Support #571

routing for a_jbhad and b_jbbgd/b_jbbgm/b_jbbgy

Added by Zoe Oldfield about 4 years ago. Updated about 4 years ago.

Data documentation
Target version:
Start date:
Due date:
% Done:


Estimated time:



I am trying to work out the routing for some of the employment variables and for some variables, the routing in the questionnaire doesn't seem to match the routing in the data.

1. In wave 1, the routing stated in the questionnaire for a_jbhad is:


However, when I do the following (stata code):

gen route = a_jbhas==2 & a_jboff==2
tab a_jbhad route

I get the following:

ever had paid |         route
employment | 0 1 | Total
missing | 0 29 | 29
inapplicable | 26,525 1,190 | 27,715
proxy respondent | 2,036 1,226 | 3,262
refused | 0 2 | 2
don't know | 0 14 | 14
yes | 0 16,108 | 16,108
no | 0 3,864 | 3,864
Total | 28,561 22,433 | 50,994

So there are 1190 observations that look like they should have been asked jbhad but were not. Why is this? Am I missing something in the routing?

2. The routing for b_jbbgd/b_jbbgm/b_jbbgy is the following:

     IF     (JBHAS = 1 OR JBOFF = 1)

However when I do the following (stata code):

gen route = (b_jbhas==1|b_jboff==1) & (b_ff_ivlolw~=1 & b_ff_everint~=1)
tab b_jbbgy route

I get this:

year started |         route
current job  |         0          1 |     Total
missing      |         0          1 |         1
inapplicable |    50,010      1,982 |    51,992
don't know   |         0         16 |        16

So there are 1982 observations who do not get asked the date that their current job started even though they have never been interviewed before. Why is this? Again, am I missing something?



#1 Updated by Gundi Knies about 4 years ago

  • Category set to Data documentation
  • Status changed from New to In Progress
  • Assignee set to Zoe Oldfield
  • Target version set to X M
  • % Done changed from 0 to 50

Hi Zoe,

on your query regarding a_jbhad. In Wave 1 JBHAD had the following routing:
if (JBHAS = 2 and JBOFF = 2)
if ((months 1-6 and if any LESHST is 1-3) or (months > 6))
This is currently not apparent from the questionnaire and we will update this as soon as we can. In the meantime, you can use the information from the employment history file to construct the indicator for whether or not a person has ever had a job.

On you second query, I looked at the questionnaire for Wave 2 and it says that the separate year, month and day variables were only asked
If (JBHAS = 1 OR JBOFF = 1) // Has a job
And If ((ff_ivlolw <> 1 & Ff_everint <> 1)) // New entrant never interviewed
And If (ANY JBBGDat = DK) // Full date started current job not known
It may be that the cases you have pointed out as inconsistent do not fulfil the third criterion. JBBGBDat is not included in the released data so it will take us a little bit of time to investigate this. Please bear with us.


#2 Updated by Gundi Knies about 4 years ago

  • Status changed from In Progress to Feedback
  • % Done changed from 50 to 100

Dear Zoe,
further to my previous reply, I can now confirm that there has been a data processing error in Wave 2. The start date for respondents who knew the exact date, stored in JBBGDat, has not been processed. We will fix this error for the next data release. The error only occurs in Wave 2.
Thank you for making us aware of the issue!
Best wishes,

#3 Updated by Zoe Oldfield about 4 years ago

Hi Gundi,

Thanks for your reply.

Do you know shen is the next data release schedule for?

#4 Updated by Gundi Knies about 4 years ago

Hi Zoe,
the next release is planned for end of November 2016. We publish the latest plans here

Best wishes,

#5 Updated by Victoria Nolan about 4 years ago

  • Status changed from Feedback to Closed
  • Private changed from Yes to No

Also available in: Atom PDF