Mapping SOC2000 and SOC1990 to RGSC
I would like to recode the yasoc00 and ypsoc00 variables, which use SOC2000 from UKHLS to RGSC categories, and I would also like to do this in BHPS youth panel, which uses SOC1990. In the guidance, it says that for the current job variable, jobsoc, this is translated to RGSC using the Camsin linking file. I have downloaded the linking file but I can't see exactly how this translates. It has code mapping the four-digit SOC2000 to estimated RGSC codes, but many of these have missing RGSC codes, or have numerous different RGSC codes within the same 4-digit SOC2000 code.
Firstly, Do you have the mapping files that you used in the survey to map from SOC1990 and SOC2000 to RGSC codes?
Secondly, in addition, up to Wave 2, the main job is coded using SOC1990 as well as using SOC2000, but this variable drops out after Wave 2. Do you have the mapping file so that I can map the main job to SOC90 for Wave 3 onwards? I want it to be consistent with BHPS.
Thanks very much
Updated by Megan Scott over 5 years ago
That should say CAMSIS linking file, which is found here: http://www.cardiff.ac.uk/socsi/CAMSIS/occunits/distribution.html
Updated by Gundi Knies over 5 years ago
- Category set to Data analysis
- Status changed from New to Feedback
- Assignee set to Megan Scott
- Target version set to X M
- % Done changed from 0 to 90
sorry for the delay in response. All our code to derive socio-economic classifications relies on SOC 2000. The look-up file we use for the current job RGSC classification (_JBRGSC_DV) relies on occupation (SOC2000, _JBSOC00) and employment status (ES2000, _JBES2000). In the online variable documentation you can see that the latter depends on _JBSEMP _JBMNGR _JBSIZE _JSSIZE _JSBOSS. Comparable information is not available in the youth questionnaire (note that this is the occupation young people would like to do, not an occupation they are already working in!).
There may be ways to generate ES2000 by making assumptions about what the typical employment status for a given occupation is - fyi: ONS provides documentation on how this can be done in the context of deriving NS-SEC (and we use this cross-walk to derive parent's SOC, see, e.g., _PANSSEC_DV).
On 'translating' SOC 1990 codes to SOC 2000 codes, I am afraid we have not got a solution for you. We code occupations from the freetext information we collect during the interview, shortly after fieldwork is completed. During the first waves of the BHPS, SOC 2000 did not exist as a code frame so it was not possible to code the freetext to that level. Subject to additional resourcing it may be possible to try and code this to SOC 2000/2010 in the future. But it is quite possible that the freetext information is not detailed enough to assign a SOC code, e.g., when the occupation split into different specialisations. Again, ONS may have some advise on whether and how the missing values may be imputed.
Sorry to not be of more help,