It is common practice for tribunals to rule on liability first and then to determine remedy later. In my experience, this can often be more trouble than it is worth, especially in short unfair dismissal cases, particularly where the issues of a Polkey reduction or of contributory fault are likely to come into play. This was even more true during the ill-fated period of the statutory disciplinary and dismissal procedure in which a procedurally unfair dismissal could be rendered fair by the effect of section 98A(2) of the Employment Rights Act. The section required the consideration of exactly the same issues as Polkey but was, strictly speaking, a matter of liability rather than of remedy, with the highly unattractive result that tribunals would often have to conduct the same exercise twice in different hearings. In the case of London Waste v Scrivens, the tribunal apparently did not think that it was necessary to consider section 98A(2) or Polkey at all but, in any event, it was criticised by the Employment Appeal Tribunal for not having dealt with these matters at the liability hearing. Somewhat surprisingly, Judge Serota then went on to reach the conclusion that contributory fault should also have been considered at the same stage.

In practice, there are advantages in tribunals giving an indication of any likely deductions after the first hearing. This is when most of the relevant evidence will have just been heard and it can also focus the minds of the parties towards settlement. The problem is that, to the non-legally trained eye, it can look as though there has been a prejudgement of what is to come and, in any event, there is so little left to determine that it barely seems worth coming back just to deal with the issue of mitigation of loss. It is hard to avoid thinking that it would have been easier to have resolved the entire case in the first place.

If the burden of proof reverses in a discrimination case, the employer has to show a non-discriminatory reason for the treatment about which the complaint has been made. Sometimes this is straightforward but the decision of the Employment Appeal Tribunal in Arhin v Enfield Primary Care Trust provides an example of a situation where it can be more difficult. It was found that the respondent had simply made a mistake in not allowing the claimant, who was redundant, to compete for a job that was offered to her comparator. However, the tribunal had failed to adequately explain how it had reached this conclusion in circumstances in which it admitted that there was insufficient evidence of the scope of the new role.

In his judgment, Justice Langstaff also makes some general observations about cases in which mistake is argued as the reason for less favourable treatment. While, of course, it will occasionally be true that a mistake has been made, and this will always be a satisfactory reason which discharges the burden of proof, tribunals are entitled to be suspicious of such an explanation, especially where it has not been pleaded and emerges for the first time in oral evidence. It is easy to allege and so must be submitted to close scrutiny. This necessarily requires a tribunal to make detailed findings about how it happened.

The decision of the Court of Appeal can be found here. The reasoning on the ‘curing the breach’ point is pretty much as argued in the appeal, which is set out in my previous post. On the ‘reasonable range of responses’ point, the approach of the Employment Appeal Tribunal has largely been upheld, but with a caveat, per Lord Justice Sedley, that, “It is nevertheless arguable, I would accept, that reasonableness is one of the tools in the employment tribunal’s factual analysis kit for deciding whether there has been a fundamental breach. There are likely to be cases in which it is useful.” It seems to follow that, while a tribunal does not have to direct itself to the reasonable range of responses, it equally does not err if it takes this into account in finding that there is no fundamental breach, as in Claridge v Daler Rowney.

Nor can a dismissal that arises from such a situation ever be fair, at least if one applies the reasoning in paragraph 47 that an employer cannot possibly justify an act which it denies has occurred. As I have never known an employer to admit to a breach of the implied term of trust and confidence, this means that there is little scope for defending a case once a breach of the term has been found. I think that this is entirely right both as a matter of legal certainty and of industrial reality but it does demonstrate the importance of having an approach to testing for a breach which is both consistent and open to scrutiny.

The problem is that whether or not there is a constructive dismissal is a matter of fact and is difficult to overturn on appeal. The aim of Abbey National v Fairbrother was to prevent such findings on whimsical grounds and, now Fairbrother has been overturned, it is unclear how a similarly questionable decision could be challenged. I would argue that this can be done by allowing the range of reasonable responses test to be be used as part of the ‘toolkit’ of the appellate courts. Even if a tribunal has directed itself correctly, an analysis based on reasonableness may lead to a conclusion that its decision is perverse.

I went to the Court of Appeal on Monday and I heard some fascinating arguments. It seems that the question of whether trust and confidence can be restored after being destroyed could become moot as, according to Robin White, for Professor Buckland, there is no authority for the more general proposition that a repudiatory breach of contract can be cured by the wrongdoer, as opposed to being affirmed by the wronged party. The authorities cited by Employment Appeal Tribunal apparently say no more than that an anticipatory breach can be remedied. Also, while an employee is entitled to give an employer a chance to make amends, without affirming the contract, there is no obligation to accept what is done.

There is also a problem with the finding that the Tribunal wrongly applied a subjective test in deciding whether the breach had been cured. It is difficult to comment without having seen the judgment but the submission by Robin White was that the Tribunal repeatedly said that Professor Buckland was “entitled” to be dissatisfied with the handling of his complaint, and so they had applied an objective test after all.

In an earlier post, I questioned whether it was possible for a dismissal brought about by a breach of the implied term to ever be fair but Lord Carnwath floated an interesting possibility in his discussions with counsel. If a repudiatory breach is incapable of being cured then it follows that events after the breach, but before the breach is accepted, cannot be considered in deciding whether an employee is entitled to resign. This means that, if these events can be taken into account in deciding whether the dismissal is fair, as permitted according to Savoia v Chiltern Herb Farms, then the two questions may well produce quite different answers. However the parties agreed that the decision in Berriman v Delabole Slate, in which the reason for the dismissal is defined as the reason for the repudiatory conduct, is regarded as being more orthodox, and it will be interesting to see if this idea makes its way into the final judgment.

I can confirm that Buckland will be heard at the Court of Appeal tomorrow. Interestingly, there is also a permissions hearing for a cross-appeal by the university so it looks as though the range of reasonable responses point willhave to be considered to some extent although, it has to be said, Fairbrother does not appear to be particularly helpful on the facts of this case. There is an enormous difference between remedying a flawed process with an appeal and trying to remedy a breach of trust with a grievance. It is for this reason that I strongly believe that the appeal by Professor Buckland also ought to succeed.

In a sense, it could be argued that the ameliorative effect of an appeal is an example of a breach of the implied term being fixed and it is, I have to accept, the most plausible contender. Disciplinary action short of dismissal following a shoddy process could give rise to a breach (though many tribunals would find that it does not) which could then be remedied on appeal. This has the unfortunate effect that an employee has an incentive not to appeal but it appears to be a correct statement of the law. However, it is a relatively rare exception which reflects the fact that procedural breaches of trust and confidence are minor. There will also be cases where, if the problems with the process are severe, and are tainted by discrimination for example, they will be impossible to repair. The Employment Appeal Tribunal in this case was wrong to conclude it could decide the matter for itself without remitting it to a tribunal of fact.

Furthermore, a grievance does not work in the same way. Consider a case of sexual harassment where it is clear that a grievance finding that the acts took place, or even the dismissal of the offender, would not be sufficient to restore trust. If it were, it would have deleterious effect on our discrimination law. Now it might be said that, in Buckland, the breach was less serious than this but, if so, it is much more likely that there was no breach at all than that it was cured, and any challenge to the findings of the tribunal ought to have been mounted on that basis, applying Fairbrother if necessary (and if possible). The question of whether or not there is a breach of the implied term is a matter of fact and so, if the Fairbrother challenge fails, there is absolutely no reason to interfere with the findings of the tribunal.

In performing the analysis, unfair dismissal law is undoubtedly helpful, and even Judge Peter Clark is willing to rely upon Roberts v West Coast Trains, in which a dismissal disappeared when an appeal was upheld, as a useful analogy. However, more pertinent is the case of Taylor v OCS Group which warns against legalism and states that the overall effect of a procedure should be considered. Rather than asking whether there is a breach and then asking whether it has been fixed, the question is whether the process is adequate when considered as a whole. This is essentially the approach taken in Fairbrother. It is important to remember that, in several cases, it will be impossible for trust to be restored, at least not until years later. The alarming statement of the Employment Appeal Tribunal that the breach must have been fixed because there is nothing else that the employer could have done must be wrong.

It is now time to talk about the range of reasonable responses, a frequently misunderstood concept. As the Court of Appeal made extremely clear in HSBC v Madden, ”there will be cases where there is room for reasonable disagreement among reasonable employers as to whether dismissal for the particular misconduct is a reasonable or an unreasonable response.” To put it another way, the word ’reasonable’ means the same as within the range of reasonable responses, which is not a gloss on the literal reading of the statute. It is simply a reflection of the reality that there is frequently more than one reasonable solution. It would be a gloss on the statute to say that there is always only one solution in every case.

It might be objected that the word ‘reasonable’ appears elsewhere in employment legislation, such as in discrimination law, where reasonable adjustments must be made for disabled people and, previously, justification for indirect discrimination was required to be reasonable. There is no reasonable range in these cases but this is because an employer is expected to keep disadvantage to a minimum. Because there is just one factor to weigh in the balance, it is feasible to expect an employer to adopt the best solution. In an unfair dismissal situation, there are many considerations and it is less clear which rights of the employee are to be given the most weight. As a result, there will be multiple reasonable options.

I now want to return to Sharp. Having suggested in an earlier post that the second part of the test does not apply in the same way when the implied term is engaged, I nevertheless think that it is helpful to examine what it says about reasonableness. It states that whether there is a constructive dismissal depends upon whether there has been a fundamental breach of contract, not whether an employer has behaved unreasonably. In the case of an express breach, it follows that there can be a constructive dismissal even where an employer is found to have acted reasonably (Tapere is an example of this).

This is all very well but Sharp prefers the contract test because, otherwise, it suggests, constructive dismissal can be found on whimsical grounds. Therefore, in most cases, the test is higher than that of a reasonableness test. I would suggest that this will always be so for a breach of the implied term, not properly considered in Sharp. As ‘reasonable’ means within the range of reasonable responses and the action of an employer must be worse than unreasonable to make out a breach of trust and confidence, it is correct that a finding of such a breach arising from an action in the range of reasonable responses would necessarily be perverse. This is important because it is difficult to overturn findings on perversity grounds but it is necessary to establish a basis for challenging them to ensure consistency in the law.

It has been reported that the House of Lords has defeated several government amendments narrowing the exemption from sexual orientation discrimination currently allowed to religious organisations. This is slightly misleading because one form of the wording rejected appears in the original draft of the bill and the only amendment since was at the request of church leaders, to remove any doubt that ministers will be covered by the exemption. This followed a legal opinion by John Bowers QC for the Christian Institute. What is particularly unfortunate is the suggestion in some media sources that the rejection of the government proposals reflects a concern based on this opinion, even though the danger of the bill applying to ministers had been removed. In fact, it is opposed because the churches (no other religions have pushed the issue) want the right to be able to discriminate against lay staff who are homosexual.

