When are ‘without prejudice’ statements admissible?

Ferster v Ferster and others [2016] EWCA Civ 717
Court of Appeal, Civil Division

The Unilever Rule[1]

 The general principle is that no one can give evidence of what was said in without prejudice negotiations. It is, however, subject to a number of exceptions. In fraud litigation, the most important one is that the exclusion of the evidence would, in the words of Walker J in Unilever Plc v The Procter & Gamble Co “… act as a cloak for [1] perjury, [2] blackmail, or [3] other ‘unambiguous impropriety’.”[2] (I have added the numbering for clarity.)

The rule (for the purposes of the article, let’s call it “the Unilever rule”) is an important one. However, it cannot be taken to mean what it seems to say. If it did, the floodgates would open to a mass of allegations that people had admitted knowing their cases to be factually flawed “without prejudice.” In Savings & Investment Bank Ltd v Fincken[3] Rix LJ said:

“It is not the mere inconsistency between an admission and a pleaded case or a stated position, with the mere possibility that such a case or position, if persisted in, may lead to perjury, that loses the admitting party the protection of the privilege…it is the fact that the privilege is itself abused that does so. It is not an abuse of the privilege to tell the truth, even where the truth is contrary to one’s case. That, after all, is what the without prejudice rule is all about, to encourage parties to speak frankly to one another in aid of reaching a settlement…”

What about the words “…as a cloak for perjury” referred to in the Unilever rule? Does it make a difference if the admission of untruth relates to a previously sworn document rather than a fundamental premise of, for example, a pleading? According to Rix LJ it did not. The words were intended to refer to “a blackmailing threat of perjury” rather than an admission of it.[4] In other words, it had to be prospective (i.e. a threat of future perjury); not retrospective (an admission of past perjury).

That all seemed clear enough, but what about the reference in the Unilever rule to ‘unambiguous impropriety?’ In Ferster v Ferster, the three Lord Justices have agreed on an answer. It is not as clear as it might be.

Ferster v Ferster – the facts

Three brothers were shareholders in a company. Two of them used their powers to procure that the third (“Jonathan”) was sued by the company for, amongst other things, breach of fiduciary duty in its management (“the company action”). Jonathan struck back. He commenced a shareholders’ petition under s994 CA 2006, saying that the affairs of the company were being conducted in a way that was unfairly prejudicial to his interests as a shareholder. He claimed that the company action had been procured by his two brothers for an improper purpose – to pressurize him into buying their shares at a price in excess of their value.

The offending email

After an unsuccessful mediation the two brothers negotiated with Jonathan on a “without prejudice” basis by exchanging emails. One particular email that they sent withdrew an offer to sell their shares at a certain price and demand a higher one. The stated basis for the price increase was that the two brothers had become aware of “further wrongdoings” by Jonathan, which included, so they said, his holding bank accounts in England and offshore, either in his own name or through his partner. Those banks accounts, it was said, should have been disclosed by Jonathan pursuant to a disclosure order in the proceedings by the Company. The email said that the discovery of these bank accounts would lead to “…charges of perjury, perverting the course of justice, and contempt of court.”

The email alluded to the possibility that their allegations would become public within 48 hours. It also mentioned possible imprisonment for contempt of court. The prospect of similar penalties being imposed upon Jonathan’s partner was referred to, as was the ruination of his reputation and ability to continue in his field of business.

Subsequently, the two brothers’ solicitors purported to clarify their client’s position, saying that their clients “…neither sought nor intended that criminal proceedings would be issued nor allegations of perjury made if the offer was not accepted.” Jonathan was not placated. He sought permission to plead the threats made in the email in his s994 petition.  His two brothers objected saying that the email was inadmissible as it was sent “without prejudice.”

This was no case of bullying the unprotected. The email was sent on behalf of the two brothers to Jonathan’s solicitors via an experienced mediator. Nevertheless, the Court of Appeal held in Jonathan’s favour on the basis that the threats that his two brothers made were sufficiently improper to justify breaking the without prejudice rule.

Why was the email unambiguously improper?

Floyd LJ said:

“Firstly, the threats went far beyond what was reasonable in pursuit of civil proceedings, by making the threat of criminal action, (not limited to civil contempt proceedings). Secondly, the threats were said to have serious implications for Jonathan’s family because of Jonathan’s wrongdoings. Thirdly, the threats were of immediate publicity being given to the allegations. It is nothing to the point in this connection that Warren and Stuart may have believed the allegations to be true. The threat to publicise allegations of extreme severity against Jonathan and his partner, and within such a short timescale, placed quite improper pressure on Jonathan. Fourthly, the purpose of the threats was to obtain for the brothers an immediate financial advantage arising out of circumstances which should accrue, if they had basis in fact, to the benefit of the company. Finally, there was no attempt to make any connection between the alleged wrong and the increased demand.”

So no single factor is identified as a sufficient self-standing ground of “unambiguous impropriety.” From the emphasis of argument in the judgment, however, the threats of contempt and imminent publicity, coupled with the fact that the action by the company was for its own benefit and not for Jonathan’s brothers’ benefit were obviously very significant.

Some years ago, I became concerned regarding the lack of guidance for mediators as to what were and what were not proper negotiating tactics in mediations. My concerns were not widely shared until I conducted a survey of mediators, barristers and solicitors, in which each was invited to judge whether a number of morally problematic situations would require a mediator to terminate the mediation. I was able to show that my concerns were justified: The disparity of views as to what was and what was not proper was simply enormous. It is interesting to note that the ‘unambiguously improper’ email may well have been drafted by reputable solicitors and had been relayed through an experienced mediator. The notion that there is consensus on what is or is not proper in everyday legal situations is, for practitioners, one of law’s most hazardous fictions.[5]

A question of wording?

