Today’s outrage against objectivity and rational analysis is brought to you courtesy of the founder and Director of Civitas, an independent ‘think tank’. David G Green has written an article in the Telegraph giving a pithy critique of the Equality Act 2010.
David G Green is not a fan of the Equality Act, regarding it chiefly as a vehicle for avaricous lawyers and dishonest employees to rob innocent hard working entrepreneurs of what meagre earnings they have managed to scrape together. This is clearly something about which Mr Green feels deeply. So deeply in fact that he has not felt the slightest need to learn anything about what the Act actually says before committing his rage to paper. His anger comes straight from the heart, not the head.
Nor is his anger merely the fleeting mood of a moment. Mr Green has felt this way from the very beginning, as this article from the Telegraph back in 2009 shows (thanks @Tribunalwatch – well spotted). There is some really bizarre stuff in the 2009 article which seems to confuse the concept of the burden of proof in discrimination cases with the definition of indirect discrimination and prays in aid Kipling and the Marquess of Queensbury rules. But that was 2009 and we have moved on now. I am happy to say that there are fewer errors in the new article – although, to be fair, it is also quite a lot shorter.
Now you may gather that I am not entirely sympathetic to Mr Green’s point of view. I happen to think that it should be unlawful, for example, for an employer to refuse to employ black people, or deliberately pay women less than men for doing the same job (call me an incorrigible lefty if you like). It seems that Mr Green does not agree. Fair enough, that is his right. He is entitled to his opinion. What riles me is not the extreme position that he takes, but the fact that in both articles he grossly misrepresents the burden of proof in discrimination cases. I know that seems to be a rather technical and prosaic thing to get worked up about, but there it is. This is the sort of thing that bothers me.
Lets focus on yesterday’s article. What Mr Green says is this:
Worst of all, the burden of proof in discrimination cases has been reversed so that once an accusation has been made it is up to employers to prove their innocence. The requirement that citizens should have to prove their innocence was imposed by the EU, contrary to our longstanding heritage of protecting citizens against false accusations by requiring accusers to prove their case. Innocent until proven guilty used to be the watchword of our judicial system – but not now that the EU has insinuated itself into our legal processes
Reading that paragraph over again, I think what really makes the red mist descend is not so much that he is wrong, but the lofty and supercilious tone that he adopts in denouncing the way in which the EU has perverted the principles of British justice – it really hasn’t. I’m also annoyed because Mr Green has not completely made up the concept of a reverse burden of proof, he has merely misrepresented it. The danger is that in explaining where he has gone wrong, I could end up writing a rather boring essay and it could look to the casual observer as though I am just being picky and that Mr Green’s general point stands. I’m not and it doesn’t, but there is obviously a risk that it will look like that to the casual observer. Luckily if you have got this far, you aren’t that casual, so here goes.
Proving Discrimination
We start with the problem. Nowadays, most employers are unlikely to wear their prejudices on their sleeve. If they are going to discriminate against someone, they are unlikely to say ‘sorry love, but this isn’t really a suitable job for women’ or ‘we don’t want your kind round here, we only employ white people’. Every so often a case like this comes along of course, but they are quite rare. In general, direct evidence of discrimination is hard to come by.
So how can we spot discrimination when it does happen? It’s all about drawing inferences from the facts that have been proved. Even before the EU began to think about a reverse burden of proof, the UK courts had all but come up with the concept on their own.
The classic explanation comes from the case of King v Great Britain China Centre [1991] IRLR 513. Here the Court of Appeal distilled some general principles from the case law and said that where someone had been treated less favourably than someone of a different race (for example), then that might ‘point to the possibility’ of race discrimination. Lord Justice Neil then said:
In such circumstances the tribunal will look to the employer for an explanation. If no explanation is then put forward or if the tribunal considers the explanation to be inadequate or unsatisfactory, it will be legitimate for the tribunal to infer that the discrimination was on racial grounds.
