Monday, 13 November 2017

My submission to the Parliamentary Inquiry on ESA and PIP

Response - PIP and ESA Assessments inquiry

Name: Tom Evans
Welfare rights adviser 2008-2009; 2013-present


I have been working for an advice charity advising clients on benefits matters and entitlement full time since 2013 with a focus on representing clients through the process of appealing work-capability and PIP decisions. I also had some experience of this process in 2008/09.

Common features of ESA and PIP

Entitlement to both benefits is determined by assessing a claimant’s disabilities against a set of descriptors which prescribe specific problems. These descriptors cover 17 activities in ESA and 12 in PIP.

A claimant who is assessed as meeting or satisfying a descriptor scores a corresponding number of points. In most cases, points will be needed across more than one activity to qualify for benefit.


In both benefits, the assessment is carried out by requesting the claimant’s own account of their relevant problems in a questionnaire, and then by asking about this further at a face to face meeting with a qualified healthcare professional (known as an HCP or assessor). The assessor then recommends to the DWP which points-scoring descriptors are satisfied. This is based on their training by the DWP on the criteria, and to an extent (which has never been fully disclosed by the DWP) by purportedly expert-system software. In ESA an application known as LiMA[1]. is involved, and there is an unknown equivalent in PIP.

Common problems in the assessment process of both benefits

Points Criteria

It follows from the above that a correct understanding by the assessor (and to the extent relevant, by the computer programme) of the criteria for satisfying each descriptor is absolutely fundamental to the process.

One of the most common complaints by claimants is that an assessor was not qualified to assess them (e.g. due to not being a doctor or specialist in the condition). While it may be a factor, this probably misses the point in most cases. Assessments by doctors are, in our experience, of no higher quality.

The issue in a given assessment is far more likely to be poor application of the descriptors to a claimant’s circumstances, even where these factual circumstances are agreed. The extent to which this is caused by an incorrect but innocent belief about the test, or by deliberately incorrect training or working to targets, remains extremely contentious.

Issues around anchoring to the top descriptor, and bottom-up counting

Within the 17 activities in ESA and 12 in PIP, each activity carries multiple potential point-scoring descriptors. The scores available range from 6-15 in ESA and 2-8 in PIP.

In both benefits, it is very unusual for claimants to score more than the lowest descriptor in an activity. Assessors seem very reluctant to award higher scoring descriptors, which can make qualifying for the benefit a much more difficult exercise, particularly for people whose needs are severe but focused in one area. This has been referred to as bottom-up counting: reversing the order in which descriptors are considered so that only the lowest scoring one is counted, even if a higher score might also be justified.

Conversely assessors also seem anchored to the concept (see ‘rules of thumb’) that to score any points a claimant must be completely unable to carry out an activity (which generally relates to the highest descriptor). This leads to a double harshening of the test so that to score any points a claimant must factually satisfy the highest descriptor while only being allocated the points for the lowest.

ESA Example: Mrs W is losing control of her bowel on a daily basis. Instead of the appropriate 15-point descriptor, she was only awarded the lower descriptor for being ‘at risk’ of this happening.

PIP Examples: -

Mr C has autism but was awarded the lowest scoring two-point descriptor for social engagement: that he needs encouragement. The activity is supposed to encompass understanding body language and forming relationships reliably, which he is unable to do.

Many claimants who are physically unable to complete an activity at all are only awarded two points for ‘needing an aid or appliance’, where this does not seem relevant to their problem, and it is not explained how one would help.

Weight placed on purported counter-examples

Much of the assessment process for disability benefits relies on looking for counter-examples: activities a claimant can manage which would be inconsistent with a points score. Explored correctly, this is a valid technique: for example the fact a person who has recently moved house themselves without difficulty or receiving any help would be good evidence they could lift a cardboard box.

Use of this technique is problematic in ESA and PIP assessments, where alongside the ‘rules of thumb’, very dubious examples are often given disproportionate evidential weight. Often the examples are activities managed very occasionally, or in other cases, are examples lacking proper elements of the test in the descriptors. The LiMA software in ESA works particularly heavily along such lines[2].

