One very tricky aspect of the Census definition of dwelling unit shows up with respect to units that are under construction. The Census first considers a unit under construction to be a vacant and unoccupied dwelling when it reaches the stage known in the building world as "framed to lock-up." That is, a dwelling that has its exterior sheathing and roof sheathing installed, and windows and doors in place so that it can be locked up and secured. It does not yet have insulation, wiring, sheetrock, plumbing fixtures, etc. installed, and would not yet be considered a habitable dwelling by any local government. Rather, we would still consider it be a "construction site," not yet approved for occupancy. This can play havoc with your baseline figure for number of dwellings in the decennial Census year, especially in those newly-developing neighborhoods (i.e., tracts or block groups) that have significant projects under construction (e.g., an apartment building with 100 units). You may have to spend a lot of effort adjusting the Census data to correspond with building records for permits taken out versus final inspections/occupancy permits, in order to get a true picture.
- Pete Swensson
(now retired, but not expired, formerly of Thurston Regional Planning Council, Olympia, WA)
From: "Robert. Allen" <Robert.Allen@abilenetx.com>
To: ctpp-news@chrispy.net
Sent: Tuesday, April 26, 2011 12:14:32 PM
Subject: RE: [CTPP] Census Vacant Structure Determination
My experience over three censuses reflects Patty’s experience. I, too, have the impression from a partial review of the redistricting data that more apparently uninhabitable units are being left on, but that is better than the last census in which we had a least one case in which a clearly occupied unit was listed as nonresidential. Hastily trained temporary Census workers have authority to make the decisions for Census purposes and all the documentation that a local professional may have on hand will not alter the record in my experience. I have even had an entire section of an important, and quite visible, boundary creek removed from Census maps by Census field workers for two censuses running. Such things give me no confidence in the precision of their records but the census is still the most official data and as good as any other source as far as I can tell.
Robert R. Allen, AICP
Abilene MPO
400 Oak St., #102
Abilene, TX, 79602
Ph. (325) 676-6243
On Behalf Of Patty Becker
Sent: Tuesday, April 26, 2011 11:00 AM
The problem is that vacancy status is determined, in the end game, by the enumerator who goes out there and looks at the building/unit. Some units that are really occupied will be declared vacant because the occupants, who may be squatters or illegal immigrants, are hiding from the census. Sometimes there's a fine line between a vacant unit and one which should be taken out of the stock as unlivable. I have the general impression that in 2010 the tendency here was toward vacant, whereas in 2000 and 1990 the tendency was toward removing it from the stock. I doubt that there's any way to prove this one way or the other.
Adam - does your ordinance also cover vacant apartments in multi-unit buildings?
Bottom line - the census is not clearly definitive on this point. Its mission is to count people, not vacant units.
Patty Becker