §2a (c) require election from congressional districts.   But the two sections are in conflict.  Since 2 USC §2c was added after 2 USC §2a (c), it should be presumed to take precedence, and reflect the intent of Congress to establish a more restrictive process.   In addition, 2 USC §2c was added a few years after Wesberry v Sanders, which established, under Article I of the Constitution, that the population of congressional districts be as equal as practicable (Wesberry is not based on the equal protection clause of the 14th Amendment.).

Let's first look at what 2 USC §2a (c) says.  It says that a State should elect representatives from districts that are established after an apportionment.  But it provides a number of exceptions, including:

• If a State gains representatives, it shall continue to elect its representatives from the previously established districts, and elect the added representatives at large.  Note that this does not permit a State to switch to a combination of district and at large representatives, or to change the district boundaries while also electing other representatives are elected at large.  It is just a stop gap until the State redistricted.  However States sometimes went several elections or even censuses under this provision, or similar earlier versions.
• If a State loses representatives, it shall elect all of its representatives at large, until it manages to draw new districts.
• If a State loses representative, but there are the correct number of districts, then it shall elect the representatives from the old districts.  This might happen if a State gained a representative at one census, then never bothered to redistrict, but rather simply elected the additional representative at large, and then lost a representative in a subsequent census.
• If there is no change in representation, a State shall continue to elect its representatives from existing districts.

There is nothing optional about 2 USC §2a (c).  It is Congress directing the States in the manner of electing representatives after a new apportionment.  The only option that a State had was to not bother with updating its districts.

Georgia lost 2 representatives (from 12 to 10) following the 1930 census.  It drew 10 districts in 1931.  After the 1960 census, it still had 10 districts, and was still using the district boundaries drawn in 1931.   By that time, a district comprised of Fulton, DeKalb, and Rockdale counties had 3 times the population of the least populous district, and twice the population of the average district.  In Wesberry v Sanders, the Supreme Court ruled that the district boundaries violated the US Constitution, even though Georgia had fully complied with the manner regulations provided by Congress in  2 USC §2a (c), and similar earlier provisions in US law.

Georgia's districts were unconstitutional not because Georgia had not drawn new districts in 30 years, but rather because the districts were not as equal in population as practicable.  Georgia could have drawn new district boundaries in 1960, that were not equal enough.  Following the 1980 census, New Jersey drew new congressional district boundaries, that were thrown out by Supreme Court because they were not equal enough.

Wesberry was decided in 1964.  A few years later, Congress enacted 2 USC §2c.  It says that following an apportionment, that a State shall elect its representatives from districts established on the basis of the apportionment.  It does not say anything about equality, but that can be presumed, on the basis of Wesberry.  Since 2 USC §2c provides none of the exceptions of 2 USC §2a (c), it means the conflicting exceptions no longer exist.  The one exception in 2 USC §2c applied only to the 1968 election, and permitted Hawaii and New Mexico to elect their 2 representatives each at large for one final election.  New Mexico actually used districts in 1968, though not required to do so.  Hawaii elected its two representatives at large in 1968, and then used two districts in 1970.

Under 2 USC §2c the sequence of events is:

(1) Census.

(2) Apportionment.

(3) State creates new districts.  These must comply with equal population standards, the Voting Rights Act, etc.

(4) State elects number of representatives apportioned to it on the basis of the latest census.

If Maine elects 2 representatives in 2012, it will be because it was apportioned 2 representatives on the basis of the 2010 census - not because it was apportioned 2 representatives in 2000, or because it had district maps left over from the previous decade.  Congress has dictated the manner in which Maine will conduct its election of representatives in 2012.  They shall be elected from districts established subsequent to the apportionment following the census. 

If Maine uses the old district boundaries in 2012, it will be in violation of 2 USC §2c.  Congress carefully and deliberately rewrote the exception under 2 USC §2a (c), to eliminate the exception that Georgia had used to avoid redistricting from 1931 to 1965.  But it also meant that Maine can not use the same districts drawn for the 2004 election, for the 2012 election.