For many people, this raises all sorts of issues including, unsurprisingly, the presence of the Lords Spiritual, who undoubtedly swung the result. Fortunately, however, for those who believe in equality, democracy and secularism, the intervention will prove to be entirely futile. The approved amendment restores the wording of the Sexual Orientation Regulations, which have been declared by the European Commission, in a recent reasoned opinion, to be inconsistent with the Employment Equality Framework Directive. This leads to the threat of action against the United Kingdom in the European Court of Justice and, therefore, the likelihood that the Parliament Act will used. It is also probable that, as it is a matter of compliance with European law, the government can make secondary legislation to force it through.

Even if the amendment were to survive, it would not last long against a challenge in a higher court. The government has stated that its proposed wording does not change the legal position and this is basically correct. The Sexual Orientation Regulations would have to be interpreted consistently with the Directive and the same will apply equally to the Equality Bill. The fact that the amendment is clearly intended to evade the Directive will not help and, if anything, it will be counter-productive. Churches may be afraid that they will be subjected to a few months of tricky litigation at a local employment tribunal but they are now faced with the possibility of dragging it out for several years, including a visit to Luxembourg.

In Buckland, Judge Clark sets out a four stage test. Firstly, is there a breach of trust and confidence? Secondly, does acceptance of that breach entitle an employee to resign, using the authority of Sharp v Western Excavating? Thirdly, is the dismissal for a potentially fair reason? Fourthly, is the dismissal fair? I want to concentrate on the second stage and it is necessary to go straight to the case of Woods.

My first observation is that Woods expresses some reservations about Sharp, which it is said has led employers to act totally unreasonably while stopping just short of a breach of contract. Justice Browne-Wilkinson feels that the implied term of trust and confidence can be used to fill this gap. More strikingly, in Malik, the House of Lords does not cite Sharp at all, which suggests that, if the two tests come into conflict, then Woods should prevail, notwithstanding the fact that Sharp is the decision of a higher court.

And there may be a conflict. Whereas Sharp emphasises that, once a breach is found, a tribunal must go on to consider whether an employee is entitled to resign, Woods says that a breach of the implied term is necessarily repudiatory, a point which has been followed in Morrow v Safeway Stores. However, “the Tribunals’ function is to look at the employer’s conduct as a whole and determine whether it is such that its effect, judged reasonably and sensibly, is such that the employee cannot be expected to put up with it”. In other words, it is a requirement of the test to consider whether an employee is entitled to resign but this is part of the test of whether there is a breach in the first place, not whether the breach is repudiatory. Buckland is wrong to divide this into two stages. Woods, as followed in Malik, creates an entirely new kind of constructive dismissal, to which principles different from those in Sharp will apply.

The reason is obvious. It is utterly artificial to say that an act breaches trust and confidence but does not entitle an employee to resign. If an employee is not entitled to resign then trust and confidence has not been breached. A stage-like approach is appropriate for a contractual analysis based on the breach of an express term but not for a breach of the implied term. I would go further and say that stages three and four become equally nonsensical when the implied term has been breached. Clearly it is hard to see how a common law concept can incorporate the statutory test of fairness but, fortunately, it is unnecessary to suggest that it does. This is because, as I will go on to argue, the bar for breaching the implied term is higher than for failing the statutory test so, it the latter is met, there can be no breach.

The decision of the Employment Appeal Tribunal in Allen v Jobcentre Plus raises an interesting question. It is common practice for tribunals, when dealing with a point of whether there is a continuing act which brings claims that would otherwise be out of time within their jurisdiction, to reserve the matter to the final hearing without a pre-hearing review. The reason is that claimants have to do no more than show a prima facie case at the preliminary stage, with the effect that much of the evidence is likely to be heard at the pre-hearing review, without disposing of any issues. It must then be heard all over again.

In this instance, unusually, it had been possible to conclude that no prima facie case had been made out, because the acts in question had been carried out by different colleagues and were of a completely different character from one another. This decision demonstrates that an application by a respondent for a pre-hearing review may not always be entirely futile, at least if the facts are equally favourable. Judge Birtles quite rightly rejects a submission that a tribunal is always obliged to defer the matter to a full hearing. The purpose of doing so is to avoid wasting time, not to prevent injustice to claimants.

To start my review of constructive dismissal law, I must deal with a point of which, in my opinion, far too much has been made. In the only House of Lords decision in this area, Malik v BCCI, Lord Steyn appears to state that the test of whether there is a breach of the implied term of trust and confidence is whether an employer has, without reasonable or proper cause, conducted itself in a manner “calculated and likely” to destroy or seriously damage the relationship of confidence and trust between employer and employee. This is different from that formulated by Justice Browne-Wilkinson (as he then was) in the authoritative case of Woods v W M Car Services, which says “calculated or likely”. It would be all too easy to get the impression that there is a major conflict in the authorities that needs to be resolved.

However, there is not. So that people do not get the impression that I am giving Judge Peter Clark too much of a hard time about Buckland, I wish to make it clear that I consider his judgment in Baldwin v Brighton & Hove Council to be an example of judicial reasoning at its best. He shows that the test in Woods had been cited in numerous cases before Malik, always with the word ‘or’. Further, the passage in which Lord Steyn uses the word ‘and’ purports to be a direct citation from Woods, which suggests that it is a transcription error. Judge Clark could also have mentioned that Lord Steyn asserts that he is stating the test again, having used ‘or’ previously. There is clearly no intention to amend the legal test.

Unfortunately, this is not quite the end of the story. One consequence of Baldwin is that, by describing the test as disjunctive, it could be understood as saying that an intention for an act to destroy trust and confidence could be sufficient, even if it is not likely to do so. Reading the decision in context, it is doubtful that this is what it means but the point is an arguable one. More worryingly, the reasonable range of responses cases, Abbey National v Fairbrother and Claridge v Daler Rowney, both adopt the formulation “calculated” with no mention of “likely”. A brilliant solution is to be found in a location which makes it less than obiter, so I shall leave readers to search for it for themselves. It relies on the case of Norweb v Dixon which states that ‘calculated’ means the same thing as ’likely’ and it does not mean ‘intended’. Therefore, “calculated or likely”, “calculated and likely” and “calculated” all mean exactly the same thing - likely - which, by the way, is how Lord Nicholls formulates the test in Malik. Case closed.

Although intention is not required to breach trust and confidence, it does not follow that the bar is low. Despite the word games that have gone on over the years, there has been no real analysis of the word ‘trust’. In everyday use, it can be broken unintentionally but it is hard to break and even harder to fix.

The government recently published draft regulations aimed at restricting the use of damages-based contingency fee agreements in which the representative receives a proportion of the damages awarded to the client. They are unlawful in the United Kingdom (although not in the United States of America) for court proceedings but are quite common in Employment Tribunal proceedings, as a reflection the fact that awards are usually relatively small and clients rarely recover their legal costs. Few people question that some regulation of the terms of such agreements is necessary but there has been criticism in the Law Gazette of the cap of 25% that is planned to be imposed upon the proportion that can be agreed.

At Employment Law Advocates, we largely support the concerns which are expressed in that article. Although, regrettably, we did not take part in the original government consultation, we have recently submitted comments about the draft regulations to the Ministry of Justice. They can be uploaded here.

I made quite a number of posts last year about the case of Bournemouth University v Buckland at the Employment Appeal Tribunal. I am not a fan of the decision which, among other things, says that it is wrong to introduce a range of reasonable responses test into the question of whether there has been a breach of the implied term of trust and confidence. This conclusion has found favour in some quarters but Judge Peter Clark also makes two more controversial claims, firstly, that constructive dismissals founded upon a breach of the implied term can sometimes be fair, and, secondly, that such a breach can be remedied fairly easily by, for example, a favourable finding in a grievance investigation. It is these two issues that are the subject of an appeal to the Court of Appeal, which is due to be heard on 8th or 9th February, and it is possible that the Court of Appeal will not have to deal with the range of reasonable responses point at all. However, in my view, it is now time for constructive dismissal to be re-examined in order to try to create a coherent whole and so, in reality, the contentions in Buckland cannot be divorced from one another. In the coming weeks, I shall outline my position on the subject.

I want to start by setting out the scope of the discussion. Judge Clark says of the range of reasonable responses that either, “it adds nothing, in which case it is superfluous; or it alters the House of Lords test” in Malik v BCCI. I wish to make clear from the outset that I have no problem with the suggestion that it adds nothing or very little to the Malik test. Indeed, I would go so far as to say that it follows logically from Malik that an act that is within the range of reasonable responses of an employer cannot possibly breach the implied term. I also note the position taken by the Employment Appeal Tribunal in several of its decisions, which are largely in agreement with the rejection of the range of reasonable responses test in Buckland but state that cases in which a reasonable act breaches the implied term will be exceptional. The latest example is the judgment of Judge Serota in NSPCC v Dear, handed down today. I cannot personally conceive of a real situation in which such an exception would ever apply but the importance of these authorities is their confirmation that there is a high bar for a claimant to reach.

Therefore, the focus on my criticism of Buckland is that I am far from convinced that it does accept the existence of quite such a high bar. The strongest evidence of this emerges from those conclusions that have led Professor Buckland to take the case to the Court of Appeal. By suggesting that a breach of the implied term of trust and confidence can often be fair and by suggesting that it is easy to cure, it is treating such a breach as a trifle and trivialising the importance of trust between an employer and an employee. That is the issue which is going to be at the core of my argument over the next four weeks.

The decision of the Scottish Employment Appeal Tribunal in Young v Hexion Speciality Chemicals deals with a short but important point. A claim for unfair dismissal had been submitted in which the date of termination of the employment had been provided but not the start date. It was rejected on the basis that it could not be determined whether the claimant had sufficient service to bring a claim. He applied for a review, supplying his start date, but it was not allowed. As the Lady Smith notes, this was wrong because the Employment Tribunal Rules specify that a claim can only be rejected if it is clear that there is no jurisdiction, not if it is unclear that there is. It could not have been rejected on the basis that “details of the claim” had not been provided, as included in Rule 1(4)(e), because this requires no more than specifying sufficient information as to whether the claim is of a type for which there is jurisdiction (Grimmer v KLM Cityhopper). Tribunals can hear some unfair dismissal claims and so that is enough.

In my experience, tribunals vary wildly in the approach they take to rejecting claims or parts of claims. A particular issue is money claims that can be pleaded as unauthorised deductions of wages or breach of contract. Parties are not asked to specify which and yet claims are sometimes rejected on the basis that there is no jurisdiction to hear a wages claim. The rationale for taking a hard line is presumably a desire to cut down on the number of claims but, in practice, there are so often appeals or applications to review that this is a false economy of time. I note in passing that those tribunals with a more liberal policy seem to be currently experiencing less backlog than those taking a more draconian approach.

