Employment and Support Allowance (ESA) is the new sickness benefit introduced in October 2008, which will eventually replace Incapacity Benefit. Its new test, envisaged by the legislator, is extremely difficult to pass and almost nobody is exempted from the test. Brighton Benefits Campaign joins the emerging national campaign against ESA, and the methods used by private provider Atos Healthcare to assess sick people for this benefit.
Medical assessment and automation: ‘The Computer says no’
A serious problem for sickness claimants today is that their medical assessment is made through the use of a computer program, the ‘Logic Integrated Medical Assessment’ (LiMA). This programme was introduced by DWP’s contractor Atos Origin in 2005 to assess claimants for the old sickness benefit (Incapacity Benefit). Already in December 2005 advice charity Child Poverty Action Group complained of serious problems with Atos’s computer-aided medical assessments in an article entitled ‘The computer says no’ [CPAG 2005].
How does LiMA work? During the examination, the ‘doctor’ asks the claimant a list of questions, which the computer shows on the screen. While the claimant speaks, the ‘doctor’ builds, through the use of a windows system and drop-down menus, a combination of words and numbers which resembles what the claimant says. For example, sentences are created by assembling bits such as ‘The customer keeps themselves busy doing’… (choice of activities)… ‘for’… (choice of numbers)… ‘minutes/hours…’. Finally, the computer uses these sentences to decide the score.
The result is an often surreal computer-generated Medical Report, based on a collection of mechanically-constructed brief sentences. More than one Social Security judge has noticed with alarm that automated reports did not reflect what the claimants had actually said and contained ‘nonsensical statements’ [see judgements CIB/476/2005, CIB/664/2005, and CIB/511/2005]. The judges also noticed that, in the medical reports,
‘Phrases appear to be capable of being produced mechanically without necessarily representing actual wording chosen and typed by the examining doctor’ [CIB/511/2005]
After aiding the ‘doctor’ to construct brief phrases, the computer automatically derives implications from these phrases. These implications are often overstretched, or even totally wrong, and options for investigation of these implications are blocked by the computer system [CAB 2010, p. 19]. For example, if a claimant cannot cook but tells the doctor that he makes himself snacks or tea, this ‘ability’ is used by the computer as ‘evidence’ of a very large range of physical and mental capacities.
Examples of unfair sentences used by the computer against claimants are: ‘usually can do light gardening for 1 minutes’ [cited by a judge in CIB/664/2005]. Or: ‘can take adequate care of his goldfish’ [Rightsnet Forum 2008]. Some sentences are complete nonsense. Examples: ‘Customer does pottery all day’ (the claimant told the ‘doctor’ that she was pottering about) [Rightsnet Forum 2008] and: ‘The client’s Amputation of Upper Limb is mild. They have seen a specialist for this problem’ [Rightsnet Forum 2010].
Last, but not least, nobody can know the logic that connects the computer-generated sentences to the scoring system. In fact the computer program is protected by commercial confidentiality – it’s Atos’s private property. A judge complained that Atos refused publishing its handbook because of commercial confidentiality [CIB/664/2005]. For this reason, one of our clients will never know why LiMA decided that he can walk 100 metres, pick things from the floor and sit on a chair for 30 minutes on the basis that he is ‘usually able to use a microwave’ [BHUWC].
Automaton and profits
Automation was made possible by the introduction, by the Tory government in the 90s, of a new test for Incapacity Benefit based on a box-ticking method and a score system.
Automation has served to deskill medical assessments. Medical examinations are made quick and superficial. The computer establishes the type of questions; and its structure does discourages its users from making further investigations or asking different questions. The ‘doctors’ may, if they want, overrule the computer’s decision, but their Handbook discourages them to do so.
‘[The EBM rules] have been tested and found to be accurate in the large majority of cases. You should think carefully before overriding the EBM rules, and should do so only on those occasions where there is a good reason as to why they would not apply in this case’ [LiMA 2004, p. 31].
