Guidance Notes On Differences Between The Annual Population Survey

guidance notes on differences between the annual population survey (aps) and business register employment survey (bres). this note has been

Guidance notes on differences between the Annual Population Survey
(APS) and Business Register Employment Survey (BRES).
This note has been produced to highlight differences between outputs
from these two ONS surveys. This is required as the BRES survey does
not report gender (which was reported in its predecessor, the Annual
Business Inquiry). Consequently, the APS needs to be used as an
interim solution to obtain information on employment by industry and
gender from 2009. The following points highlight the differences that
exist between the two datasets that the user should be aware when
using the ASP in place of BRES / ABI for this analysis.
APS data is availble on Nomis at www.nomisweb.co.uk
1 – BRES / ABI data reports employee jobs whereas APS reports people
in employment (people can hold multiple jobs, individual jobs can be
shared, so these are not like for like comparisons).
2 – BRES / ABI data relate to all persons of age 16 and over. Age is a
selectable variable within APS, so you need to ensure you are choosing
the relevant employment rate required (16+ and 16-64 are the two most
commonly used). The output from Nomis will include the level as well
as the rate.
3 - APS data will also contain the following groups which are not
contained in BRES/ABI
- HM Forces
- Farm agriculture (not included below government office region
geographies)
- Self employed
- Government sponsored trainees
These will have different impacts across different local authorities.
4 - BRES/ABI data is workplace based, the APS tables available on
Nomis for industry by gender are residence based. The workplace based
APS data on Nomis does not contain gender data.
5 - The APS datasets are not seasonally adjusted, therefore when
making comparisons over time, always choose the same time spans, e.g
compare Jan-Dec 2009 data with Jan - Dec 2008 data (on Nomis this
would be clicking the Oct Nov Dec 2009 and Oct Nov Dec 2008 boxes).
6 - For the reasons stated above, estimates from ABI/BRES are not
directly comparable with estimates from APS. Therefore only make
comparisons between data from the same dataset when comparing
estimates to those from previous years.