Happy New Year everyone. For my first post of 2010, I would like to make my selection for the most important unreported case of last year. There are always significant cases that slip under the radar of the law reporters, especially if their practical usefulness outweighs their academic interest or where novel points arise about an issue peripheral to the main case. Chondol v Liverpool City Council qualifies on both counts. It generated a frisson of excitement in the media because it is yet another case that distinguishes between treatment on the grounds of a religious belief and treatment on the grounds of acts related to that belief. There was no need to report it for that reason but what is really interesting is what it says about unfair dismissal where a person is dismissed after multiple charges of misconduct.

Such cases are surprisingly common, often because evidence of earlier acts of misconduct will emerge during the investigation of another charge, but also because some employers will reopen issues which had been resolved or overlooked in order to try to justify a dismissal, especially where the evidence about the main charge is rather shaky. It may have been concerns about the latter which led the House of Lords in to conclude, in Smith v Glasgow District Council, that a dismissal is necessarily unfair if an employer has failed to establish the truth of any one of the charges that forms an important part of the reason for dismissal. This harsh conclusion can only be avoided by an express finding that the charge is not central to the dismissal. Of course, Polkey and contributory fault could reduce the compensation.

Chondol ameliorates the effect of Smith somewhat by stating that a failure to establish the truth of even a key allegation is not always fatal, on the basis that the overall reasonableness of the procedure is the issue, rather than the reasonableness of any individual element. This is established from cases such as Taylor v OCS Group although this was not cited by Justice Underhill in his decision. Indeed, it might be said that the result in Chondol follows so clearly from the decision in Taylor that there was no need to report it at all. I disagree. The problem with Taylor is that it leaves unclear which procedural flaws will be forgiven and which will not and, in the light of Smith in particular, it might be assumed that a failure to establish the truth of a charge still would not be. Chondol provides welcome clarification.

In my last dark law post, I discussed the scepticism of tribunals regarding discrimination claims about acts which do not appear to be discriminatory on their face. It might be thought that the situation would have changed since it has become possible for the burden of proof to be reversed but, in fact, this has not really happened at all, for largely the same reasons as those that were outlined in the earlier post.

A reluctance to apply the reverse burden of proof strictly is understandable, particularly in cases where there is no jurisdiction to consider unfair dismissal. An employer who has dismissed an employee with less than a year of service for an entirely arbitrary reason is unlikely to want to go into detail about the thought processes involved, but it would be entirely unjust to make a finding of discrimination in these circumstances. In practice, this means that tribunals will often accept dubious reasons, or find entirely unpleaded reasons of their own, for dismissal in order to ensure that the burden is discharged where it is fair to do so. This tendency is only increased by the very unfortunate tactic of pleading discrimination where a complaint is really one of unfair dismissal, as a way of trying to evade the lack of jurisdiction.

However, employers should still remain wary of cases like these. They are rarely won by the claimants but are sometimes lost by the respondents who are afraid of scare stories about tribunals who find in favour of employees for obscure reasons, and therefore decide that it is wise to lie about embarrassing but non-discriminatory motives. Such scare stories are completely unfounded but tribunals have been known to allow the reverse burden of proof to operate when told such lies, even if there is no direct evidence of any discrimination. This seems harsh but actually makes a great deal of sense. If a person is lying about the basis for a decision, it is clearly possible that discrimination is part of the reason and, if this did not result in an adverse finding, the practice of lying to cover up unlawful acts will no doubt be encouraged. Parties without legal experience tend to underestimate the ability of tribunals to work out whether a witness is telling the truth and the advice to them is simple. Tell the truth, however silly it will make you feel and, if there has not been any discrimination, you will win. Lie and you will lose.

The Court of Appeal dealt with some complicated issues of European law in Duncombe v Department for Children, Schools and Families. It was not a good week for the Department, which also intervened unsuccessfully in JFS. Mr. Duncombe was employed to teach at a European School in Germany but was dismissed when his fixed-term contract came to an end after nine years, contrary to the Fixed-Term Workers Directive. He did not bring a claim in Germany, because he knew that the Department would rely upon state immunity to defeat the claim, and so he had no choice but to bring his claim in the UK.

He was allowed to pursue wrongful dismissal by the Employment Appeal Tribunal, following the earlier decision by President Elias in Bleuse v MBT Transport. It was held that a claim under the Working Time Directive, concerning work carried out in Austria, could be pursued in an Employment Tribunal because the Directive is directly effective and, otherwise, there would be no effective remedy for a breach by a UK-based employer. In the appeal, Lord Justice Mummery notes that Bleuse is not necessary for the wrongful dismissal claim. The contract was agreed to be subject to the jurisdiction of the English courts and, although the conversion of a contract from a fixed-term one to an permanent one depends on the operation of European law, it operates upon the contract itself and so no jurisdictional issue will arise.

However, the unfair dismissal claim is a different matter. Jurisdiction is governed by the case of Lawson v Serco which depends, subject to certain very limited exceptions, on the employment being based in the UK. None of the exceptions apply to this particular case. While the Fixed-Term Workers Directive is found by the Court of Appeal to be directly effective, there is not a legally-binding right in European law to be protected against unfair dismissal itself. Nevertheless, Lord Justice Mummery concludes that it is necessary to extend Lawson v Serco so that there is an effective remedy for a breach of the Directive.

It is difficult to argue with this result on the facts but the decision can hardly be said to clarify the law. The problem is that, in Bleuse, it is far from clear that the case could not have been brought in Austria and so there is no obvious reason why an English tribunal had to accept jurisdiction in order to provide an effective remedy. The argument that this is the case in Duncombe is much more compelling but this has the result that the correctness of the wider ratio in Bleuse is not considered. Perhaps the strongest conclusion to be drawn from it all is that there is a difficulty with Lawson v Serco, which really ought to give jurisdiction in a case of unfair dismissal where the UK government is the employer and can rely upon state immunity. Doubtless the Department will appeal and this will allow the point to be explored.

The judgment of the Supreme Court in this case has been all over the newspapers. I went to see it being handed down yesterday and I was pleased both that the carpet looks much better in real life than in photographs and with the decision itself, in which thirty years of direct discrimination law narrowly avoided being overturned, by a wafer-thin majority of 5 to 4. As noted in this post, it was the authority of James v Eastleigh Borough Council that was at stake and, although at least one of them purported to follow it, the Justices in the minority effectively suggested that the principle had been watered down by Nagarajan. The position of the majority is the orthodox one (in a non-religious sense) and is supported by recent decisions of the Employment Appeal Tribunal, including this one. None of them were actually cited in the speeches, which is surprising because I cannot believe they were not cited in submissions.

A less settled area of law, which gave rise to some ingenious arguments by Lord Pannick, on behalf of the school, concerns the case of Mandla v Lee, which provides the legal definition of an ethnic group. The Court of Appeal had suggested that an ethnic group had to be defined on biological grounds alone and, therefore, a convert to Sikhism (or Judaism) could not claim race discrimination. The House of Lords overturned this decision, stating that there can be a cultural element to an ethnic group. The test is uncontroversially an objective one and, as Lord Pannick pointed out, the effect is that M, the boy in the case, is Jewish like his comparator, who would have been given a place. As Orthodox Jews were not found to constitute a separate ethnic group, it was suggested that there was no race discrimination.

This approach initially appears to be compelling and is not an easy one to rebut but the Justices found a number of problems with it. The first issue is identified by Lord Kerr and Baroness Hale and concerns the fact that, although M and his mother are both Jewish, according to the Mandla test, they are not exclusively Jewish and are also part Italian. The reason for the less favourable treatment was the part that is not Jewish. This is persuasive because, otherwise, people of mixed race would not be protected.

However, it is Lord Mance, in the most detailed judgment, who really gets to grips with the issue. He points out that the intention of Mandla was to widen the definition of an ethnic group, rather than to narrow it, which means that biology is still part of the test, as well as culture. Although the boy and his comparator are members of the same ethnic group culturally, they are members of different groups defined in biological terms. Furthermore, following the case of English v Thomas Sanderson, people can be discriminated against on the grounds of their perceived membership of an ethnic group, even if the perception is objectively incorrect. The school cannot rely on Mandla to say that M is Jewish so there is no discrimination if the reason for treating him less favourably is that they do not accept him as Jewish.

An employment tribunal does not have the jurisdiction to hear a claim if it has been subject to an agreement which complies with section 203 of the Employment Rights Act, or with a similar provision in another statute. It does have the power to determine whether an agreement properly complies with the terms of the section but it had not been clear if it could rule on the question of whether an agreement is void for some other reason, such as misrepresentation, or whether this would have to be decided by a court. In Industrious v Vincent, the Employment Appeal Tribunal conducted a comprehensive review of the case law and determined that it could, disapproving several of its own previous decisions, including Byrnell v British Telecommunications. Justice Silber notes that the judgment in Byrnell does not refer to the two most relevant authorities, one of which had been expressly approved by the Court of Appeal.

This result could have much wider significance because Byrnell had also been the only authority for the proposition that a tribunal cannot reject an agreement where there has been a repudiatory breach of contract, such as the fact that a settlement has not been paid. Indeed, it is difficult to see a distinction between a rescission of a contract because of misrepresentation and one resulting from a repudiation. It would therefore seem to follow that precisely such an argument would now be expected to succeed.

In Byrnell, Judge Ansell makes a brief reference to authorities, apparently relied upon by the tribunal, which state that a compromise agreement must be enforced by a court unless it was agreed before the termination of employment, because otherwise it would not fall within the scope of the Extension of Jurisdiction Order. This is correct but, in the above situation, the claimant is not seeking to enforce the agreement in order to obtain damages for breach of contract. She is seeking to set it aside so that she can bring a straightforward tribunal claim for, say, unfair dismissal. These cases are simply irrelevant.

The consequence of all this is that employers should now be particularly wary of failing to comply with the terms of compromise agreements as they may find themselves in the tribunal rather than merely facing enforcement proceedings. On the other hand, in contrast to settlements where a claim is stayed on agreed terms, a new claim must be issued and normal time limits will apply. Therefore, employees will need to keep an eye on the deadline if they want to retain the option to elect for the tribunal route.

The single most common mistake made by inexperienced advocates and litigants in person in the Employment Tribunal is to assume that it is easy to demonstrate that an negative act by an employer was done on discriminatory grounds. For example, many think that evidence of having been bullied is sufficient even where there is no basis for suggesting that the bullying had anything to do with being a member of one of the protected groups. There is a similar problem in constructive dismissal cases where it is sometimes believed that a resignation as a result of stress is enough and that it does not matter which act of the employer, if any, actually caused the stress. However, (most) tribunals are very strict about not exceeding their jurisdiction and they will not allow what really amount to personal injury cases to succeed merely because they disapprove of a respondent or feel sorry for a claimant.