In other words, where facts are proved that indicate that discrimination may have taken place, the tribunal can look for an explanation from the employer of the difference in treatment. If that explanation is not believed by the tribunal, then the tribunal can infer that discrimination has taken place. From 1991 onwards, that was the standard way in which the burden of proof was dealt with in the UK. In formal legal terms the burden did not shift, but in reality an employer had to be able to demonstrate a non-discriminatory reason for a difference in treatment in order to defeat a claim.
Its worth stressing that this was nothing to do with any directive or ruling from the EU. Indeed the principle was developed in race discrimination cases which were not even covered by EU law at that time.
A shifting burden
The formal shift in the burden of proof came about as a result of the burden of proof directive in 1997, which has now been replaced by the Equal Treatment Framework Directive which says this:
1. Member States shall take such measures as are necessary, in accordance with their national judicial systems, to ensure that, when persons who consider themselves wronged because the principle of equal treatment has not been applied to them establish, before a court or other competent authority, facts from which it may be presumed that there has been direct or indirect discrimination, it shall be for the respondent to prove that there has been no breach of the principle of equal treatment. (Article 10)
As a result,we do now have a shifting burden of proof under the Equality Act (S.136), but the key requirement is that the claimant must first of all prove facts ‘from which the court could decide, in the absence of any other explanation’ that discrimination has taken place. It is only if those facts are proved that that the burden shifts to the employer to prove that there was no discrimination. In its essentials, the process is not so very different from that described in King back in 1991.
If anything, the bar has in practice been raised somewhat, with the Court of Appeal in Madarassy v Nomura International plc stressing that the claimant must prove ‘something more’ than a mere difference in race (for example) and a difference in treatment. Just what that ‘something more’ needs to be is not completely clear and we are currently getting quite a lot of tedious case law struggling with the precise mechanism by which the burden of proof passes to the employer. On the whole, employers seem to be getting the best of it, but that is a different story. The fundamental point is that when Mr Green says
once an accusation has been made it is up to employers to prove their innocence
he is categorically completely and unarguably wrong. He was wrong in 2009 and he is wrong today. A mere accusation simply does not place any burden of proof on the employer, the claimant has to do much, much more than just make an accusation. All employment lawyers know that Mr Green is wrong and could explain to him where his mistake lies. They could show him the case law and highlight the key passages for him so that he could understand just how wrong he is.
I wonder if he would be grateful for that?
Because the thing about this sort of article is that the author – and the paper indulging him – are not interested in getting the facts right. Mr Green is a clever man and I’m sure he could understand this subject if he chose to. It seems almost as if he wants the law to be as unreasonable and un-British as possible, the better to denounce it. Any actual understanding of the subject would simply get in the way of his rhetoric.
Business deserves better
What makes me livid about this sort of argument is that ignorance of employment law is bad for business and Mr Green’s article practically revels in its ignorance; he luxuriates in it. Clearly he enjoys the fantasy he has dreamt up of the horrid Europeans forcing good old Blighty to sacrifice its ancient liberties. However, there is a danger that employers and business owners could read Mr Green’s article and assume that he knows what he is talking about – and that’s a depressing prospect. They will think that they will indeed be forced to prove their innocence whenever any accusation of discrimination is made against them, irrespective of its merits. That could lead them to settle cases that shouldn’t settle, or avoid making decisions that their business needs them to make because of their fear of a discrimination claim. Mr Green and the Telegraph would no doubt consider themselves to be on the side of British business, but by promoting a false view of employment law they do British business a great disservice.
I know that there is a right-wing push on at the moment to weaken employment protection. I’m not too dogmatic about that and keen to have a rational debate about the issues. But can we at least start from a position of accurately assessing what the law currently is before we start hacking away at it?
