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.