When, as often occurs, there is a discrimination claim concerning numerous incidents, the tribunal will want to isolate those for which there is a discriminatory motive from those for which there is not. This can sometimes be fatal to the case. For example, to make out harassment, an employee is required to show either a violation of his dignity or the creation of an adverse environment, and one or two minor incidents may well not have this effect. Even where such an environment is created, if it is largely the result of events for which there is no jurisdiction, it will not assist if off-colour comments contributed to it if they would be insufficient on their own. Similarly, tribunals will rarely find constructive dismissal in cases of overwork and a claim of this nature will not be strengthened by a few mildly aggressive acts by an employer. Regrettably, although this reasoning is entirely justified, tribunals are reluctant to spell it out explicitly for fear of an appeal. Instead, they have a tendency to make adverse findings of fact which can be distressing for claimants and means that they often fail to understand why they have lost.

There are some more experienced representatives who understand the above problems and realise that it is necessary to selectively plead those points that will most assist the case. This is very sensible but it will not prevent tribunals from considering whether there are some unpleaded issues which may have actually been the cause of the stress felt by the claimant. It is even possible that tribunals have become more sceptical to counter the development of more sophisticated litigation tactics and that this could be as detrimental to strong claims as to weak ones. Therefore, it is important to be able to lead evidence to show which particular acts were the most upsetting for the claimant. It goes without saying that it will help a great deal if there are documents to prove they were complained about at the time.

Consider the followings two scenarios: a black person claims race discrimination on the grounds that she has wrongly been assumed to be mentally ill; and a white person claims race discrimination on the grounds that she has wrongly been assumed to be mentally ill. Even without knowing any of the facts, the first case seems immediately more plausible than the second and the reason for this is that there is a well-known and widespread tendency to overrefer and overdiagnose black people with mental illness.

The next question is whether this stereotype is a sufficent to found a tribunal claim. In the case of Johnson v Coopers Lane Primary School, the Employment Appeal Tribunal concludes that it is not, for the persuasive reason that it is unjust to make a finding of discrimination against an employer without some evidence relating specifically to the behaviour of that particular employer, rather than just people in general. This echoes some very similar reasoning by Justice Underhill in the decision of B and C v A.

However, Judge Ansell appears to go further than this. He agrees, in paragraph 40, with a submission of the respondent that tribunals should not take judicial notice of the existence of stereotypes, even in cases where (unlike this one) there is statistical evidence to support them. It is true that such evidence is never determinative on its own but surely it must be right that it can be used to bolster a case where other reasons to suspect discrimination exist. A fundamental rule of evidence is that the more unlikely an event, the stronger evidence is required to show that it took place. A corollary of this is that events which can be shown to be more likely than others will require less compelling evidence in their support.

I am not even certain that statistical evidence would be strictly necessary in a case like this. There is a convention that a tribunal panel hearing a race discrimination case should include at least one member with special experience of race relations. Such a person would undoubtedly be aware of the stereotype explored here, and also numerous others, and would therefore be in a position to be alive to their use.

The judgment of the Employment Appeal Tribunal in McFarlane v Relate Avon is another in a long line of authorities in which discrimination on the grounds of religion and sexual orientation have appeared to come into conflict. In such cases, there is often unhelpful talk about one set of rights ‘trumping’ another but this just misses the point. In McFarlane, Justice Underhill followed the decision of his predecessor as President, Justice Elias, in London Borough of Islington v Ladele and was undoubtedly correct to do so.

The reason why ‘trumping’ is the wrong word is that there are certainly plenty of situations in which Christians and other religious people are subjected to criticism that goes far beyond what is necessary to ensure equality for gay men and lesbians. In such cases, a claim for religious discrimination ought to succeed, even if the disciplinary action which provides the background is justified. There is perhaps an example here although, as noted at paragraph 12, it was not appealed. This is the claim of harassment.

It does not appear to be disputed that colleagues of the claimant circulated a letter describing him as ‘homophobic’. This account is supported by at least one source unlikely to favour his case. The next question is why this suggestion was made and it simply does not follow that, because his dismissal was on the grounds of his actions rather than his beliefs, the same can also be said of this remark. To call someone homophobic is an attack on the beliefs at least as much as the actions of that person, if not more so. Justice Underhill himself noted in Richmond Pharmacology v Dhaliwal that some examples of harassment do not require a comparator and I would submit that this falls squarely into that category.

A libel court would surely agree with me that the label ‘homophobic’ is more than merely a criticism of behaviour but, if this seems extreme, consider a scenario in which the word is preceded or followed by a stream of unpleasant expletives. I do hope that this case is not regarded as excusing employers from the need to be more careful about their use of intemperate language in these situations because, if so, then the argument that one set of rights has ‘trumped’ another may begin to seem like it is a valid one.

The Employment Appeal Tribunal decision in Yorkshire Maintenance Company v Farr concerned penalty clauses in employment contracts, an area of law in which there are surprisingly few authorities. The respondent had attempted to rely upon a clause stating that wages could be deducted in the event that timesheets were not signed by the client. The tribunal did not accept this, not expressly because the clause was a penalty, but because it was held to require timely deductions and so the right to do so had lapsed. Judge Pugsley rejected this approach and felt that more exploration of the facts was necessary.

The argument was that a deduction was justified due to the frequency with which clients refuse to pay if a timesheet has not been authorised by them. This seems a surprising claim on the facts, the claimant having stated that he had never previously had worksheets signed, and the tribunal having appeared to find in his favour on this point. It is hard to deny that it is reasonable to make such a deduction if a client does not, in fact, pay, and I would be willing to accept that a clause which realistically reflects the chance of a client failing to pay would also be justified. However, an automatic deduction of wages for the entire period covered by a timesheet surely does not constitute a genuine pre-estimate of the loss.

An article in the Guardian reports on a fascinating case which began at the Manchester Employment Tribunal earlier in the week. It represents part of the fallout from the construction industry blacklist which I discussed in this post. Now many of the people affected (twenty three of them) have brought claims against the companies who put their names on the list in the first place. As I said at the time, there is no obvious remedy in employment law but there are some interesting submissions to be made.

Paragraph 355.02 of Division NI of Harvey on Industrial Relations sets out the basic structure of the argument. Workers are protected from being subjected to detriments on grounds related to trade union activities. From the case of Woodward v Abbey National, providing a negative reference is capable of being a detriment even where an employment relationship has come to an end and so the same must surely apply to the act of adding a name to a blacklist. However, there are two further problems in this case. Firstly, the claims are out of time by several years because the clock starts running from the date of the detriment. Secondly, there has only been protection for workers against post-employment detriments since the Employment Relations Act 2004 but many of the claims in this case precede the date of it coming into force. Furthermore, some of the workers were victimised by end users with whom they had no contract, and so they would seem not to qualify as workers, even under the amended law.

The time point is easier to deal with. There is considerable assistance from the decision of the Court of Appeal in London Borough of Southwark v Afolabi, in which a race discrimination claim was allowed to proceed when the claimant did not discover the relevant acts until nine years later. Afolabi concerned the extension of time on a just and equitable basis while, in this case, it will have to be shown that it was not reasonably practicable to bring the claims in time, but the point is certainly a persuasive one.

Moving on to the extent of the protection, before the changes, only employees were protected and only against action short of dismissal. When the amendments were made, the transitional provisions made it clear that they did not apply to detriments taking place before 1st October 2004. However, it is obvious that the main reason for the changes was criticism that the old law did not provide sufficient protection for freedom of association to comply with Article 11 of the European Convention of Human Rights. If so, perhaps the interpretative obligations in section 3 of the Human Rights Act will allow the unamended statute to be read in accordance with the current one and the transitional provisions to be disregarded.

This leaves the claimants who are suing end users in the absence of a contract. For them to succeed, either the definition of a worker would have to be extended further than it is in any other legislation or it would have to be found that human rights make it necessary to imply a contract to provide personal services, distinguishing James v Greenwich. Still, it would be rather surprising if they were not entitled to some sort of protection and so it is entirely conceivable that this case could end up in Strasbourg.

The Court of Appeal decision in Royal Bank of Scotland v Allen concerns the provision of goods and services under the Disability Discrimination Act. The case turned largely on its facts and was decided using the existing authorities about the meaning of a ‘reasonable alternative method’ of providing the service. However, there was one point which could later become relevant in an employment context.

Section 21(2) of the Act says that a service provider must take such steps, by way of adjustments, as it is reasonable “for him to have to take”. This awkward wording appears deliberate but the judgment confirms that the words “have to” add nothing of significance to the test. Neither Lord Justice Dyson nor Lord Justice Wall could think of any situation in which it would be reasonable to take certain steps but it would not be reasonable to have to take them. It is hard to argue with this and so the outcome reflects badly on the draughtsman of the statute rather than the Lord Justices of Appeal. In any event, identical statutory wording appears in section 4A(1) of the Act, which concerns the adjustments that should be made by an employer, and so it is extremely likely that this section will be interpreted in the same way.

The decision in Orr v Milton Keynes Council gets to the heart of a very difficult area of discrimination law, usually described as causation, but this is itself a problematic term, as the case demonstrates. The claimant suffered harassment for which his employer was vicariously liable, but was dismissed fairly for his rude and abusive response to it. It was held that the dismissal was not tainted by discrimination notwithstanding the racist provocation because a white employee would also have been dismissed for a comparable act of misconduct. The problem is not an entirely new one and a similar result was reached by the Court of Appeal in Sidhu v Aerospace Composite Technology, not mentioned in the judgment.

However, the analysis in Sidhu focused on the question of whether or not it was race discrimination to fail to take the context of the racist abuse into account as possible mitigating circumstances. It does not appear to have been suggested explicitly by anyone that the dismissal flowed from the tortious act of discrimination itself, even at the Employment Appeal Tribunal where the provocation was held to have occurred in the course of employment. Nevertheless, precisely such an outcome would result from the use of a ‘but for’ test, as indeed was unsuccessfully argued in Orr. Strict causation does not apply.

The problem is that, in practice, a ‘but for’ test is often used as a shorthand, an approach endorsed by no less an authority by Lord Goff in James v Eastleigh Borough Council, although it should be noted that he only said that it was appropriate in ‘most cases’. A more cautious note has been sounded in the subsequent decisions of the Law Lords, in a series of cases beginning with Nagarajan v London Regional Transport. In West Yorkshire Police v Khan, Lord Nicholls said, “Contrary to views sometimes stated, the third ingredient (’by reason that’) does not raise a question of causation as that expression is usually understood.” Instead, the reason why should be assessed subjectively, as a question of fact.

On the other hand, the approach in Nagarajan, which requires consideration of the state of mind of the alleged discriminator, can also be misinterpreted. It does not affect the principle in James v Eastleigh Borough Council that the motive for discrimination is irrelevant. Motive must be assessed only for the purposes of determining whether the treatment is less favourable, not whether the less favourable treatment can be justified in some way. The key lesson is that both the ‘but for’ approach and the ’state of mind’ approach can sometimes be convenient shorthands but they are not substitutes for a proper understanding and analysis of the case law. In particular, where an act is inherently race-neutral, such as a dismissal for aggressive behaviour, the Nagarajan approach is often likely to be more appropriate.