A Range of Reasonable Responses 
The question of what the burden of proof rules actually mean in practice is fraught with difficulty, and I can see where a lot of this “guilty until proven innocent” fear comes from. Indeed, from reading King v Great Britain Chine Centre it is easy enough to conclude that where an employee shows that he has been treated less favourably than someone of another race, this is of itself something which could lead the tribunal to “look to the employer” for an explanation, and to find for the claimant if no adequate explanation is provided. We lawyers cling to the idea that “something more” is required than just a difference in treatment and a difference in (say) race, but it really is almost impossible to draw a clear line as to what is actually required to tip the balance. In some cases, the “something more” could simply be the lack of an explanation (e.g. the CA decision in Bahl v Law Society, para 101, or the failure to respond to a discrimination questionnaire (S. 138, Equality Act 2010).
I personally have my doubts whether silence on the employer’s part is of itself something which ought to be taken into account in deciding whether the burden shifts. It is as if the tribunal were saying “The reason we require you to provide an innocent explanation is because you have not yet provided an innocent explanation”.
On the other hand, the burden of proof rules do not in practice seem to have caused a major change in the ease with which discrimination can be proven. As the House of Lords has recently emphasised in Hewage v Grampian Health Board, at para 32, the burden of proof is really only relevant in fringe cases, where there really isn’t enough evidence to decide the matter one way or the other. In most cases, the tribunal will simply come to a decision on the evidence as to which party it believes, and the burden of proof as such doesn’t come into it.
More importantly, tribunal statistics demonstrate that a discrimination claim (particularly a race claim) is still the hardest claim for an employee to win, despite the burden of proof rules supposedly being stacked in their favour, as I blogged 2 years ago.
Personally I think the change in the burden of proof rules has actually had a negative impact, but not for the reasons David G Green suggests. It has led to no great improvement in success rates for claimants, but rather has led to appeal after expensive appeal to the EAT and Court of Appeal as to whether the tribunal in a given case has shifted (or not shifted) the burden at the right time. While there were appeals under the previous rules (i.e. under King v GB China Centre) there has been a phenomenal growth in such appeals since the burden of proof rules came in. Appeals on points of law such as this are costly, time consuming, and in my view deny justice to deserving claimants, who have their compensation payments delayed as a result of such appeals and have to to spend a significant wedge of whatever compensation they do get on legal fees to fight the appeal.
Furthermore, the rules have been a gift to the “red tape” lobby, giving them something to whine about endlessly in articles in the Mail and the Telegraph complaining about the EU, employment rights, or the so-called “compensation culture”.
I think there is something structurally wrong with our approach to proving discrimination. I came to this conclusion when I started a project to explain discrimination law in the simplest way I could. It proved insanely difficult.
It seems to me that most people expect anti-discrimination law to stop employers treating employees adversely because of their protected characteristic. At the moment only Pregnancy and Maternity discrimination are formulated in that way. It\’s a key point, so I want to emphasise it: We lawyers think the legislation is aimed at tackling ”less favourable” treatment rather than ”unfavourable treatment” because of a protected characteristic. I think that is wrong. The better way of understanding the legislation is that ”less favourable treatment” is used as a tool to identify ”unfavourable treatment”. It is part of a statutory test rather than a definition of the mischief the legislation is aimed at.
Assume that the legislation simply banned unfavourable treatment because of a protected characteristic, how would we establish that adverse treatment is ”because of” the protected characteristic? Well, absent an admission, the single most telling piece of evidence would be disparity of treatment. Where there is no evidence of disparity of treatment we would have to ask what else there was that pointed to the protected characteristic being some part of the reason for the adverse treatment.
The present legislative formulation is confusing. It is the treatment of less favourable treatment as a mischief in itself rather than as an indicator of the protected characteristic being a reason for action that creates the problem. Because the two stage test asks Tribunals to start by identifying LFT and then to move to ask what the respondent’s reason for action was, it appears to suggest that identifying LFT is somehow unrelated to the process of identifying the respondent’s reasons. In other words, rather than emphasising its importance as an indicator of the respondent\’s reasons, it camouflages it. The artificiality of this carving out is demonstrated by what happens when there is no actual comparator. Applying the two stage test in cases using hypothetical comparators has caused tribunal brain implosions. The truth is that cases like Shamoon which suggest you can approach the test backwards – looking at the second stage first, really implicitly acknowledge the absurd artificiality of having a two stage process at all. Rather than being allowed to say:
(1) Absent evidence of LFT, we were nevertheless able to find evidence that convinced us that the respondent treated the employee in the way he did because of his race; and
(2) The treatment was adverse.