Who is paying for these cheap medical examinations? The claimant. Who gains? Atos: with the introduction of a computer programme who ‘thinks’ for the doctor, Atos can now use non-doctors staff, who are no doubt cheaper than a real doctor [CPAG 2005, p. 5].
A history of complaints
Considering the above, it is not a surprise that advice agencies were very worried. Back in February 2006, the Citizen Advice Bureau (CAB) complained that:
‘Doctors pay more attention to the computer than the client; the system is inflexible and gives rise to inappropriate stock phrases in reports; options for investigation and findings are blocked off by the system inappropriately; doctors sign off reports without checking what they say, because the phrases have been generated by the system, not the doctor’ [CAB 2006, p. 10].
In the same document CAB showed how Atos’s ‘doctors’ produced inaccurate reports, reporting incorrectly what the claimant had said about their own conditions and taking answers out of context [CAB 2006].
The extent to which the government’s took notice of this report can be evinced from a new CAB’s report, written three years later (September 2009):
‘Doctors pay more attention to the computer than the client; the system is inflexible and gives rise to inappropriate stock phrases in reports; options for investigation and findings are blocked off by the system inappropriately; doctors sign off reports without checking what they say, because the phrases have been generated by the system, not the doctor’ (sic) [CAB 2009, p. 4].
And, the CAB adds:
‘…Doctors produce inaccurate reports… reporting incorrectly what the claimant has said about their own conditions and taking answers out of context’ [CAB 2009, p. 3].
It is worthwhile mentioning that in their recent report of March 2010 the CAB complains about the same problems:
‘We still hear repeated reports of rushed assessments, assumptions being made without exploration, inaccurate recording’ [CAB 2010, p. 27].
Despite criticism from judges and complaints, LiMA continued being used; in fact it has now been adapted to the new, and tougher, ESA test. Let us now consider this new test.
The ‘Limited Capability for Work Assessment’ (CWA)
After the introduction of LiMA, the old test for Incapacity Benefit (Personal Capability Assessment, PCA) became much harder. But the new test for ESA is worse. This is due to a combination of the application of the assessment methods described above and the harshness of the test itself.
Both old and new tests are based on scoring points with respect to a number of physical and mental ‘activities’, such as ‘walking’, ‘sitting’, etc. (these activities are called ‘functional areas’). Levels of capacity are described by ‘descriptors’ such as ‘can’t walk at all’; ‘can’t walk more than 50 metres on level ground without repeatedly stopping ’; can’t walk more than 100 metres’, etc. In the new test these descriptors score 15, 9, 6 or 0 points. A total score of 15 is needed to get ESA.
In their report of March 2010, the CAB explains in detail why this test is harsher than the previous.
Firstly, there are almost no exemptions from the test for very ill claimants [CAB 2010, pp. 8-14]. This is a problem, for example, for mentally ill claimants who are now aware of their illness.
Secondly, low level problems score zero, while they were scoring something (three points) in the old test; yet the effect of combined low level problems can make people incapable to work [CAB, pp. 13-15].
Third, the ‘doctors’ are recommended to disregard symptoms which come and go (variable conditions).
Back in 1996, a judicial decision had obliged the government to take variable conditions into account in the old test; now, using the excuse that the WCA is a new test, the government has wiped out what had been established!
If we compare the old and new tests, descriptor by descriptor, we are surprised how strict the new test is.
Under the old PCA, for example, if a claimant was unable to walk up or down a flight of stairs of 12 steps, he scored 15 points and was ‘unfit to work’; if he could climb 12 steps only with the help of a banister, he could still score some points. Under the new WCA, in order to score more than zero, one should not be able to climb two steps, even with the help of a handrail. This is extreme.