I was delighted to discover last week that a case I conducted, Tapere v South London and Maudsley NHS Trust, which I discussed in this post, has been reported in the Industrial Relations Law Reports at [2009] IRLR 972. It was analysed by Michael Rubenstein in his ‘Highlights’ section but I must admit that I found his reasoning rather odd.  He said, ‘However laudable such a purposive interpretation, the result of this reasoning can be seen as rather harsh in this case and could create practical difficulties more generally. A TUPE transferee is in no more advantageous position in respect of contractual rights than the transferor, but he is in no worse position either. If there had been no TUPE transfer, given the mobility clause, it is hard to see how the locations to which an employee could be sent would be frozen as at the time they entered the employment. If the locations operated by the PCT expanded, that would “increase the scope of the geographical area in which the employee could be required to work.” Why should the same principle not apply in respect of the locations operated by the transferee?‘

This is a weak analogy and, moreover, a similar suggestion could be made for most contractual terms protected by TUPE. For example, it is easy to see how a personal assistant role could evolve slowly over time into a general administrative role but it does not follow at all that a transferee could refuse to employ a personal assistant and offer her a job in administration instead, for three reasons. Firstly, in the former case, the changes have happened slowly over time, giving the employee time to adapt to them. Secondly, in the former case, there has clearly been agreement, either express or implied, to the changes. Finally, and most crucially, it is of enormous significance whether the changes are made for a reason related to the transfer. This is why the TUPE Regulations were originally enacted, after all.

As for practical difficulties, what no commentator seems to have realised, perhaps because it was not necessary to decide the case, is that there were, as a matter of fact, no practical difficulties in Tapere at all. The transferor and the transferee were both NHS Trusts with a close relationship. They frequently did work for one another (hence the service provision change) and they often allowed one another to use their property. This is how the claimant was able to remain at her old place of work for six months after the transfer and there would have been utterly no impediment for her to have continued to do so.

A harder situation which might arise under TUPE would be if there was a transfer between entities that did not share any property. However, the appropriate non-TUPE analogy would be with the closure of a place of work and its reopening some miles away. In such a case, best practice would be to make the employees who do not want to move redundant and any of them who unreasonably refused alternative employment would lose their entitlement to redundancy payments. An employer that forced employees to move outside the scope of their employment contracts would expect to face claims for constructive unfair dismissal. This approach has commendable flexibility because it encourages employers to offer a variety of alternatives, avoiding the very real problem that employees have no choice about the effects of TUPE. It also prevents TUPE from being misused as a mechanism to avoid redundancy payments.

A possible objection is that such a redundancy would be a dismissal related to the transfer to which my answer would be that, following Tapere, so would a unilateral variation of a mobility clause. However, TUPE allows such a dismissal if for an economic, technical or organisational reason entailing changes in the workforce. It is easy to see how a redundancy situation caused by employees not wishing to move location can entail such changes but it is more difficult to see how a change of location alone can do so.

Since Malcolm, the issue of knowledge of a disability has gained a great deal of importance because it is a requirement for a successful claim for a failure to make reasonable adjustments, which is now the most important head of claim. It has also become a requirement for disability-related discrimination in the unlikely event that the comparator problem is overcome. Although there is no statutory test in the latter case, the principles are likely to be similar to those that arise in the statutory test for knowledge in a reasonable adjustments case and so clarification of these principles is of considerable importance.

Clarification has not been achieved by the case of Department of Work and Pensions v Alam, in which the Scottish Employment Appeal Tribunal states that an employer is able to make out this defence if it is ignorant of either the disability or its effects and, therefore, it must be aware of both for a claim to succeed. This is contrary to the earlier authority of Eastern and Coastal Kent PCT v Grey, which said that ignorance of both is required for the defence, although this was admittedly not very helpful, being notable as the only Employment Appeal Tribunal judgment in the last year to be internally inconsistent. The decision by Justice Silber to remit the case when it was apparently quite unnecessary to do so was interpreted by Lady Smith as showing that it did not really mean what it appeared to say on its face.

However, the real problem with both of these judgments is that they are trying to carry out statutory interpretation in a vacuum. It is difficult to properly get a grip on the meaning of a word like ’and’ in a complex piece of legislation without trying to understand what is being said from a broader perspective and there are many issues which are completely ignored but which could be of considerable assistance. For example, the relevant section also includes the word ‘disability’ and, as I discussed in this post, this can mean both the underlying medical condition and the impairment caused by it. As this has the result of making the ‘disability’ and its effects equivalent for all purposes, neutralising the use of the word ‘and’, this suggests that Grey is correct, although for reasons different from those given in the decision.

On the other hand, I am sympathetic to the result of the Alam case, and of Ridout v T C Group, which it cites. This is another reason why I think that the word ’disability’ should refer only to the effects of the impairment and not at all to the diagnostic label. If so, there would be a defence if an employer knew about a medical condition but not its effects but no defence if, for example, it knew that an employee used a wheelchair without knowing the reason why. This approach is also consistent with the logic of constructive knowledge. A person who knows the effects of a disability knows everything necessary in order to make reasonable adjustments but a person who knows only the label does not. This issue must be urgently considered by a higher court to ensure the coherence of disability discrimination law.

Judgment has now also been given by the Employment Appeal Tribunal in this case, further to my earlier discussion of the submissions here. The outcome, that a passionate belief in climate change, accompanied by a moral imperative to do something about it, is capable of constituting a philosophical belief for the purpose of the Religion or Belief Regulations, is contrary to my original intuition but unsurprising, having had the benefit of hearing the arguments made on the day. The case of John Bowers QC that the test is narrower than the one for Article 9 of the Human Rights Act is attractive as a matter of policy, to avoid the opening of floodgates, but is impossible to reconcile with comments made by Justice Elias, as he then was, in Eweida, a view that is supported by the Framework Directive itself.

My original disquiet remains. Parliament did not originally legislate for political belief, and Justice Burton is in no doubt that the belief is political in this case, but now it seems that, not only will this lead to more claims, it will also be inconsistent in its effects, protecting some political beliefs but not others. Worse, it may protect those political beliefs which are dogmatically held more easily than those based on the wisdom of experience. Perhaps the solution is to extend protection to all political beliefs, as in Northern Ireland, although exemptions would presumably have to be made, for example, for editorial staff at the Guardian or the Daily Telegraph. However, legally speaking, the decision cannot be faulted.

The Court of Appeal judgment, which I first discussed in this post, has now been made available online.

When the Court of Appeal gave judgment in R (on the application of E) v The Governing Body of JFS, it was suggested to me that this might be a good topic for the blog. I did not take the opportunity to write about it then because it seemed to me that, like the vast majority of judicial decisions, it was entirely sound and consistent with authority and, being the contrarian that I am, this was uninteresting to me. I could not dispute the reasoning that to favour a boy of Progressive Jewish faith who is of Jewish blood over a boy of Progressive Jewish faith who is not of Jewish blood must necessarily be an example of direct discrimination on the grounds of ethnicity. As was established by James v Eastleigh Borough Council, a potentially benign motive for the discrimination will not prevent a claim from being made out.

However, I have learned that the case is being heard at the Supreme Court this week, exceptionally by a panel of nine judges, a measure of the high public importance which has been attached to it, as is the large number of interveners, many of whom have been involved since the case began in the High Court. Unfortunately, cases raising important issues of policy do not always encourage the consistent application of law and I am now worrying about an interventionist judgment with negative side effects.

It is worth examining the legislative history behind the key issues that arise. There has never been a justification defence for direct race discrimination but employment law has the concept of a Genuine Occupational Requirement, which allows practices that would otherwise be discriminatory if there is a strong enough reason for them. Rather curiously, when Parliament extended discrimination protection to other areas of life, it chose not to introduce a similar defence. In fact, the idea of a Genuine Service Requirement was discussed during the consultation for the Equality Bill but was subsequently rejected.

As a consequence, there are striking and alarming parallels with the infamous situation in Lewisham v Malcolm. Just like Malcolm, this is another case in which discrimination law is being used outside of the employment sphere, in circumstances in which employment law provides a sensible and proportionate defence but the other field in question does not. As in Malcolm, the temptation for the Justices will be to narrow the concept of direct discrimination in order to achieve a desired result, but in a way that will have knock-on effects for employment law, where the policy concerns in this case simply do not arise.

They should resist that temptation. If they feel strongly that a Genuine Service Requirement defence should have been introduced, they are entitled to say so in their judgments. Indeed, the Equality Bill is still passing through Parliament and it is not inconceivable that it will be changed as a result. Ultimately, however, introducing such a defence is a matter for our elected representatives. The Justices must not undermine the decision of James v Eastleigh Borough Council to resolve a narrow issue in a hard case.

I have one final thought. The central argument of the school is that its admission policy should be lawful because it is following the requirements of Jewish law. The implication is that English law should bend in order to accomodate it. Imagine the outrage if a similar argument were to be used about Sharia law.

According to Judge Peter Clark in the Employment Appeal Tribunal, in Allen v Murdoch, the statutory disciplinary and dismissal procedures apply to claims for redundancy payments and for pay in lieu of accrued holiday arising on termination and, therefore, there is no need for a grievance to be raised in respect of these matters. This result would have knock-on effects for time limits and would be highly significant for other cases were it not for the fact that the statutory procedures have been abolished.

Nevertheless, the reasoning in this case is worth closer examination. It is based on regulation 6(5) of the Dispute Resolution Regulations which state that the grievance procedures do not apply when the complaint is about having been dismissed. It is easy to see how a claim for a redundancy payment is a complaint about having been dismissed, as is a claim for notice pay, but it is harder to see how this is the case for accrued holiday pay, where surely the complaint is about not having received holiday pay.

This is a straightforward matter of causation, the issue being the reason for which the compensation is being claimed. Notice pay is compensation for having been wrongfully dismissed and redundancy pay is compensation for having been made redundant but holiday pay on termination is compensation for not having been able to take annual leave during the employment, and not for the dismissal itself. This is significant for tax purposes because there is a tax-free allowance of £30,000 for payments resulting from the termination of employment, with the effect that redundancy payments and payments in lieu of notice are often made gross. Some tribunals also award holiday pay gross but it is submitted that this is an error for the reasons stated above. Holiday pay is analogous to wages in that a tribunal can order it to be paid through payroll, with the appropriate deductions made, but cannot order it to be paid directly to an employee without deductions, as this would give the employee a windfall and would provide an incentive not to take any leave, completely contrary to the intention of the Working Time Regulations.