Examples: -

Mr M was found able to cope with unexpected changes, and travel to unfamiliar places alone, based on him being able to sit in his car at the seafront near his home.

Miss P was said to be able to plan and organise tasks based on being able to drink water.

Mr R was said to be able to walk 200 metres without stopping based on getting to the foodbank, an unknown distance away.

Miss F was found to be able to cope with social engagement reliably due to attending rape counselling.

Many of our clients have been found to be able to cope with social engagement reliably based on going into a shop or to the GP, without any more detail about these encounters being asked.

The LiMA software in ESA is known to search for keywords in a claimant’s alleged typical day and to carry these examples over to its advice on activities, where it is used to affect the recommended score. This can be observed when a nonsensical example is occasionally picked up by the computer: for example, a statement that a claimant ‘sees his children’ might be carried over to the section about vision. Other examples of known keyword recognition include any statement including the word ‘shop’ being carried into social engagement, and any use of ‘dog’ being carried into mobility.

Irrelevant and pseudo-scientific mental health tests

In addition to often dubious counter-examples, ESA and PIP assessments both include tests such as asking the claimant to count down from 100, spell ‘world’ backwards, and remember three objects. While these have some use as tests for dementia and severe learning disability, they have no known validity in testing for mental health or emotional disorders, or autism-spectrum disorders.

Despite this, these tests are used routinely as purported evidence that a claimant can function socially, reliably plan and organise tasks at home or in a workplace, or cope with stresses.

Failure to refer to the claimants’ own evidence and to other evidence

In principle the DWP’s decision should take into account the claimant’s own evidence, either in the claim forms or other statements, and other relevant evidence, as should the assessor’s report.

In practice we do not find that this happens. The assessor’s report follows the pattern of focusing on the claimant’s account of the typical day and checking this for purported counter-examples. This also has the (possibly unintended) side-effect that because the LiMA manual asks for positive examples of ability only, it is very hard for the claimant to get evidence on record that might suggest they meet a descriptor. It biases the process towards non-entitlement.

Over-reliance on absence of treatments or input

A very common theme in ESA assessments is an observation that a claimant is not in receipt of certain treatments. Often this is summarised as ‘no specialist input’ or similar. Sometimes the particular treatment in mind is specified, such as ‘not been referred to the memory clinic’, or appropriate treatment is minimized, such as ‘only on standard medication’ or ‘a standard dose of painkillers’.

While it can, with care, be possible to draw appropriate inferences where an expected treatment is absent, there is reason to think this assessment technique is not being used appropriately. In a recent housing case, overuse of a similar asessment technique was criticized by the Central London County Court[3].

Many treatments may not be appropriate or available, and there may be other good reasons for a claimant not receiving them other than an absence of need. Receipt of treatments or certain medical input is not part of the criteria for receiving sickness or disability benefits and it is wholly wrong for assessors to substitute this as a proxy for the proper criteria. It is especially worrying when an implication that to qualify, a claimant should be overdosing on painkillers.

Problems specifically known to affect ESA

Use of rules of thumb

In 2013 Dr Greg Wood resigned from Atos and revealed that he had been trained to apply various rules of thumb which ‘he believed deliberately traduced the more nuanced written instructions in the assessors' handbook on five critical areas, so that assessors were less likely to award points to claimants’[4]. Training of these ‘rules’ was also independently recorded in undercover filming[5].

Despite the denial of this by Atos, and some attempts to backtrack from the more obviously incorrect rules, evidence of these rules or variants of them persists in reports.

Examples: We assisted Mr S to appeal against a decision incorporating a purported finding that he could mobilise more than 200 metres without stopping. The evidence cited was that he could walk briefly close to his caravan and was observed walking with a stick from room to room at the assessment.

A further effect of rules of thumb is that they discourage the assessor from recording, or even asking, enough information to arrive at a proper decision, since the cursory information is wrongly thought to be enough. (See also ‘counter-examples’ below)

Use of statistical norms

Rumours of targets or incentives to find more claimants ineligible for benefit have long haunted ESA. Then Minister for Work and Pensions, Chris Grayling, denied the existence of any targets anywhere in the system[6].