Like the physical health test, the mental health test in the WCA identifies a number of activities such as ‘learning or comprehension of tasks’, ‘memory and concentration’, ‘awareness of hazards’, etc. A significant impairment in each one of these functions makes a person unemployable. But this is not what the legislator has decided! In fact, under the new test, if a person has a significant problem with only one of these tasks (and perhaps low level problems with other tasks) he is likely to not pass the ESA test. An example: how poor should my learning ability be to make me ‘unfit to work’? The legislator answers: unbelievably poor. If, in order to learn how to use a new kettle, I need to witness more than one demonstration, and I also need to receive a verbal prompting the day after, I will not get the 15 points which qualify me for ESA. If I have l score zero elsewhere, here you are, I am ‘fit for work’, despite my capacity to learn is so bad.
Scoring even a minimum of 6 points is very difficult, because each activity is defined too broadly. For example, let’s consider the ‘functional area’ of ‘coping with social situations’:
‘Normal activities, for example visiting new places or engaging in social contacts, are precluded because of overwhelming fear or anxiety’’ (One has six points if this is happens ‘frequently’, zero if it’s less ‘frequently’).
We don’t need to be geniuses to realise that this is a trap. Any human being, with the exception of those who are cataleptic or bed-bound, do some ‘normal activities’: so, whatever I tell the ‘doctor’ I do, if this can be taken as evidence that ‘normal activity’ is not ‘precluded’.
It must be now clear how, thanks to the combination of this new test and the use of LiMA, anything from shopping, making ourselves tea or taking care of a goldfish will give us zero points.
Finally, the Handbook for Atos’s ‘doctors’ asks them to apply the harshest possible interpretation to the test. For example, the handbook says that if a claimant’s memory is not so bad that he forgets to get dressed in the morning, he cannot really score anything at all for memory problems. CAB commented,
‘Forgetting to get dressed in the morning as an extreme sort of memory loss and clearly ought to be awarded more than six points. How many employers would take on someone who forgot to get dressed?!’ [CAB 2010, p. 17].
It is a fact that LiMA can decide the score automatically, and we suspect that unlawful interpretations such as the one above have been incorporated in the programme – the doctors are instructed by their Handbook to think according to the same lines so as not to be surprised by, and antagonise with, the computer’s results.
The ESA50 questionnaire
To make things more difficult, the ‘ESA50’, a self-assessment form which the claimant is asked to fill in before the medical assessment, is another trap. Benefits Advisers have complained that this questionnaire does not reflect the scoring system and in many pages its tick boxes do not include possible high-scoring answers.
Ticking the wrong box will allow the ‘doctor’ to give claimants zero points for given activities without having to ask questions during the medical assessment [BHUWC].
Media response and political reaction
Not surprisingly, the new test has already found seven in ten sick claimants ‘fit for work’. CAB says that ‘an adviser from a community mental health team reported that almost all their new clients applying for ESA are being refused benefit’ [CAB, 2010, p. 11]. This outcome is used by government’s agencies to claim that most sick claimants are well enough to work [Daily Express 2010].
The truth has however emerged in the media, that the new test is too tough [The Guardian, 2009 and 2010]. A recent article in the Herald of Scotland mentions a claimant who has attempted suicide after failing his ESA assessment [The Herald, 2010].
In January false news spread that ministers backed down about the ESA test [The Guardian, 2010]. In fact the government has amended the law to make the test tougher [Benefits and Work, 2010].
The consequences of an unfair assessment
The government’s statistics show that out of 100 sickness claimants who are assessed, 69% are told that they are ‘fit for work’. Those who fail the ESA test will have to seek jobs and claim JSA, but will have less chances of finding a job than healthy people, often none. As CAB wrote:
‘Research by the Chartered Institute of Personnel and Development (CIPD) discovered that “more than 60 per cent of employers said they disregarded applications from people with… a history of mental health problems or incapacity… Research carried out by Rethink highlights the fact that fewer than four in ten employers would consider employing someone with a history of mental health problems”’ [CAB, 2010].
