Project

General

Profile

Support #455

Residential mobility indicators

Added by Rory Coulter over 8 years ago. Updated over 8 years ago.

Status:
Closed
Priority:
Normal
Assignee:
-
Category:
Derived variables
Start date:
11/27/2015
% Done:

100%


Description

Hello,

I have the latest version (W5 release) of SN6614 and SN7248 and have a question about residential mobility indicators. In a way this builds on query 257 I raised last year.

I've attached 2011 census LSOA codes to UKHLS and have derived an indicator measuring whether the respondent changed LSOA since the previous wave. For each wave this yields roughly a 5-6% residential mobility rate. Although the relationships between other variables and this mobility indicator make sense, this rate seems very low so I'm trying to see whether there are a lot of very short moves within LSOAs I'm missing. These moves may also be important in other ways too.

In past releases of UKHLS, at least up until W3, there were at least two ways to identify moves in SN6614: (a) use the plnow variables based on interview responses or (b) use the xwhist variable on xwavedat to examine whether respondents have changed postcode sector since last wave.

Both of these have problems so I like to compare both methods but in the latest UKHLS I can't find the xwhist indicator. Is this no longer being provided or have I just overlooked it? If xwhist is no longer available is there an alternative way to measure changes in address that comes from the administrative records of addresses rather than the interview itself? BHPS had a useful movest variable to do this and triangulation with the plnew variable helped reconcile inconsistencies.

Any advice would be much appreciated. Obviously it would be possible to use the secure access datasets but I'd rather avoid that if possible.

Many thanks,
Rory

Also available in: Atom PDF