This is the week where everyone is talking about the BNP. The appearance by Nick Griffin on Question Time yesterday came shortly after he finally conceded the need to change a constitution which restricts membership on racial grounds. Frustratingly, the media coverage has not analysed the relevant law which is not as straightforward as it seems and so it is worth giving some thought to this tricky issue.

The Equality and Human Rights Commission, which used its powers to bring the case, provided the best summary of issues in its letter before claim. What is clear is that there is no express provision in the Race Relations Act for political parties. Rather, they have been found to qualify as associations covered by section 25 of the Act, in the House of Lords case of Watt (formerly Carter) v Ahsan. Incidentally, this was a case brought against the Labour Party and funded by the Commission for Racial Equality, now part of the EHRC, so the suggestion by Nick Griffin that his party has been singled out is simply untrue.

However, the Race Relations Act provides a special defence to a section 25 claim for associations set up for the benefit of people of a particular racial group. It seems that the BNP was attempting to bring itself within this exception. One of the less-trumpeted consequences of the new Equality Bill is that this defence is being abolished with the consequence that, even if the BNP had been successful in arguing the point, it would have been futile. This is why Nick Griffin is reported to have made the concession.

In case you think I have forgotten that this is an employment law blog, there is a point to be made in that area as well. Application forms for jobs with the BNP ask candidates to supply their membership numbers, implying that only party members can be employed. This has the effect that employees are also selected on the basis of their race, which is undoubtedly unlawful. The BNP would have had more chance of succeeding if it had tried to defend its membership rules without this additional problem. On the other hand, a requirement that employees be members is probably lawful so long as membership is not restricted on the grounds of race. Ironically, to defend this practice, the BNP would have to assert that membership does not constitute a philosophical belief for the purposes of the Religion or Belief Regulations, where previously they have said that it is. Still, the BNP seems unworried by incoherence.

The case of Autoclenz v Belcher is the last in a long series which caused a lot of confusion but resulted in the law being in much the same position it was in when the confusion began. The story starts with the case of Consistent Group v Kalwak in the Employment Appeal Tribunal in May 2007, in which President Elias stated the seemingly uncontroversial proposition that tribunals should examine the reality of the situation to see whether an employment relationship is perceived in the way that the contract suggests.

The tribunal in Autoclenz gave judgment On 1st March 2008, relying expressly upon Kalwak. On 28th April, President Elias employed similar reasoning in Protectacoat Firthglow v Szilagyi. Just a day later, everything was suddenly changed by the decision of the Court of Appeal in Kalwak in which the appeal was allowed, primarily on the basis of inadequate reasons, but Lord Justice Rimer also criticised the reasoning of the Employment Appeal Tribunal, and relied on the comments of Lord Diplock of Snook v London and West Riding Investment as authority for the proposition that, in order for there to be a finding that a contract is sham, there must be an intention to deceive a third party as to the nature of the relationship and it is insufficient that the parties understood it differently from the written contract.

On 4th June, Autoclenz reached the Employment Appeal Tribunal. Judge Peter Clark followed the Court of Appeal in Kalwak and allowed an appeal against the finding that the claimants were employees of the respondent, on the grounds that no Snook sham had been identified. However, he dismissed an appeal against the finding that they were workers. Autoclenz appealed this finding to the Court of Appeal which they are probably now regretting as it then allowed Belcher and his colleagues, who it appears were not intending to appeal, to put in a late cross-appeal regarding the finding that they were not employees.

The tide began to turn on 10th October 2008 with the case of Redrow Homes v Buckborough in the Employment Appeal Tribunal, in which Judge Burke noted that the House of Lords had used a different definition of a sham in Street v Mountford and this had come into employment law in Echo and Express Publications v Tanton. He suggested that there is no conflict between the the two approaches in Kalwak and that of Lord Justice Rimer was, in any event, obiter, given his other finding of inadequate reasons.

The extent to which all these decisions can really be reconciled is questionable. However, what is surely true is that the Street v Mountford definition is preferable to a narrow reading of the Snook definition. In the context of a situation in which an employer has a dominant bargaining position and the employee is being deceived as much as any third party, it is absurd to require that he or she also takes part in that deception. Indeed, it could be argued that tribunals should be much less sympathetic to workers who do willingly enter into sham contracts as they are likely to be tainted by illegality in most instances.

Orthodoxy has been restored in 2009, firstly when Szilagyi reached the Court of Appeal in February and now with the decision this week. Lady Justice Smith and Lord Justice Sedley gave judgment in both cases, agreeing with Judge Burke that there is no conflict but suggesting euphemistically that the formulation of Justice Elias may be of more assistance to employment judges than that of Lord Justice Rimer. There is no appeal in Redrow Homes so that should hopefully be an end to the matter for a few years at least.

The case of Ministry of Defence v Fletcher got a great deal of media attention at first instance (see for example here), mainly because the claimant was a female soldier who was awarded £186,895.52 for sexual harassment and victimisation. Of this, £100,000 represented injury to feelings, including £30,000 for the upper band of Vento adjusted for inflation, £20,000 for aggravated damages, and £50,000 for exemplary damages. It comes as absolutely no surprise that the Ministry of Defence decided to appeal.

There was no dispute that exemplary damages can be awarded in discrimination cases but only in two situations, the relevant one here being “oppressive, arbitrary or unconstitutional action by servants of government.” It has been established that they must only be awarded in exceptional circumstances where the wrongdoing is conscious and contumelious. In this case, the Ministry of Defence had been heavily criticised for its systematic failure to provide redress for the complaints of Ms. Fletcher but this was not considered to be serious enough by the Employment Appeal Tribunal to justify such an award.

There were two interesting obiter remarks by Justice Slade. Firstly, she rejected an alternative ground of appeal that systematic failures could not be taken into account because they had not been pleaded. She stated that “remedy is at large” and anything can be taken into account so long as findings of fact have been made. She also said that, even if exemplary damages had been appropriate, they would have been reduced to £7,500. An award of £50,000 is considered to be the maximum for cases of wrongful arrest and false imprisonment and a failure to provide redress is quite simply not comparable.

Although Ms. Fletcher was undoubtedly treated appallingly, this decision is equally undoubtedly correct. As noted elsewhere in the judgment, £50,000 is a typical award for moderate brain injury or for severe post-traumatic stress disorder and so £100,000 for injury to feelings is excessive. In any event, even allowing for the aggravated damages being reduced to £8,000 and a substantial reduction of interest as a result, the overall compensation will remain in six figures despite a total reduction of around £70,000.

I had the pleasure of attending the Employment Appeal Tribunal as an observer yesterday in the case of Grainger v Nicholson, which I previously discussed here. As a result, I am now much less confident of the views which I had then. John Bowers QC, for the appellant employer, argued that the definition of a religious or philosophical belief in employment law must be much narrower than the definition of a religion or belief under Article 9 of the European Convention of Human Rights, which considers a belief in the efficacy of smacking children to qualify for protection. Otherwise, he argued, it would open a floodgate of claims and could potentially also give undeserved protection to the British National Party.

Dinah Rose QC, who represented Mr. Nicholson, countered that the wording of the Religion or Belief Regulations, and the Framework Directive which they implement, makes it clear that the test should be just as wide as it is for Article 9. So, she said, do the comments of Baroness Scotland, sponsoring the bill, in Hansard. This approach is also supported by Justice Elias, as he then was, in Eweida v British Airways (see paragraphs 26-27). The British National Party may nevertheless not easily be protected, as a result of Article 17 of the Convention, which prevents it being used to limit the rights of others.

Because the belief of Mr. Nicholson includes a moral imperative to act to prevent climate change, it is easier to see it as a philosophical belief instead of a scientific theory unlike, say, Darwinism. However, Dinah Rose went on to argue that even Darwinism should be protected. Otherwise, a teacher could be dismissed for a belief in Darwinism but not for belief in creationism. Because a belief in Darwinism is not the same as a non-belief in creationism, it would not always be protected in that way. Leaving aside the fact that modern creationists say that Intelligent Design is itself a scientific theory, this presents a real problem. I have personally been struggling to resolve this matter in a way that I find satisfactory, as neither Darwinism qualifying as a religion nor going without any protection is particularly attractive.

Fortunately perhaps for Justice Burton, he does not have to decide this issue, only the one concerning the environmental beliefs of Mr. Nicholson. He has reserved his judgment and so I shall say no more.

There has been an interesting recent judgment from the Court of Appeal, reversing the decision of the Employment Appeal Tribunal in Oakland v Wellswood (the link is to the now overruled judgment). The case is unusual because the insolvency issue did not arise in the context of an employee who had been made redundant immediately afterwards but rather someone who had been transferred to a new company but then dismissed less than a year later. Therefore, the question which arose was whether service with the old company could be taken into account in considering his continuity of employment.

Because of these facts, the Court of Appeal did not need to consider TUPE at all in making its decision. As a result of section 218(2) of the Employment Rights Act 1996, continuity of employment is preserved where there is a transfer from one employer to another, whether or not TUPE applies. Therefore, there was jurisdiction for a claim of unfair dismissal and it did not matter that the point was not argued below.

Nevertheless, the obiter comments of their Lord Justices may prove to be more significant. By virtue of regulation 8(7) of TUPE, the regulations do not apply to insolvency proceedings instituted with a view to liquidation of the assets of a company. Judge Peter Clark at the EAT concluded that this includes “pre-pack” administrations where an administrator is appointed solely for the purpose of selling the company assets to a new company before liquidation is commenced. This would seem to be an easy way to avoid the effects of TUPE and, with respect to Judge Clark, it does not really promote a rescue culture because of the frequency with which pre-packs are used by owners to sell businesses to themselves.

In any event, the Court of Appeal doubted that regulation 8(7) would apply in these circumstances and, although it was not necessary to resolve the case, it is likely that their approach will be followed in the future. As usual, I shall be posting a link to the judgment on this blog as soon as it is published online.

The Employment Appeal Tribunal case of Henderson v Connect (South Tyneside) is not very likely to be reported as there is not much new law in it. However, it is a nice example of a judgment which provides a useful survey of the existing authorities, in an area in which many of them are quite old and not all of them are particularly clear. It concerns the tricky issues that arise when an employer sacks someone at the behest of a client, especially if the client does not have good grounds for making such a demand.

The background is the seminal case of Devis v Atkins, which established that the fairness of a dismissal does not depend on justice being done. A dismissal can be unfair even where justice is done (although it may be equitable to reduce the compensation awarded) but, conversely, a dismissal can be fair notwithstanding the fact that there is injustice, such as when the employer has a reasonable belief in the guilt of an employee who is, in fact, innocent. The issue is whether the employer acted reasonably.

