Producing working-life histories in the BHPS and UKHLS 2017-2020

DOI

The British Household Panel Survey (BHPS) and United Kingdom Household Longitudinal Study (UKHLS) contain many questions related to working-life histories. Responses to these questions are not always consistent and cleaning the data is a large task. This collection contains Stata code to produce a sequence of main activities - from birth to last interview (where possible) - for adult participants in the two surveys. The collection also contains a report which explains the logic behind the code and describes the main working-life history questions from the two surveys (including questionnaire routing) in detail. Stata syntax for producing working-life histories (a sequence of main economic activity, with start and end dates) for participants in the British Household Panel Survey (BHPS) and United Kingdom Household Longitudinal Study (UKHLS).

The deposit contains Stata code and no data. BHPS and UKHLS data were collected by the Institute for Social and Economic Research at the University of Essex. I have no affiliation with that organisation. The BHPS and UKHLS contain three modules related to working-life histories: an annual activity history, an annual education history, and an intermittent lifetime activity history. The modules overlap in the period questions cover. Further, date information is missing for some participants. Decisions must be made on rules for combining data from separate modules and how to handle missing date data. The logic behind the code is as follows. Missing dates are clean for each module separately by wave. When the range of possible dates is sufficiently small (default six months), the date is set to the middle of the range. If not sufficiently small, the dates are set at their maximum known range (this leaves gap). Next, sequences of activities from a single module are collapsed across survey waves. To minimise recall bias, precedence is given to responses from earlier waves. Finally, modules are combined. Where there is overlap, precedence is given to the annual education history over the annual activity history, and both are given precedence over the lifetime activity history. A spell of full-time education from birth to first leaving full-time education (or last date in education, if never left) is also added to this combined data, which draws on information from other survey questions. After running the code, the combined working-life histories are saved in a file. The file contains the type of activity, its start and end dates, along reason for taking up the activity and reason for ending it, if this data is available. Cleaned data from each module is also saved separately, in case the user is interested in just one data collection method.

Identifier
DOI https://doi.org/10.5255/UKDA-SN-854327
Metadata Access https://datacatalogue.cessda.eu/oai-pmh/v0/oai?verb=GetRecord&metadataPrefix=oai_ddi25&identifier=3946b043471d02bab01cf3ba434cfd9d013745d434088d239686826ac84b3c93
Provenance
Creator Wright, L, University College London
Publisher UK Data Service
Publication Year 2020
Funding Reference Economic and Social Research Council
Rights Liam Wright, University College London; The Data Collection is available from an external repository. Access is available via Related Resources.
OpenAccess true
Representation
Language English
Resource Type Software
Discipline Economics; History; Humanities; Social and Behavioural Sciences
Spatial Coverage United Kingdom