Reinstatement of APS disability estimates

In November 2017, the Office for National Statistics (ONS) suspended the publication of Labour Force Survey (LFS) tables relating to disability due to an apparent discontinuity between estimates for April to June 2017 and July to September 2017. The survey estimates showed a step increase of around 400,000 in the number of people reporting themselves as disabled. Publication of disability related estimates from the Annual Population Survey (APS) on Nomis straddling these periods were also suspended at this time.

Since then ONS has been investigating this apparent discontinuity, along with stakeholders such as the Department for Work and Pensions.

These investigations have proved inconclusive with no change to the survey instrument or immediate external factor being attributable to the sudden increase.

Following these inconclusive investigations, ONS has decided to reinstate the LFS disability tables from 15 May 2018, however they are advising caution in making comparisons between April to June 2017 and subsequent quarters. Similarly, the APS estimates on Nomis are being reinstated, with the advice that estimates for datasets straddling the apparent discontinuity should be used with caution.

ONS have produced an article with details of the issue identified and the investigations that have been carried out.

ONS intend continuing investigations and reporting again in November, with further recommendations regarding the use of the data.

Update 13/11/2018:

In November 2018, ONS concluded that there is no evidence to suggest that the large increase in the number of people reporting disabilities between April to June 2017 and July to September 2017 is caused by a discontinuity in the data. They have therefore removed the "health warnings" that were previously applied to the dataset.

A further article on the ONS website explains the quality assurance investigations they conducted.