Returning to Ferster v Ferster, one thing is clear. The rule was (and remains) that ‘unambiguous impropriety’ is a pragmatic test. It will not be satisfied if the court has to wade through a mass of evidence as to context or interpret what was said in the heat of the moment. The following passage from a judgment Hoffman LJ in Forster v Friedland was cited with approval:

“The rule is designed to encourage parties to express themselves freely and without inhibition. I think it is quite wrong for the tape recorded words of a layman, who has used colourful or even exaggerated language, to be picked over in order to support an argument that he intends to raise defences which he does not really believe to be true.”

Perhaps, then, the test is really one of “self-evident and unambiguous impropriety” (although nobody said so). If so, that leads to a further question of whether the brothers could have conveyed much of what they wanted to say had they expressed it differently. Very possibly they could have done. Floyd LJ said:

“It would have been entirely possible for Stuart and Warren or their advisors to make an increased offer for the sale of their shares by reference to what they regarded as their increased value (once the contents of the undisclosed bank account, and the interests of the company and its creditors had been allowed for), or because of what they perceived as the consequences for them personally as a result of increased prospects of the company succeeding in the company’s action. That is, however, not the offer which they chose to make. Instead, a fair reading of the email is that they wanted more for their shares because they had learned of their ability to cause the company to take the steps identified. Thus the offer was increased.”

Separate categories or a question of degree?

Probably the most central aspect of the judgment, however, remains the fact that the without prejudice communication was used not to make an admission but to make a threat. This chimes with the thinking in Saving & Investment Bank v Fincken: admissions are a proper purpose of without prejudice discussions and so they are safe. Threats are not necessarily so. They can be proper (“if you do not settle we will take the case to trial”) or improper (“if you do not settle you will be shot”). However, Floyd LJ said:

“…it may be easier to show that there is unambiguous impropriety where there is an improper threat than where there is simply an unambiguous admission of the truth” (emphasis added).

But what is meant by “easier”? According to the thinking in Savings of Investment Bank Limited v Fincken, an admission of past dishonesty (or even perjury) cannot be enough. After all, as Rix LJ pointed out (para 59), “…cases of fraud or dishonesty [would] become almost impossible to settle” if they were. So what kind of admission can be enough?

One possibility is that some admissions are sufficiently serious to disentitle the maker to without prejudice protection.[6] Take, as a hypothetical example, the Hoogstraten litigation in 2004, where the Defendant faced a civil action for murder (which he denied). If the Defendant had admitted murder in without prejudice discussions but said that he never intended to commit it again, could his admission be used in evidence?[7] Or perhaps it might refer to admissions utterly unconnected with the subject matter of the negotiation? E.g. “in the other action I had against you last year, I gave false evidence at trial.” On balance, it is best to assume the word “easier” simply to be a judicial escape door providing room to manoeuvre in situations not yet envisioned.

Despite the attraction of clear rules, perhaps some flexibility is sensible. The distinction between admissions and threats may be a little blurred. Presumably some admissions are close to “threats” where it is clear that the person who makes them intends to continue in court proceedings with the conduct he has admitted (e.g. tell a lie. “The court believed me and always will”). That seems to be what Rix LJ had in mind by his reference to lies “…if persisted in.” Perhaps some threats are in a sense, also close to admissions. Jonathan certainly wanted the offending email to be taken as a tacit statement that the brothers were pursuing an action in the name of the company for improper reasons.

So who won?

Let’s step back a bit. Did the email really help Jonathan to show that the action by the company was without basis? Not really. On 15th November 2016, Morgan J heard the s994 Petition and held him to have acted dishonestly and unreasonably, and that Jonathan “deliberately gave false evidence when he tried to explain away the contents of the financial documents which he had signed.” The Judge also held that he was “quite satisfied that in large parts of his evidence he was deliberately trying to mislead the court.” In other words (and by different means), Jonathan’s own Petition brought about some of the adverse consequences that he had been improperly threatened with.

All fraud litigators encounter parties who forget (or simply pay no heed) to the ultimate merits of a case. Some come to regard litigation as a “game” to be won by some technical quibble or elusive psychological “knock out” blow. Like many litigants who indulge in interim applications to support a case they know to be flawed, Jonathan may have felt “king for a day” following the Court of Appeal ruling.

But so did Macbeth. Briefly.

Robert Hunter, Partner


[1] Unilever plc v The Procter & Gamble Co [2001] 1 WLR 2436 per Walker LJ at page 2444.

[2]The expression “unambiguous impropriety” was borrowed in the Unilever decision from a judgment of Hoffmann L.J. in Forster v. Friedland (unreported), 10 November 1992; Court of Appeal (Civil Division) Transcript No. 1052 1992).

[3] [2004] 1 W.L.R. 667 [2003] EWCA Civ 1630.

[4] Based upon an obiter statement of Peter Gibson LJ in Berry Trade Lld v Moussavi. (No 2) [2013] EWCA Civ 715.

[5] See, for example, Brown Jenkinson & Co Ltd v Percy Dalton (London) Ltd [1957] 2QB 621.

[6] Floyd LJ explicitly stated that “unambiguous impropriety” did not have to amount to the criminal offence of blackmail.

[7] He denied murder (unsuccessfully in the Civil Proceedings).