In October 2009 The Guardian commented:
‘The new ESA is meant to offer disabled people better and more personalised support to get back to work. But if the assessment is made so tough that people are not getting the benefit in the first place, then they will also not be getting the support that could help them get into employment’ [The Guardian, 2009].
The CAB now demands a ‘real-world test’, which assesses the employability of claimants and not only their inability to perform abstract ‘tasks’ and which should also consider concrete obstacles such as discrimination in the labour market [CAB, 2010]. However, without a political national campaign, appeals from good-hearted charities will be dismissed by a government whose interests coincide with those of capital.
Employment and Support Allowance: a new brutal regime
Out of 100 sickness claimants who are assessed, 22% receive ESA, but are considered sufficiently well to undergo compulsory ‘Work Focused Interviews’ – they can be asked to follow compulsory ‘work-related’ activity.
Is this help?
The government has presented ESA as a new regime which offers ‘personalised help’ to the sick to make themselves more employable. But is this true?
No, it is not. For many years, sick claimants have been able to receive free personalised help to make themselves more employable, find suitable training and courses and even find a job. ESA has only introduced a factor of compulsion.
In fact, the government’s propaganda that people need compulsion is misleading and hides the real problem: a tough medical assessment. Many sick people, although unable to work, do wish to attend courses, and they don’t because they are afraid of losing their benefit. And they are right! Any mention of having done any activity at all will be used to disqualify a claimant from his sickness benefit. An extreme example from CAB: a claimant was found ‘fit to work’ because he attended a compulsory ‘Work Focused Interview’! The automated medical report said: ‘the client is actively seeking work through Jobcentre Plus’ [CAB, 2010, p. 19].
Compulsion will only cause trouble for those who are not ready for activity. Claimants who are asked to go to training or courses may fail to attend, cause disruption, etc. and as a result they will be sanctioned. Many will be too ill to seek help if they are sanctioned, and will opt to live on a reduced rate of benefits. These people will not be made more suitable for jobs, they will only increase the ranks of the forgotten, and helpless, poor.
Last but not least: the issue of compulsory medical treatment
Originally, ‘work-focused activity’ could include compulsory medical treatment. However, on 22 June 2009 an amendment of the section 14 of the Welfare Reform Act 2007 was presented to the House of Commons, which excluded medical treatment [Hansard, 22 June 2009]. However, we still need to be alert. Although accepting medical treatment is not explicitly part of a ‘work-focused activity’, it can still be a condition for entitlement to ESA, as it has been in the past for Incapacity benefit and Disability Living Allowance – with the approval of Social Security judges [see, e.g., CDLA/3908/2000].
No to ESA, no to Atos
A fight for a more human medical assessment is also a fight for the abolition of Atos’s mechanised methods – for the dismissal of Atos and the return to a professional medical service, publicly run, and run not for profits.
On 12 March 2010 Jonathan Shaw, Parliamentary Under-Secretary for the DWP, recently confirmed that about £9,3m of taxpayers’ money has been paid to Atos per month in the six months from September 2009 to February 2010 (totalling about £111m per year), for providing their medical examinations, and that this figure included the cost for developing their computer program (‘new technology and other service improvements’). Not only does Atos provide cheap and automated medical examinations, but the taxpayers are asked to pay for this! [Hansard, 2010].
The latest contract with Atos Origin (now called by its ‘trading name’ Atos Healthcare) was renewed in March 2005, in a deal worth £500m over 7 years. On 12 March 2007, Computer Weekly said that this contract could be extended by a maximum of five years under two separate extension clauses: the first for three years and the second for two years, and that these extensions could take the potential total contract value to more than £850m over 12 years [Computer Weekly, 2005].
The first deadline for renewal seems to be in 2012, so we are still on time for a campaign to sack Atos. And perhaps this will not be too difficult, as the DWP has already considered terminating its costly contract with Atos! Indeed, in October 2008 Atos Origin lost a memory stick containing details and passwords to access a major database of 12m benefit claimants, the government ‘Gateway’ website. Following this, in November the government thought about terminating their contract [Computer Weekly, 2008].