Under US code, Maine need not draw two equal population districts.  It must draw two districts subsequent to the census.  But any court will require the districts to be equal in population based on Wesberry and subsequent decisions.

[jimrtex]

What about Legislative districts?  How sure a winner would a suit against the current existing State House and State Senate being used in 2012 be (assuming there is more than a 10% deviation between those districts as of the 2010 census, which there will be)?  You may have tried to answer that question but it got lost by me among everything else.  (I'm sure many people experience that with my posts. ) If the districts were thrown out in 2012 on population grounds (a suit lauched too soon before that might be determined to be not ripe for review), coming up with a practicable solution that would abide by Maine's Constitution as much as possible while passing federal muster would be very difficult, much moreso than for Congress where you could extend the filing deadline if necessary for the first district candidates to collect more signatures from the smaller district depending on when the ruling went down (a suitable map could probably be drawn that would only shift territory from the first district to the second).

Is Maine required to use the federal census for purposes of drawing its legislative district boundaries?   No.

May Maine elect legislators in 2008 or 2010 from districts that were drawn on the basis of the 2000 census, and that quite likely have greater population deviation than is generally regarded by the Supreme Court as lawful.  I suspect that Maine can legally do so.  So what is different about 2012?  Why are districts drawn based on 10-year old data OK, and not districts using 12-year old data.

Is it because that we couldn't easily prove in 2008 that there was an excessive deviation?  What happens if the ACS is determined to be reliable enough for districting?  Must districting then be done before every election?  The city of Houston, Texas does redistrict every two years, so it is not totally impractical.

Or if deviation in excess of 5% violates equal protection, why can't Maine be required to conduct more frequent censuses?  Why couldn't someone in early 2001 sue to overturn the November 2000 elections that were conducted on boundaries that could be shown to have excessive variation in population?

I don't think there is a clear cut case against Maine in the case of legislative district boundaries being drawn for the 2014 election, rather than the 2012 election.

[jimrtex]

No.  The Census is based on April 1, 2010.

In 2000, census forms were mailed starting on March 13.  Some early operations in the Alaska interior started as early as January.

The Census bureau reported that 61% of households had sent their completed form in by April 11.  Ultimately this would reach 66%,  Near the end of April, census workers started contacting the 42 million households that had not been heard from.  This stage was reported as completed at the end of June.  They then begin on their quality control, which could include followups on forms that made no sense.

Interesting theoretical question (maybe you know the answer):

If the Census forms are mailed out starting in March, are you allowed to send them back before April 1, or do they have to be postmarked April 1 and later to count in the Census ?

I´m asking, because the official Day of Count is April 1, and let's say you send your form back on March 27, yet your child is born March 28, does your child that was born a day later count in the Census ?

Probably not, because the Census Bureau will never know that the child was born a day later and therefore it won't count. Or does the Census Bureau have any method to include these cases ?

The question is, if not already implemented, are the April 1 postmarked forms even better for the accuracy of the Census ? Because in a 2 week timespan, about 100.000 babies would be left out, if parents send their forms back to the Bureau right away.

But more adults would be left out if the forms had to be postmarked April 1 and later, because if you get the form in Mid-March you can easily forget to send it back 2 weeks later.

So, it's more or less a no-win situation ...

I found the 2000 census forms, but didn't find a separate set of instructions.  It could be that they decided that people don't read instructions, and tried to include everything on the form itself.

The forms all say April 1 (eg how many people were living at this location on April 1, 2000; rather than how many people will be living at this location on April 1, 2000), but that could be simply to make the wording more direct.

I looked through the press releases from 2000.  And there was one on March 27, that said that they had a 42% return rate by that date.  It also suggested that those who had not already returned their form might set aside "Census Day" to complet the form.

The USA releases the census forms 70 years after the fact (these are available for genealogical research).  In times past, the census was taken my enumerators, who went door to door.  Those were started on the census date, and presumably back-dated (eg my grandparents lived on a farm that was one of the last enumerated in 1920, which was a couple of months after the census date).

[jimrtex]

I just sent a query to the Census Bureau.  I'll let you know what they say.

I think the most common source of double counting is college students who are counted at college, and also reported at their parent's home.