Feb 232013

It has always been assumed that only 11% of new ESA claimants would meet the criteria of the Support Group, although reality shows otherwise. And this figure has been repeatedly used by DWP in almost all their communications. But where does this 11% come from?

Through various Freedom of Information requests, we have been able to learn that:

“..within the document that has been referred to (ESA RFP.pdf) there was a planning assumption (based on historical/statistical evidence?[1]) made prior to the introduction of ESA for commercial purposes to assist Atos Healthcare in the recruitment of sufficient resources to deliver the medical services contract”

Or in another FOI:

“…the original estimate of the mix of Work Capability Assessment (WCA) outcomes (for initial assessments) was Fit For Work (FFW) (15%); Work Related Activity Group (WRAG) (76.5%) and SG (8.5%). These estimates were based on Incapacity Benefits Personal Capability Assessment (PCA) data and pilot data”

But the response to a new Freedom of Information request lets the cat out of the bag. To the same question, where does this 11% come from, the DWP answers:

“The WCA was designed in consultation with medical experts and representative groups.  It builds on the experience gained from the operation of the Personal Capability Assessment (PCA) which was used for Incapacity Benefit, which was replaced by ESA. To estimate the probable assessment results the figures derived in the ESA RFP were based on historical Incapacity Benefit results of Personal Capability Reports information.

In year 2008/09 total ESA New Claim referrals were 151,523 – Support Group criteria 16,274 (10.74%); the percentage was constant throughout.  Therefore, DWP use the derivation of 11% from these findings”.

Below is the table from which these figures are taken:


ESA New and Repeat claims -referral at Week 4

ESA New and Repeat claims -referral at Week 4


So the DWP tells us that this 11% is derived from the 151,523 persons assessed in 2008/2009, and that this has been confirmed by figures obtained in the following years. Really?

In fact, this number is part of what is called a planning assumption, which does not represent the actual number of people assessed during the year, but what DWP thought was a workable figure for budgeting costs in its ESA contract with Atos. The document, which contains these figures and other figures up to 2015, was completed in 2007 and the little note underneath the box referring to the drop out rate gives it away.   So while DWP presented this 11% in the last FOI as the findings derived from actual figures, it was in fact a percentage derived from a planning assumption, or to put it differently, in this FOI the DWP was confirming an assumption with another assumption.

But let us look more closely at the figures, because there it becomes really interesting. It is complicated, but worth reading until the end. We need to disregard the figures in the 2008-2009 row, because they are atypical and do not tally as it was the first year of ESA roll-out.

For all the following years, the number of people in the Support Group column also amounts to 11% of the total number of referrals, exactly as DWP stated, but we need to remind ourselves that these figures were only assumptions. Now let us do some calculations. If we take the total number of ESA claimant referrals, then subtract the number found in the Support Group column, and then subtract the number found in the Drop out Pre-exam column, we consistently arrive at the number of people found in the Expected at Examination column[2].

What does it mean? It means that the assumption of 11% does not represent, either in this table or in any other DWP documents, the total number of people in the Support Group, but only the number of people, meeting certain criteria, who would be put directly into the Support Group without going through a WCA.  Of the number of people undergoing a WCA in the Expected at Examination column, a certain number would also be assessed as meeting the criteria of the Support Group. And do not forget that these figures represent the number of people being assessed, before the decision of the Decision-Maker, and before any appeals, and the number of appeals would have impacted this group more than it would have impacted the total number of referrals, because by definition it contains all the people likely to appeal a decision.

Or to put it differently, 11% is the basic percentage to which should be added the percentage of people going through a WCA and meeting the Support Group criteria.

These assumptions were given to Atos for costing and planning purposes only. The 11% was never intended to represent the actual number of people in the Support Group, but only those who did not need a medical examination, so that Atos could plan and budget staff, premises and other resources around the number of physical WCAs it would actually have to do.
The DWP people who compiled these tables would have known the 11% represented only the number of people who would have to be deducted from the number of referrals in order to work out how many claimants will need to attend a physical examination, but they seem to have been the only ones to know this.
It is obvious from the last FOI request that DWP staff do not understand what these figures refer to, as no explanation accompanies the table, and in the contract, these figures are contained in a stand-alone document. It is also obvious that the DWP staff who wrote the contract for the introduction of ESA with Atos at the same time must have been different from the staff who compiled this table as they did not understand what these figures were supposed to mean either, or they would not have written:

‘It is estimated that approximately 11% of new customers will fall into the Support Group. Atos Healthcare MUST base their solution and costs the figures in Appendix 8’,

when their own contemporary contract document shows that 11% only represents the fast-tracked claimants.  It could not have been in DWP’s interest to deceive Atos on such a scale.

 To understand what happens, look at this graphic:


Atos would not have had any reason to doubt this figure, around which it would have planned and incurred expenses. Anything deviating from this assumption would have cost Atos money, and would therefore have created the financial pressure to keep the Support Group percentage as close as possible to 11%.  Moreover, although 11% does not constitute a target as such, the financial pressure coupled with DWP’s estimations based on wrong assumptions would have necessarily driven the WCA’s outcomes.


Apart from the DWP’s real desire to cut the number of benefit claimants, this is a glaring example and evidence of confusion and incompetence at several levels of the DWP, which goes right to the top of the department. Not only are DWP staff unable to differentiate between assumptions and statistics, but they are also wrongly interpreting their own figures, and this incompetence would have led Atos to attempt to mitigate any unplanned costs it incurred because of this blunder by cutting the number of people meeting the Support Group criteria. That would have therefore influenced the assessment process and the outcomes of the WCA.


It also explains why neither Grayling, IDS nor any other DWP spokesperson can afford to reveal the real numbers of people meeting the Support Group criteria, which have been grossly under-estimated, without losing face, as they show their incompetence.  They would rather be seen as nasty than incompetent. But it is incompetence that disabled people are paying dearly for.


[1] This was written in red in the original document, with the question mark

[2] I have used the Expected at Examination column rather than the Completed PCA Examinations for simplicity purposes (the number of people non attending is 17%). It does not affect the calculations in any significant way, and does not impact on the argument