Benefits and Work, blog, 2010, ‘Thousands will lose benefits as harsher medical approved’: http://www.benefitsandwork.co.uk/blogs/2010/04/13/thousands-will-lose-benefits-as-harsher-medical-approved/
Brighton and Hove Keep our NHS Public. This campaign group found out that, on the 16th June 2005, the under-secretary for the Department of Health, Caroline Flint, declared to the House of Commons that NHS patients would not be refused medical treatment if they refused having medical information on national databases. However, this is still a condition for having a smear and the Department for Health refuses to answer to letters which quote Caroline Flint.
BHUWC, Brighton and Hove Unemployed Workers Centre. We can provide (anonymised) pages of Reports.
Citizen Advice Bureau, February 2006, ‘What the doctor ordered?’: http://www.citizensadvice.org.uk/index/campaigns/policy_campaign_publications/evidence_reports/er_benefitsandtaxcredits/what_the_doctor_ordered
Citizen Advice Bureau, September 2009, ‘Decision making and appeals in the benefit system’: http://www.citizensadvice.org.uk/index/campaigns/policy_campaign_publications/consultation_responses/cr_benefitsandtaxcredits/decision_making_and_appeals_in_the_benefits_system
Citizen Advice Bureau, March 2010, ‘Not working, CAB evidence on the ESA Work Capability Assessment’: http://www.citizensadvice.org.uk/index/campaigns/policy_campaign_publications/evidence_reports/er_benefitsandtaxcredits/not_working
Computer Weekly, 24 March 2005, ‘Atos wins DWP deal extension’: http://www.computerweekly.com/Articles/2005/04/01/209047/Atos-wins-DWP-deal-extension.htm
Computer Weekly, 3 November 2008, ‘Latest data blunder rocks faith in government security record’: http://www.computerweekly.com/Articles/2008/11/03/233173/latest-data-blunder-rocks-faith-in-government-security.htm
Child Poverty Action Group, 24 March 2005, ‘The computer says no’: http://www.cpag.org.uk/cro/wrb/wrb189/computer.htm
Daily Express, 1 February 2010, ‘Benefit Britain exposed’: http://www.express.co.uk/posts/view/155413/Benefits-Britain-exposed
Hansard, 22 June 2009, http://www.publications.parliament.uk/pa/ld200809/ldhansrd/text/90622-gc0007.htm
Hansard, 12 March 2010, Column 515W, link: http://www.publications.parliament.uk/pa/cm200910/cmhansrd/cm100312/text/100312w0009.htm#10031228008975
Rightsnet Forum, 18 June 2008, ‘IB85 PCA reports – it’s all in the mind (of the doctor)’: http://www.rightsnet.org.uk/dc/dcboard.php?az=show_topic&forum=111&topic_id=2964&mesg_id=2964&page=14
Rightsnet Forum, 15 January2010, topic: ‘The client’s amputation of upper limb is mild’: http://www.rightsnet.org.uk/dc/dcboard.php?az=show_topic&forum=111&topic_id=4761&mesg_id=4761&page
The Guardian, 14 October 2009, ‘Testing time for disabled people’: http://www.guardian.co.uk/commentisfree/2009/oct/14/disabled-people-sickness-benefit
The Guardian, 3 January 2010, ‘Retreat on draconian disability testing’; 23 March, ‘Flawed benefit system classifies terminally ill man ‘fit for work’; see also Amelia Gentleman, The Guardian blog, ‘Readers detail experience of being found fit when not able to work’: http://www.guardian.co.uk/society/joepublic/2010/mar/29/reader-responses-fit-to-work
The Herald of Scotland, 1 February 2010, ‘Back-to-work benefit row’: http://www.heraldscotland.com/news/health/back-to-work-benefits-row-1.1002753