What has been now admitted is that assessors for ESA work to ‘statistical norms’[7].These are said to be different from targets, as an assessor does not know what the ‘target’ is. This is still problematic, as an assessor will know their numbers are being tracked so will still indirectly be influenced.

Since tracking norms requires assessors to follow the behaviour of other assessors, this leads to a self-perpetuating culture of incorrect outcomes. And since it is widely accepted that the initial tendency was to pass too few people, this set the course of the norm in an unmovable direction.

DWP objections to purported ‘double counting’

While not a consistent problem, the DWP has made repeated attempts to restrict what it purports to be ‘double counting’ - that is to alter the criteria so that a limitation cannot score points in more than one assessment activity. Most notably, an attempt at this succeeded in 2011 when three of the mental health activities were removed from ESA on this basis, significantly harshening the test[8].

As noted in the introduction above, it is normally necessary for a claimant to score in more than one area to qualify for benefit. Far from being ‘double counting’, this is a basic feature of the assessment process that has been created. Any future attempts by the DWP to resurrect this as an issue should be viewed with considerable suspicion.

Ministers statements on criteria

It is often the case that claimants’ expectations have been poorly managed by statements from ministers. People who are then found not to be eligible are made to feel as if they have been caught out or in some way accused of making a false claim. This is especially unfair in ESA, as a subsistence benefit for those unable to work, who may well have no other source of income and who will have come to be on ESA simply through being signed off work by their GP with no opportunity to make a positive decision to claim on the merits of their case. The assessment is usually the first time they learn that their GP’s note is not the final word.

There has been a long trend for ministers to make statements such as ‘we are committed to ensuring that those who cannot work are protected’ or that the ‘most vulnerable’ will not lose out.

When a claimant actually tries to claim the benefit and is found ineligible, the Department appears to retreat from such statements and to replace them (privately) with platitudes about the test requiring a ‘very high degree of disability’, or that despite significant disabilities being accepted, they are ‘not enough to be found to have limited capability for work’.

This disparity between ministers’ statements and the actual tests have led to increased suspicion and abuse of claimants, such as that seen in the right-wing press who feel empowered to assert that anyone who was ‘genuine’ would have easily passed their assessment[9]. This situation risks causing misguided campaigns for eligibility to be tightened even further.

This inconsistency was again seen in 2016 when the work-related component of ESA was abolished, with many MPs expressing an incorrect belief that anyone incapable of work was entitled to be in the support group – in effect, that being allocated to the support group was easy and expected. Even the newly appointed Secretary of State at the time had to correct a misleading statement to this effect[10]. This has led to a noticeable rise in people who feel that being placed in the work-related activity group of ESA is an insulting denial of their condition, or that it is equivalent to being found fit for work.

We would urge that the DWP should be required to ensure that there is a consistent message in public and towards claimants. In particular, if the policy is that significantly disabled or vulnerable people can nevertheless be ineligible for extra support, or should fall within the sanctions system, ownership of this needs to be publicly accepted by the DWP.

[1]              See for example the discussion in the Upper Tribunal in NH v Secretary of State [2014] UKUT 114 (AAC)
[2]              Extract from the LiMA Technical Manual “Tip – Say what the claimant CAN do. LiMA works best when it has positive information about the claimant’s abilities. For example, the phrase ‘Can walk 400m to the shops’ is much more useful than the phrase ‘Cannot walk 800m to the town centre.’ Wherever possible, try to gather information about what the claimant CAN do.” [In an independent report, one is left to speculate why this is seen as more useful.]
[3]              Thomas v Lambeth LBC
[5]              Channel 4 - Dispatches – Britain on the Sick -
[6]              Channel 4 - Dispatches  -
[7]              Freedom of Information response
[9]              Daily Express 14th Oct 2009, 26th Jan 2011, 27th Jul 2011 “75% on sick benefits are faking” This story was repeated across the press in similar terms. It is clear that the figures were sourced from a closed DWP press release.

No comments:

Post a Comment