AHS 2015 Geographic Coverage (was RE: Time is Running Out to Make Your Voice Heard)

Dear All:

I think Patty's comments are right on. I don't think people are aware of the issues surrounding group quarters (e.g. dorms, prisons) on both the decennial and the ACS. Take a look at this NAS report. http://www.nap.edu/openbook.php?record_id=11727

Primary residence or usual residence is gone from the ACS.

I don't think the AHS should be morphed. I have used it in court cases with the metro area data. It has many things that are neither on the ACS nor the Census.

But I would like a second house question. Remember the unit in the AHS is the housing unit and not the household.

Andrew A. Beveridge
andrew.beveridge@qc.cuny.edu

I have been on vacation for the past week, and thus only dimly aware of this discussion. I have now read through the discussion in the various emails a bit more closely.

What is fascinating here is the desire to make the AHS something it's not. The AHS is designed to describe traditional housing in the U.S. By traditional housing, I mean single family detached and attached, small multiple buildings (2-4 or 6 or 8 units), apartment buildings, and mobile homes. It is explicitly not designed to cover or include group quarters housing or to define housing which comes in and out of the stock, such as permanent resident hotel rooms.

Yes, it would be great to have a clear understanding of all the varieties of student housing, but it would take a study explicitly designed to do that to come to such an understanding. Post-secondary students live in such a variety of housing--university owned and private, group quarters and regular apartments/homes, part of traditional households and not, with their parental families, their own nuclear families, their peers, and on their own, that it's impossible to cover all the bases in a survey such as the AHS. ACS can't do it, either. The decennial census tries to find them all, wherever they live, but can't precisely describe all the variety of places that house them and definitely can't give us socio-economic characteristics along with the count.

The AHS also cannot provide great geographic detail. I just doesn't have enough funding to collect enough cases. Its original concept was to do two things at once: to provide a national sample with little or no geographic detail, and to provide data for a select number of metro areas and their central cities. When originally designed, it was fair to assume that the decennial census, with its long form, would provide the socio-economic data for specific categories of housing, especially through use of PUMS files which provide geographic detail down to the 100,000 population level. Now the ACS PUMS files have to do that.

My experience in Detroit is that the metro data are greatly underutilized. The great value of AHS has been in the details about housing which are not available elsewhere, such as condition and respondent attitudes toward and evaluation of their own housing. I think that should and will continue. There is also great value in the longitudinal aspect of the survey, if done properly. Let's not expect the AHS to do what the decennial and ACS are better equipped to do.

I think there should be a fight to get structure type on the decennial, i.e., on the short form. It was taken off because it wasn't/still isn't needed for measuring undercount, and that was the rationale for the short list of short form items. I think it's needed because it's a critical socio-economic variable in its own right and because it ties the geographic identification of the data to the ground at the block level.

Patty Becker

Patricia C. (Patty) Becker
APB Associates/Southeast Michigan Census Council (SEMCC)
28300 Franklin Rd, Southfield, MI 48034
office: 248-354-6520
home:248-355-2428
pbecker@umich.edu