Census Bureau wants to halt immigration raids during 2010 Census

[jimrtex]

On a related note, Libertarians and Republicans better not start whining that the Census invades their privacy like they did in 2000. I agree that some of the questions are not needed, the ones about plumbing, etc. on the long form. But to not fill out the short census form is, for lack of a better word, unpatriotic. It is part of the Constitution after all.

The long form won't be used for the regular census, having been replaced by the ACS.

[jimrtex]

On a related note, Libertarians and Republicans better not start whining that the Census invades their privacy like they did in 2000. I agree that some of the questions are not needed, the ones about plumbing, etc. on the long form. But to not fill out the short census form is, for lack of a better word, unpatriotic. It is part of the Constitution after all.

The long form won't be used for the regular census, having been replaced by the ACS.

Well, it looks like the complainers won that round. The ACS is very suspect in my opinion. The 2005 one shows lots of places declining from 2000 but only because it is very limited and leaves out large groups of people.

When they have enough data, it should be as statistically valid as the results from the long form (other than not being tied to a single date), and for larger populations on an annual basis.

The data that has been released so far excluded persons living in group quarters, so areas with a lot of college dormitories are undercounted.  Note that group quarters were excluded from the long form in the 2000 Census.

[jimrtex]

WASHINGTON - The Census Bureau wants immigration agents to suspend enforcement raids during the 2010 census so the government can better count illegal immigrants.

Census Issues Clarification on 2010 Census and Law Enforcement

“The Census Bureau has not requested that the Immigration and Customs Enforcement (ICE) agency refrain from enforcing immigration laws. While previous Administrations sought to have law enforcement raids curtailed somewhat to help obtain greater accuracy, we respect ICE's statement that they will not suspend raids even if a decision were made to ask them to do so. The Census Bureau fully recognizes that times have changed, with new challenges facing immigration authorities, and Census will change with those times.”

[jimrtex]

On a related note, Libertarians and Republicans better not start whining that the Census invades their privacy like they did in 2000. I agree that some of the questions are not needed, the ones about plumbing, etc. on the long form. But to not fill out the short census form is, for lack of a better word, unpatriotic. It is part of the Constitution after all.

The long form won't be used for the regular census, having been replaced by the ACS.

Well, it looks like the complainers won that round. The ACS is very suspect in my opinion. The 2005 one shows lots of places declining from 2000 but only because it is very limited and leaves out large groups of people.

When they have enough data, it should be as statistically valid as the results from the long form (other than not being tied to a single date), and for larger populations on an annual basis.

The data that has been released so far excluded persons living in group quarters, so areas with a lot of college dormitories are undercounted.  Note that group quarters were excluded from the long form in the 2000 Census.

And what is their justification for excluding group quarters? It's much easier to call up a single jail or college and find out how many people live there, than it is to get the public at large to participate. If I remember from 2000, only 67% of Americans were capable of returning their census forms.

In 2000, 67% responded.  This says nothing about the capabilities of the non-respondents.

The purpose of the ACS is to gather information equivalent to the long form on a continuing basis.  Each year, about 1/40 of households are surveyed,   Over a five-year running period this represents about the same number of persons as received the long form.  The 1/8 sample is considered to be statistically valid for most census tracts.    A one year sample is valid for areas with population of 65,000 greater.

For Group Quarters (GQ), sampling is also used.   GQ believed to have less than 15 residents are sampled just like Housing Units (HU).  Larger GQ are sub-sampled into groups of 10 residents, with 1/40 of these groups interviewed each month.  For a GQ with 400 residents, 10 would be interviewed each month.  The operator of the GQ provides a list of the residents to the census bureau, then the census bureau selects those to be interviewed.  The interview is done in person, while for ordinary households, the initial contact is by mail.

During the early part of the decade, only tests of the ACS were run.  2005 was the first year for full nation-wide implementation for housing units.  My guess is that the Census Bureau decided it was better to get the full sample for GQ rather than the more specialize GQ population (which is 2.8% of the population according to the 2000 Census).

2.0 m in correctional institutions.
1.7 m in nursing home.
0.4 m other institutions.

2.0 m dormitories
0.4 m barracks
1.3 m other non-institutional