The Tribunal has to take the additional and pointless step of inferring that the adverse treatment would not have been meted out to an entirely non-existent comparator.
I’ll come on to the consequences for the burden of proof of this artificial two stage analysis once I’ve done my telecon!
So … for part 2.
One of the most striking things about the s. 136 burden of proof is how no-one ever seems to be able to explain it without first describing King. King is much easier to understand.
Put all law aside for a minute. Assume that you are told that an employer has two employees – one English and one Australian. He has handed out an annual bonus but has paid the Australian half what the English employee received. You are asked to find out whether the difference in nationality is the reason for the difference in treatment. What is the first thing you would do? Almost any person unburdened by a legal qualification would ask the employer to explain the difference. That is why King is easy to understand. It matches our intuitive approach.
The statutory formulation of the burden requires the employee to prove facts from which the court could decide, in the absence of any other explanation, that the employer has paid the Australian less because of her nationality. The provision does not say what kind of facts will do. Nothing on the face of the provision says a disparity in treatment is insufficient. In theory, therefore, it could simply match the first part of the King guidance. It would, in fact, improve on it since it would deal more comfortably with cases where no comparison with a colleague is possible. That is because it does not contain any explicit requirement that LFT be proven before an explanation is sought.
As Mrs Markleham points out, however, the courts have taken a different approach. A mere disparity in treatment and difference in “status” (by which is meant that the claimant has the protected characteristic whereas the comparator does not) is insufficient to shift the burden: “Something more” must be shown (Igen).
In our example of the underpaid Australian, the Law, as presently interpreted, stops you at the office door and says “before you ask the employer for his explanation, go away and find something else that suggests her nationality was part of the reason she got a lower bonus”. The only sensible response to this is to ask “like what?”, to which the courts have said, in essence, “we can’t tell you. You know, stuff. You’ll recognise it when you see it.” We have begun our drift away from an intuitive approach.
It gets worse. You come back and say: “This is proving difficult, can’t I have some sort of steer as to the explanation the employer is going to offer”. To which the Law replies: “Of course, in practice you always get to hear the explanation before you decide whether the burden has shifted and thus whether an explanation is called for (Laing). However, please try to remember these three principles as to how you take the explanation into account:
(1) You are to assume the explanation is inadequate until the burden has shifted (Hewage);
(2) If you think the explanation actually is inadequate you can take that into account in deciding whether the burden shifts (Millwood); and
(3) If the explanation is really really good (so that no other explanation makes sense) you can find the burden has not shifted (Hewage).”
It’s enough to make you cry. Why have we opted for this mind-wrenchingly complex approach? The short answer is the following words in the burden of proof provision:
Whereas, under King an inadequate explanation allowed a tribunal to find there was unlawful discrimination, under the new rules (simplifying somewhat) the tribunal is compelled to do so. Though they never say so out loud, the only sensible explanation for the present interpretation of the burden is a hostility to this compulsion on the part of the Courts and Tribunals. I share that hostility. Human behaviour is often inconsistent, unreasonable or even stupid. Once the burden has shifted, the employer has to prove a negative. Indeed, which is worse, he has to prove nationality was not part of his reasoning in circumstances where the courts and tribunals claim to be able to read parts of his sub-conscious that are closed to him. In practice, an employer armed only with an inadequate explanation for his actions is in real trouble. Under the King guidance, a sympathetic tribunal was free to determine that an explanation was inadequate but genuine. The courts and tribunals fear that that is harder now and the hammering into the formulation of the burden of a requirement that disparate treatment and difference of status is not enough is designed to replicate the effect of the discretion that existed under King. Personally, I think the discretion is valuable but at present it is secured only at the cost of complicating the Law to the point that one gets to watch some of the greatest judicial minds stumbling around it clutching at handholds.