Revisions to the Claimant Count from the inclusion of Full Universal Credit Service claimants

The Claimant Count is a measure of the number of people who are claiming benefit principally for the reason of being unemployed. Between October 1996 and April 2013, it was measured using the number of people claiming Jobseeker’s Allowance (JSA).

In April 2013, the Department for Work and Pensions (DWP) introduced Universal Credit (UC) to replace a number of means-tested benefits, including income-based JSA. Since then, we have been working with DWP to produce figures incorporating unemployed claimants of UC with those people claiming JSA to produce a Claimant Count reflecting both benefits.

The initial Pathfinder roll-out of UC only covered a limited subset of claimants at a small number of Jobcentre Plus offices. Since then, the coverage of the benefit has expanded both geographically and in the type of claimants receiving the benefit. Although UC is now available at all Jobcentre Plus offices in Great Britain, there are still limits on the types of claimant receiving the benefit, with different coverage in different parts of the country. The roll-out to full coverage is expected to extend beyond 2020.

At first there was no information available on the number of UC claimants. Later, a data series became available that included all claimants of UC, rather than just unemployed claimants. Although these methods initially gave an undercount, then an overcount, the size of UC rollout at the time meant the overall impact on the Claimant Count was small.

The next development was a combined JSA and UC estimate, which included those UC claimants who were identified by DWP as “out of work”. DWP’s method for identifying those who were out of work was based on information on whether claimants had any earnings over a 4-week period, including the count date. Those without earnings were identified as out of work. Having no earnings did not guarantee that the out of work claimant was claiming for unemployment reasons in the particular month contributing towards an overcount. At the same time, this method would identify any claimant who had earnings over a 4-week period including the count date as in work, rather than whether or not they were in work on the count date, contributing towards an undercount.

From the labour market releases on 17 August 2016, DWP were able to provide information from the Live Service, identifying those UC claimants who were required to be available for and seeking work on the relevant count date. The information became available for periods from April 2015. Although this led to a discontinuity, due to the improvement in the definition of UC claimants included within the Claimant Count, between March and April 2015, the size of the discontinuity was small. The revisions to the series resulting from this improvement were upward. This is because the number of claimants that were being wrongly considered as in work on the count date, due to earnings during the month, exceeded the number of out of work claimants who were not required to be available for and seeking work.

Although the August revisions brought us towards a final definition of the Claimant Count, as was stated alongside those revisions the new revised Claimant Count still only included claims from the DWP Live Service and not from the new Full Universal Credit Service.

From the labour market releases on 16 November 2016, DWP are able to provide information from the Full Universal Credit Service, in addition to the Live Service. This information is available for periods from December 2014, when the Full Universal Credit Service started, to the latest data.

As can be seen from Figures 1 and 2, although revisions stretch back to December 2014, it is only in the most recent periods, since Full Universal Credit Service has started to be rolled out across a larger number of Jobcentre Plus offices, that the impact has grown.

The revisions have not significantly changed the general pattern of the Claimant Count, although the resulting recent increases are stronger than previously published.

Although this latest revision brings us towards a final definition and coverage of the Claimant Count, there is still further work to do before we can consider the new Claimant Count to be a National Statistic. In particular, we need to replace the current seasonal adjustment methods to better reflect the data in the new series. We also need to better understand to what extent the introduction of Universal Credit is leading to people being included within the Claimant Count that would not have been previously counted.

Also, until we have all cases transferred into the Full Universal Credit Service, it is unlikely that we will be able to produce any reliable statistics relating to claimant flows or duration within the Claimant Count.