Enriched ASHE Quick Start Guide
This document describes the ‘Enriched ASHE’ dataset created by the code, which can be applied to the standard ASHE dataset. The code can also be applied to the ASHE-2011 Census dataset, creating an ‘Enriched ASHE – 2011 Census’ dataset.
QA of Main Job Marker (mjob)
The variable ‘mjob’ is supposed to indicate whether this job is the ‘main job’, if a person holds more than one job. It appears that this definition is based upon the job with the largest number of hours worked. However, there are a substantial number of multiple job holders where this is not the case. In addition, the definition of ‘mjob’ is not helpful (if you have only one job, ‘mjob’ is set to ‘false’). We have added new variables to the dataset, with main job definitions based on different hours/earnings measure. We have also created a new ‘mainjob_ons’ which is set to 1 if the only job, or if multiple jobs and the ONS ‘mjob’ variable is set to 1.
QA of sjd (‘same job’ marker) and sernol (serial number in previous year)
The variable ‘sjd’ reports whether a person has been in the same job (note, not the ‘same employer’) for 12 months or more. This is a survey question, and valuable for research into within-work progression. If marked ‘yes’, the understanding (in the 2000s) was that information from the previous year was copied over, but this is not currently the ONS understanding. How sjd is used and how it relates to the independence of other variables therefore has the potential to affect analysis. In the WED drop, a variable ‘sernol’ is included, indicating the reference number of the same job in the previous year. This is highly valuable, as it allows reliable longitudinal jobs to be constructed. It is not fully understood yet how this variable is constructed, or how well it represents job continuity.
QA of work location (wpost) problem (one page summary)
ASHE records an employee’s home and workplace location. This is of great potential value in assessing the role of geographical location in labour market analysis and in estimating geographical impacts of policy. There is concern about the accuracy of the work location variable as the underlying survey question is pre-filled by ONS with the registered PAYE address of an employer. We tested the hypothesis that there is measurement error in the ASHE workplace location variable, leading to possible errors in employee work location for multi-site organisations. We make a number of recommendations for short-term measures that can be used to reduce the potential for unrepresentative descriptive statistics or model estimates arising because of workplace measurement errors.
QA of work location (wpost) problem (detailed version)
This paper includes a detailed breakdown of our analysis of the workplace location problem, as well as a suggested strategy and recommendations for addressing the issue.
Longitudinal Attrition in ASHE (Methodology Paper)
The longitudinal integrity of the ASHE sample has been the subject of little prior investigation, with the panel sample generally assumed free of any attrition biases that might compromise longitudinal analysis. Our analysis confirms the existence of systematic patterns of longitudinal attrition in ASHE, which have the potential to introduce bias into longitudinal analyses of these data. We have constructed longitudinal weights that correct for estimated attrition biases over adjacent years in ASHE.