It follows that, if a third party has a contract with an employer which gives it a veto on who carries out work for it, and the employer does everything possible to mitigate the effects of such a decision, a dismissal will be fair. So said the Court of Appeal in Dobie v Burns. Unfortunately, however, the Court of Appeal did not quite say this but rather it said that the injustice to the employee must be taken into account when considering the fairness of the dismissal and it appeared to distinguish Devis in this respect. This is ambiguous. Although it seems to have always been interpreted in accordance with the first sentence of this paragraph, which is also supported by the EAT decision in the same case, there is another possible interpretation, which is that a dismissal which results in injustice to an employee can be unfair even in circumstances where the employer has done everything that can be done to avoid it.

Henderson rejects the latter approach although, strictly speaking, it does not rule upon the matter because it was not argued before it. In any event, there is another point from Dobie v Burns which does remain outstanding, which is the suggestion of the EAT that, where a employer has a client with a right of veto, this right should be expressly incorporated into the contracts of its employees or else a dismissal might be rendered unfair. This is something which does go directly to the reasonableness of the actions of the employer and so it could be something which tribunals will want to take into account.

The case of Eagle Place Services v Rudd had an unusual outcome in that the tribunal found direct disability discrimination but not disability-related discrimination. Both sides were not happy with the outcome, with the employer challenging the former finding and the employee challenging the latter. It was the cross-appeal which succeeded, with the effect that the claimant received a six-figure award.

The reasoning of Judge Serota is throrough and sheds a great deal of light on Malcolm and the relationship between the two forms of disability discrimination being considered. At paragraph 76, he expresses the orthodox view that, “There is no issue that as a result of Malcolm the comparator in cases of both direct and disability related discrimination are likely to be the same.” This is cause for worry in the light of the very narrow interpretation given to direct disability discrimination in the only two reported cases, High Quality Lifestyles v Watts and Stockton-on-Tees Borough Council v Aylott.

This case provides fewer grounds for pessimism. Unlike in Aylott, it was considered permissible for a tribunal to find that a stereotypical view of disability consitituted direct discrimination. In particular, an unreasonable belief that a disabled employee is a commercial liability is not a relevant circumstance that must be attributed to a comparator. Furthermore, where the reason for treatment is clear, it is not always necessary to identify a comparator, following the case of London Borough of Islington v Ladele.

The analysis of disability-related discrimination is also encouraging. The tribunal had found that the case was not made out because the disability was not in the mind of the employer at the time of the dismissal. Its reasoning was that, because it did not have a basis for believing that the employee was a commercial liability, it cannot have held such a belief for a reason related to his disability. This is sensibly rejected. What matters is why the belief is held rather than whether the belief is accurate.

Further to this post, I can report that the judgment of the Court of Appeal in this case is now available.

Barack Obama appeared US television this week on the Late Show with David Letterman. This followed accusations by his predecessor Jimmy Carter that criticism of his policies about healthcare reform were motivated by racism. When asked about this, Obama observed that, “I was black before the election.”

A simiar problem can arise in employment tribunal cases, especially where it is alleged that a dismissal is discriminatory. If an employer really wants to avoid having employees from, say, a particular ethnic group, the best way to acheive this is surely not to employ anybody from that group in the first place, rather than employing someone only to be dismissed at a later date. Tribunals often specifically ask claimants to explain why their employers recruited them if they are as racist as they are alleged to be.

The problem is not insoluable and there are situations in which there is a straightforward answer, such as when companies are taken over by new managers who are less enlightened than their predecessors and decide to undo their good work. A more complex variation of this situation can happen in public bodies, and other large centralised institutions, where there is a strong desire from managment to promote diverse recruitment but this culture is not always supported by less senior staff. Finally, the saddest cases are those in which employers mean well and strive to avoid discrimination but are hamstrung by stereotypical assumptions, which means they are perfectly happy, for example, to select black job applicants for jobs, but will be more likely to dismiss them for perceived agression than their white colleagues, sometimes without even being aware of it. In any event, anyone wishing to argue that a dismissal is discriminatory should expect to provide an explanation and evidence to support it.

By contrast, dismissals based on pregnancy discrimination and victimisation are much easier to prove, precisely because the employee in question usually was not pregnant and usually had not made any complaints of discrimination when first employed. If the dismissal occurs immediately after a change of status, the employer has a lot of explaining to do. The same can also be true of disability discrimination where the employer was initially unaware of the disability, either because of a failure to ask about it or because of a failure to spot the obvious, such as that a person being interviewed has a prosthetic arm.

There has already been a great deal of discussion about the decision of the European Court of Justice in Pereda v Madrid Movilidad, in which it was held that workers who become sick while on holiday can reclaim their supposedly lost annual leave to be taken at a later date. Not surprisingly, much of the commentary has focused on the fact that this may allow some people to abuse the system, especially given the well-known sickness-inducing properties of sangria. In the past, I have largely been willing to defend Luxembourg on the holiday pay issue, and I had nothing to say in criticism of the controversial decision in Stringer v HMRC earlier in the year. However, I do agree that it has gone too far this time.

What is frustrating is that the court ruled on an issue that was not before it to be decided, in saying that not only can holiday be taken at a later date, but also that it can be rolled over if it is not practicable to take it in the same leave year. This upset the delicate balance established in Stringer, which appeared to give member states a choice between allowing employees to take holiday while off sick, and allowing untaken holiday to roll over from one year to the next. The former choice suited the United Kingdom rather well because the Working Time Regulations expressly forbid leave from being carried over from year to year and, in retrospect, it is even more unfortunate that the House of Lords did not expressly endorse this option when it had the chance to do so. Pereda has now thrown a spanner into the works.

Because of the minor possibility that a worker might become sick while on holiday, it appears that it is no longer lawful for member states to have a blanket ban on carrying leave over, with the result that the Working Time Regulations are incompatible with European law. This must have been contemplated by the ECJ who, in Stringer, appeared to be so sensitive to differences between member states in the implementation of the Working Time Directive. It is entirely disproportionate that the government will have to draft new regulations in order to deal with this piece of trivia. In recent posts, I have discussed the importance of tribunals making careful decisions on remedy when a lot of money is at stake but this case will do the opposite and introduce an unnecessary level of complexity into small value claims. I have already had a one-hour hearing relisted for a whole day and this will now inevitably be repeated.

Today, I am travelling in my DeLorean back to the year 1985, when Mikhail Gorbachev became leader of the Soviet Union, Boris Becker won Wimbledon as an unseeded 17-year-old, and the mobile phone was introduced into the UK by a tiny company called Racal-Vodaphone. Number one singles included ‘Nineteen’ by Paul Hardcastle, ‘The Power of Love’ by Jennifer Rush, and ‘Dancing in the Street’, from the Live Aid concert, by David Bowie and Mick Jagger. In employment law, it was a crucial year in the fight over the correct test for constructive dismissal, one which some say is still being fought in 2009.

The modern orthodoxy was represented by the case of Lewis v Motorworld Garages in the Court of Appeal, which approved the formulation by President Browne-Wilkinson in Woods v W M Car Services, that conduct without reasonable or proper cause, which is calculated or likely to destroy or seriously damage trust and confidence between employer and employee, constitutes a repudiatory breach of contract. There would later be further approval from the House of Lords in Malik v BCCI. The radical alternative, which everyone (?) agrees is now heretical, was expressed in the case of Dutton & Clark v Daly, in which it was suggested that an act by an employer within the reasonable range of responses can never be a repudiatory breach but, more controversially, an act outside the reasonable range of response will always be so. The latter formulation would certainly seem to place the bar too low and makes the test too close to a reasonableness test, whether or not you believe the reasonable range of responses test is just a reasonableness test, and this is inconsistent with Western Excavating v Sharp.

However, the contemporary debate is about whether or not the first part of Daly is correct and, if so, whether or not it is useful. In my view, it must be correct because, if Daly is setting the bar too low, anything that does not even pass the Daly test cannot be sufficient. I think it is useful because the Woods definition of a trust and confidence breach is far from precise and is apt to lead to inconsistent application. If this means that a gloss is being placed on the test, this is no great disaster. Constructive dismissal is a common law principle rather than a statutory one and, moreover, the ordinary English meaning of the word ‘trust’ clearly evokes concepts such as fairness and reasonableness. If I say that a friend has not acted unreasonably but I no longer trust him, he is likely to find this hard to accept.

For me, the biggest question about the trust and confidence test is where reasonableness comes in, if not in establishing a breach, and this is where cases like Buckland run into trouble. Buckland cites the 1985 case of Delabole Slate v Berriman, in which it is confirmed that a constructive dismissal can be fair. This is fine but trust and confidence was not breached in that case. In another case from the same year, Cawley v South Wales Electricity Board, it was held to be perverse to find that a disproportionate disciplinary sanction breached trust and confidence and yet gave rise to a fair dismissal, and it is hard to see how it could be otherwise. It was also said that considerations of fairness and the considerations affecting constructive dismissal are two sides of the same coin. Given the importance of fairness and reasonableness to unfair dismissal law, they must be key in deciding whether a breach is established.

While I am on the subject of decisions on remedy which are upheld by the Employment Appeal Tribunal but overturned by the Court of Appeal, another example was first mentioned in this post. I am pleased to report that the judgment of the Court of Appeal in Stuart Peters v Bell has now been produced. Lord Justice Elais gave the leading judgment, in which he confirms that the narrow Norton Tool principle, as defined in the case of Burlo v Langley, does not apply to constructive dismissal because it is simply not the case that it is good industrial practice to make a payment in lieu of notice in such a situation. As Burlo also confirms that the narrow Norton Tool principle is not to be extended, the appeal was allowed.

There was an interesting decision by the Court of Appeal a few weeks ago, concerning compensation for unfair dismissal. Unfortunately, the judgment has yet to appear but the key background reading in any event is the Employment Appeal Tribunal judgment in Aegon v Roberts, earlier in the year, which it overturned. The facts are unremarkable but the same cannot be said of the approach taken to the law.

The original tribunal noted that the pension scheme of the dismissing employer was far more generous than that of the job which was subsequently taken by the claimant. Accordingly, she was able to claim the difference as part of her loss. This would have been uncontroversial were it not for the fact that the second job also provided a larger salary and so it would be expected that any loss of pension would be offset by this additional income. The tribunal took another view, reasoning that there had been a break in the chain of causation with respect to the loss of salary but not with respect to the loss of pension.

The Employment Appeal Tribunal upheld the decision but this is no longer surprising. It once described the approach of tribunals to compensation as “rough and ready” and it continues to be very reluctant to interfere with such assessments. The problem is that this is an area in which there is a great deal of case law, which cannot readily be ignored. There is no point in tribunals making referemce to a break in the chain of causation if they are going to conclude that there can be a partial break or that it can be broken more than once, which is not unknown. Therefore, there is an increasingly common sequence of events in which a tribunal makes a rough and ready assessment of compensation which is upheld by the Employment Appeal Tribunal and then overturned by the Court of Appeal, as happened in this case.

