Skip to main content

Re: 2002 AHS: weighted total numbers of recent movers

HUD.GOV HUDUser.gov
eList
AHS
thanks to David and Gregory for your quick responses.

I followed David's suggestion to identify movers/person who moved within a
year prior to the interview; I got new numbers showing that things are
more complicated than I have thought. This is what I found:

1. All movers in the mover module are indeed persons who moved within ONE
year before the interview date. (this is consistent with the description
in the Code book).

2. The mover module contains all persons who are identified as recent
movers, who could be a householder, a reference person, or any household
member. It is very straight forward to identify a recent mover/person, but
things get tricky when it comes to households. You can get quite different
numbers depending on how you define a household as a recently moved one.
(a) if a household is defined as a recent mover/household when the
reference person is a recent mover regardless of whether the reference
person is a householder, you get the same total estimate numbers as those
published on the AHS web site.

(b) if a household is defined as a recent mover/household as long as at
least one household member is identified as a recent mover, you get total
estimates quite larger than those published numbers. (what I got the
first time).

(c) if a household is defined as a recent mover/household only if the
householder moved within the past year, you get total estimates slightly
smaller than the published ones.

Yizhao

>
>>>I tried to estimate the total numbers of recently moved households in
>>> the
> 13 MSAs, using the data in the mover module and aggregating the data to
> the
> household level. But the numbers (weighted total)are different from those
> in the AHS data charts posted on the AHS website (
> http://www.census.gov/hhes/www/housing/ahs/metropolitandata.html ). Mine
> are quite larger than those on the web. For example, the web shows 163.2
> recent respondents in Anaheim-Santa Ana MSA; mine is 245.6.<<
>
> The published table shows households that have moved in the past year. We
> collect data on households that have moved in the past two years. Thus,
> the microdata identifies more recent movers than the report.
>
> From the table spec file for 2002 (available from our web site), recent
> movers in the past year are identified by:
>
> IF (DATEY eq MOVE or (DATEY - MOVE eq 1 and MOVEM ge DATEM)) then MOVEDLY
> 'Y';
>
> DATEY and DATEM are the year and month of interview, which you can get by
> parsing DATE.
>
> Dav Vandenbroucke
> Economist
> U.S. Dept. HUD
> david_a._vandenbroucke@hud.gov
> 202-708-1060 ext. 5890