Hot on the heels of our post on indirect discrimination in employee tests for promotion comes another decision posing similar challenges for employers.

Government Legal Service –v- Brookes concerned the Situational Judgement Test (SJT) which forms part of the recruitment process for lawyers wanting to join the Service. Ms Brookes told the GLS in advance that she was likely to find the multiple-choice format of the SJT particularly difficult because of her Asperger’s Syndrome. She wanted to answer the same questions but in short narrative rather than multiple-choice form. That option was not provided by the GLS.  Brookes narrowly missed the SJT’s pass mark and so her candidature for the role fell at that stage.

Brookes claimed that the requirement to use a multiple-choice format represented both indirect disability discrimination and a failure by the GLS to make reasonable adjustments. It was accepted by the GLS that Asperger’s could make dealing with the multiple-choice format more difficult for sufferers than for those without it. However, an indirect discrimination claimant needs to show not just group but also individual disadvantage.

The principal question at issue in the EAT was therefore whether Brookes could show that she herself had been prejudiced in that way. The medical evidence was not definitive – even the GLS’s own expert conceded that some people with Asperger’s would find that format a problem while others would not. In addition, the SJT and other tests were a substantial part of the GLS’s process for winnowing several thousand applications down to about 35 successful candidates each year. Statistically there would therefore be perhaps several hundred non-disabled people who equally narrowly failed the SJT. How could Brookes show in those circumstances that it was the Asperger’s that had seen her miss the mark rather than something else?

Both the indirect discrimination and reasonable adjustment claims were upheld by the Employment Tribunal, Brookes representing herself and so potentially putting beyond argument that she was in fact up to the role.

In rejecting the Service’s appeal, the EAT set a standard of proof of individual disadvantage which some employers may regard as disconcertingly low. There was no direct evidence about whether the Asperger’s had caused her to fail but the ET essentially formed its own view from peripheral circumstances, not least Brookes’ cogent presentation of her own case, that she had had adjustments made for her at university, had obtained an ok degree and seemed outwardly committed and intelligent. Although it did not say so in terms, it then appeared to place the burden of explaining the failure on the GLS. Since the Service obviously had no better idea of why Brookes had ticked slightly too many of the wrong boxes than it had as to hundreds of others, it could put forward no counter-explanation.

In fact, the EAT went further. Not only was the ET entitled to reach a finding on that largely circumstantial evidence that Brookes had failed the test because of her disability but actually it had not needed even to go that far – it would have been enough for the ET to find merely that Brookes’ Asperger’s made the test substantially more difficult for her to pass, and not that it had actually been what caused her to fail.

On the reasonable adjustments front, both the ET and EAT accepted that multiple-choice as a format made good sense for the Service because answers could be marked by computer quickly, cheaply and without human subjectivity. Nonetheless, given the scale and duration of the recruitment process and the resources of the GLS, both agreed it would have been possible for a small number of people to have been marked manually on narrative-format answers without undue delay or expense. The Service’s refusal was found to be a matter of principle, not practicability, and so it went down on the reasonable adjustments claim too.

This case plus Essop seems to allow a substantial distance between being statutorily disabled on the one hand and establishing some consequent personal disadvantage from an employer’s PCP on the other. However, neither case means you need to make recruitment or promotion standards lower for disabled people. Neither one requires the employer to water down its tests to the point where they no longer act as a proper filter of good enough from not good enough. The GLS was fully entitled to insist on high-level decision making skills in its candidates, but there is nothing about that skill which requires it to be tested by multiple-choice only.  The skill and the test were different. You can not assess the ability to bang nails in straight by a written test, but you can assess decision-making by other formats which do not have an adverse impact on those with certain mental health conditions.

Lessons for employers

  • If on disability grounds someone asks you to make an adjustment to your assessment processes for promotion or recruitment which (i) does not prejudice its effectiveness as a test of whatever skill set you are testing; and (ii) is logistically doable, then you should probably make it.
  • But even if a request for adjustments is made which you believe cuts across (i) or (ii) above, you should still be seen to examine it with an open mind before rejecting it. Refusal as a matter of unthinking principle will almost always be fatal to your defence.