In many ways, the approach of the Employment Appeal Tribunal is understandable. It wants to prevent tribunal litigation from becoming complicated by technical points of law and it wants to avoid endless appeals concerning small sums of money. However, I agree with the Court of Appeal on this issue. The overriding objective requires proceedings to be conducted in a manner proportionate to the amount of money at stake. Therefore, it is right that one-hour wage claims worth a few hundred pounds do not get bogged down in technicalities but, conversely, when tens of thousands of pounds are being claimed, as in this case, tribunals surely ought to take the time and care to produce a more thorough analysis of the losses of a successful party. It should also go without saying that, however badly an employer has acted, compensation should reflect actual loss and should not be a punishment for these actions or for failing to settle, or a reward for a claimant whose successful mitigation means the case has little value.

There have been very few new decisions recently but there was another case on time limits that caught my eye, McFadyen v PB Recovery, which is a reminder that employment tribunals in Scotland are a separate jurisdiction from those in England and Wales. Usually this makes no difference because the law is identical to the point that decisions of the Scottish Employment Appeal Tribunal are binding in England and Wales and vice versa. However, it affects the issue of where claims should be submitted.

When claims are made online, they go to a central server and are then forwarded to the tribunal office local to the place of work of the claimant. If the claimant did not work at the main address of the respondent, there is a separate part of the ET1 form in which the actual place of work can be indicated. In this case, the employees worked in Scotland for a company based in Bristol but they did not put their place of work on the form and so the claim was received in Bristol. When the error was spotted, the Glasgow office was able to process the claims but they were now out of time and were struck out.

When a claim is sent to the wrong office within the same jurisdiction, no problem arises because it can simply be transferred to the correct office. The problem in this case was that the Bristol tribunal said that it could not transfer the claims to Glasgow as it did not have jurisdiction to deal with the claims in the first place. In fact, this was wrong. Tribunals can deal with claims against employers based within the jurisdiction, even those brought by employees who worked elsewhere. This does not mean that there would necessarily be jurisdiction to deal with a case of unfair dismissal but there would at least be sufficient jurisdiction to transfer the claim to the correct office. The problem in this case is that the appeal was brought against the decision of the Glasgow office rather than the one of the Bristol office.

The case of Gisda Cyf v Barratt concerns the thorny issue which arises when a dismissal letter is sent to an employee but is not read until several days later. A well-established principle was set down by the Employment Appeal Tribunal in Brown v Southall & Knight, as long ago as 1980, but it came under challenge for the first time in the Court of Appeal. The majority upheld its ratio that, so long as there is no deliberate avoidance of reading a dismissal letter, it will not come into force until it is actually read.

This result differs from the common law position in the case of The Brimnes, well-known to first year law students, in which a contract was brought to an end by a fax sent within office hours, even though it had apparently not been seen. This discrepancy is the basis of a persuasive dissent by Lord Justice Lloyd. However, Lord Justice Mummery, giving the leading judgment for the majority, emphasises the policy issue which arises from the three-month limitation period. An employee who does not find out about being dismissed until some time afterwards will have even less time to decide to bring a claim.

An intriguing point arises from the unreported constructive dismissal case of Potter v RJ Temple, in which a resignation letter faxed to an employer late in the evening was deemed to have taken effect immediately, causing the time for bringing a claim to start to run. The Court of Appeal does not deal with this contradiction entirely satisfactorily but it is clear that policy weighs in a different direction when somebody resigns, because the crucial issue is when the person bringing the claim becomes aware of a right to do so. The knowledge of the employer is, in this context, irrelevant. Lord Justice Lloyd suggests that this inconsistency undermines Brown but, as Lord Justice Mummery notes, Potter is not consistent with The Brimnes either because a termination communicated out of hours is ineffective in contract law.

The approach of tribunals towards amendments is increasingly liberal, especially if no new facts are pleaded. However, a difficulty remains in situations where an amendment is requested at the hearing itself. It is particularly prejudicial for a respondent to be faced with an allegation for which it has not prepared but, conversely, justice demands that a claim should not be defeated by the technicality of pleadings, in what is supposed to be a jurisdiction accessible to all. The overriding objective provides the possibility of granting a short postponement to allow a respondent to take further instructions. As a result, the amendment can be granted while keeping any prejudice to the respondent to a minimum.

Given the range of options that are available to a tribunal, it should not be surprising that there is so much variation in practice, particularly when it comes to the granting of last minute amendments. The guidelines from the famous case of Selkent Buses v Moore do not always make it possible to predict the outcome because there is always prejudice to a respondent in such a situation. The only question is whether it is sufficient to defeat the amendment and, on the face of it, Selkent does not answer this.

However, there is a lesser-known passage in Selkent which states that an amendment can be rejected if it is hopeless on its face, regardless of the other factors. If it is arguable then all the circumstances must be considered but the non-exhaustive list of relevant factors does not include the merits of the amendment, once the threshold of arguability has been reached. This seems inconsistent. If a case is only barely arguable then the degree of prejudice should surely have a greater weight than if is very straightforward and my experience suggests that this is indeed the case, although not always stated explicitly by tribunals. This is a classic example of dark law where a sensible gloss to the authorities is adopted by tribunals but has never been supported by a higher court, in part because the tribunals themselves are unwilling to allow it to be tested. Ironically, an appeal-proofing mentality results in the law being less clear so that parties are less likely to understand why they have lost. I am not convinced that this approach always reduces the number of appeals and, quite often, it has the opposite effect.

I am delighted to announce something of an exclusive for the Employment Law Advocates blog. Please click here to read the written reasons of the most famous (or rather infamous for Abercrombie & Fitch) employment tribunal case of the year. I would like to thank May Dean for providing me with a copy.

There is one finding of fact which is not strictly relevant to the case but nevertheless reveals a lot about the culture of the American store. One of the terms of a settlement with ethnic minority employees was to increase the level of diversity in its promotional material. However, the judgment notes that training slides shown to new staff have only been changed in order to increase the racial diversity of shoplifters.

Moving onto the law, I am most intrigued by the claim that Riam Dean lost, that of direct discrimination. The tribunal finds, at paragraph 49, that she was sent to the stock room by a manager who knew that she had a prosthetic arm, in circumstances in which an employee who did not have a prosthetic arm would have simply been asked to remove the cardigan she was wearing in breach of the look policy. However narrowly direct discrimination is defined in disability discrimination, this must surely qualify.

Apparently not, according to the tribunal. Their logic  is, firstly, that the comparator must be someone who would be believed to be reluctant to remove the cardigan and would therefore also have been sent to the stock room. The problem is that the tribunal does not make a finding as to how such a person would have been treated but instead considers the entirely different scenario of a person who refuses to remove it and is disciplined. There is no evidence that Riam Dean would have refused and she often wears no more than a band around her elbow. The tribunal completely ignores the crucial question of whether she would even have been allowed to work on the shop floor if she had not worn a cardigan.

Secondly, the tribunal says that being sent to the stock room is not necessarily a detriment because it might have been more upsetting for Riam Dean to remove her cardigan. This is an error of law as a detriment should be asssessed primarily from the point of view of the claimant. If she would prefer not to be sent to the stock room, and this is a reasonable preference, then a detriment is established. What this case shows is a marked reluctance by tribunals to make findings of direct disability discrimination, for fear of an appeal, even where there is a good chance of them being made out. However, I do not want to be too critical of the tribunal as they largely got it right and the overall outcome is a good one.

As promised, a summary of my recent appearance at the Employment Appeal Tribunal appears below.

The case of Tapere v South London and Maudsley NHS Trust concerns the scope of Regulation 4(9) of the TUPE Regulations 2006, which allows a person to resign and treat her contract of employment as having been terminated if there is a substantial change in working conditions to her material detriment. It also makes some observations about the scope of mobility clauses and the effect of TUPE upon them.

Ms. Tapere had been employed by the Lewisham Primary Care Trust for more than nine years and was based exclusively at one of their properties in Burgess Park, although a mobility clause allowed her to be moved to any property within the Trust. Following a transfer to the neighbouring South London and Maudsley NHS Trust, she continued to work at Burgess Park but was informed that she would shortly be moving to Bethlem Hospital, a property owned by her new employer. A few months later, she returned from holiday to discover that the move had taken place in her absence and she resigned and claimed that she had been constructively dismissed. The tribunal found that she had not been entitled to resign because the scope of her mobility clause had changed following the transfer to reflect the area covered by her new employer or, alternatively, because the Trust had acted reasonably in proposing the move.

The Employment Appeal Tribunal, chaired by Judge Hand QC, overturned this decision and substituted a finding that there was a constructive dismissal. The case is authority for the following propositions:

1) The geographical range of a mobility clause is not altered by a TUPE transfer and cannot be replaced by one of substantial equivalence in the same way as, for example, an employee profit share scheme.

2) A tribunal may not find an implied mobility clause which differs from an express clause if there is no implied qualification as to reasonableness. There was no evidence of such an implied term in this case.

3) In Regulation 4(9), a material detriment is defined in the same way as a detriment in discrimination law, in the case of Shamoon v Royal Ulster Constabulary. The word ‘material’, which does not appear in the Acquired Rights Directive, adds nothing. Therefore, the tribunal erred by applying an objective test.

An intriguing but under-analysed problem is how to deal with a situation in which some symptoms of a medical condition would be able to qualify as a disability but others, in themselves, would not, such as a psychological condition resulting in impaired understanding but also a tendency to steal. If the tendency to steal were the only symptom, it would be an excluded condition that is deemed not to qualify as a disability and, as the education case of X Endowed Primary School v SENDIST confirms, the excluded condition can be severed from the other symptoms. The High Court states that the word “condition” can refer to either the condition as a whole or merely to some of the manifestations that result from it, echoing the similar reasoning concerning the word “impairment” in the case of McNicol v Balfour Beatty, rather than the words having distinct meanings as a literal reading of the statute would seem to imply.

But X Endowed School does not deal with the situation in which the non-disability is not an excluded condition. For example, there are some people with cerebral palsy who have substantially impaired speech, and mobility which is impaired but not to a substantial extent. However, the wide test in McNicol would appear to suggest that the cerebral palsy can itself be considered to be the impairment, with the effect that reasonable adjustments must be made for a mobility problem which would not amount to a disability on its own. This seems strange. Clearly the McNicol decision was intended to prevent tribunals from being confused by unnecessarily elaborate arguments on such a point, but there are side effects.

This is not just an academic criticism. The McNicol formulation can result in extremely difficult cases of the most bad-law-generating kind. The obvious and unavoidable example is Malcolm in which a decision to sublet a property was supposedly caused by schizophrenia. Far from producing insuperable problems of causation, an analysis based on separating disabilities from non-disabilities would have made it much easier to resolve. If the House of Lords had disapproved McNicol rather than Clark v Novacold then Mr. Malcolm would have failed on the straightforward ground that resisting the temptation to sublet is not an everyday activity. In my opinion, this would have had a much less damaging effect upon the law.