Non-disclosure agreements: the truth behind the headlines

Is there a place for NDAs post #MeToo? Any restriction on freedom of contract will require very careful consideration, writes Jonathan Cohen QC


 ‘Wrong’, ‘Corrosive’, ‘Gagged’, ‘Cover-up’, ‘Immoral’

These are just some of the colourful descriptions given to the ‘NDA’ (non-disclosure agreement), as it has now come to be known in the public consciousness. Notwithstanding the use of NDAs for decades, it appears that we have only belatedly recognised this scandalous abuse of clauses which were originally designed to protect commercially confidential information, as one national newspaper described them. Can that really be right? The answer may be that the mood music matters. NDAs used in situations of little interest to the press or the public have unsurprisingly generated no adverse comment. NDAs used by litigants who may be thought to be on the unattractive side, Harvey Weinstein and the like, have generated a considerably different reaction. Is it justified?

Legitimate purpose? The settlement bargain

The starting point is to understand what an NDA actually is. Litigation predominantly settles. That is true of all types of litigation, whether rooted in employment law (where the NDA debate has been substantially focused) or commercial disputes. In the event of settlement, the rights and wrongs of allegations will never be determined. Reputation matters. Though not in issue in some claims, many and perhaps even most disputes will to a greater or lesser extent involve a challenge to the honesty or probity of the opposing party. It is not therefore surprising that a litigant might want, as part of an overall settlement agreement, to keep potentially damaging allegations confidential. It would be naïve to think that because allegations remain just that, ie undetermined, the risk of reputational harm vanishes. The NDA serves an undeniably legitimate purpose. As part of a settlement bargain, it protects reputation. There has never been any English public policy rule that prevents such a bargain from being made, and why should there be? If I, a wholly innocent defendant, am the subject of a false and reputationally damaging claim, which can be settled with confidentiality for £1,000, should I be forced instead to spend hundreds of thousands of pounds to trial to protect my reputation? Such an approach would certainly sit uncomfortably with the endless exhortations to litigants to settle their disputes.

This example of the innocent defendant colours the view that one takes of NDAs. The position becomes considerably less appealing when one takes the contrary example of the defendant who has engaged in a serious wrong. Allowing that defendant to escape public opprobrium is harder to justify. Not only does the guilty defendant avoid the court of public opinion by the good fortune of being able to afford to pay for it; there is also the very important consideration that other potential victims are not alerted to the risk of repeat conduct. This might be thought to be particularly significant where the guilty defendant has engaged in dishonest or discriminatory conduct. Publicity has an undeniably important function in those sorts of cases. Indeed, not only might publicity prevent repeat harm, it may also deter it in the first place. In October 2017, the Daily Telegraph reported that Weinstein’s employment contract prevented his employment from being terminated in the event of sexual harassment claims, providing he ‘paid off women to silence them’. If true, it would be a very good example of a situation where the power of the deterrence effect is permitted to be sidestepped.

Baby and the bathwater?

Deterring and preventing the repetition of wrongful conduct, particularly dishonest or discriminatory behaviour, is undoubtedly important. However, any attempt to prevent the use of NDAs in certain circumstances risks throwing the baby out with the bathwater. Distinguishing between the good defendant and the bad defendant is impossible, absent a trial. There is no way to stop the fraudster or the discriminator from obtaining confidentiality for untried allegations without, at the same time, preventing the innocent defendant from obtaining confidentiality for the rightful protection of reputation. Another approach might be to attempt to distinguish between causes of action. The law might, for example, prevent the use of NDAs in discrimination claims, but permit them in claims for breach of contract. However, that still gives rise to the problem of the innocent defendant, seeking reputational protection. Does the legislature make a hard decision, which favours deterrence and prevention, at the expense of depriving certain litigants of the ability to seek reputational protection through settlement? All remains to be seen. The Women and Equalities Committee chaired by Maria Miller MP is currently carrying out an inquiry into the use of NDAs in discrimination cases. It seems clear that firm regulation will be proposed, but how far that baton will be taken up by the government remains to be seen.

The Californian approach

Use of NDAs is an equally hot topic in the USA. California has introduced legislation, in force in 2019, banning NDAs in sex discrimination cases. However, the devil is in the detail. The ban only operates once litigation has been commenced (or at least where certain mandatory pre-litigation formalities have been complied with). Pre-litigation settlements are unaffected by the legislation. This hardly seems to be a solution to the problem. There would seem to be no greater or lesser justification for permitting confidentiality once litigation has begun as there would be beforehand. Indeed, one might think that the only real effect of legislation of this kind will be to provide a new tool in the Californian litigator’s armoury to achieve settlement at a point in time earlier than would hitherto have been the case.

The Californian approach seems an unlikely solution in this country. The government will ultimately need to decide where the pain of injustice should be felt. Is the innocent defendant to be deprived of a right to protect reputation and, necessarily, should settlement be de-incentivised in some cases, as a price to pay for deterrence and prevention? Should any new rules only apply to discrimination cases, or should they also extend to wrongs of equal seriousness, for example fraud? The easiest option will be to kick the issue into the long grass and, with plenty of other pressing issues on the legislative agenda, one can well see that happening.

Where does all this leave practitioners?

In the meantime, practitioners are left in an uncertain position. An NDA clause which seeks to prevent a complaint to a regulator or law enforcement authority is an obvious no go area. The clause would be unenforceable in any event, but even the attempt to prevent such complaint could result in disciplinary action. If that was not always clear, it has been made so as a result of a Solicitors Regulation Authority (SRA) Warning Notice, issued on 12 March 2018. What the Warning Notice expressly does not do is to prohibit the use of NDA clauses in settlement agreements. To the same effect is a practice note issued by the Law Society on 9 January 2019. There would be no legitimate reason to outlaw NDAs on the law as it stands and this guidance to solicitors rightly does not do so. Bar Standards Board guidance is awaited, but is unlikely to be materially different to that issued by the SRA. The dividing line between permissible and impermissible conduct in respect of NDAs will only become clear if and when practitioners are prosecuted. The author is aware of at least one successful Solicitors Disciplinary Tribunal (SDT) prosecution of a solicitor for entering into a settlement agreement, on his own account, which prevented allegations being repeated. The NDA clause in that case made no mention of repetition to the SRA, but the general nature of the prohibition was, in the view of the SDT, enough to catch an SRA complaint. Such prohibition was, in the view of the SDT, a breach of Outcome 10.7 of the SRA Code of Conduct, ie prohibiting a solicitor from attempting to prevent anyone from providing information to the SRA. If an intention to prevent regulatory complaint can at least in principle be identified from general words of confidentiality, it would be wise for practitioners to include specific wording in an NDA clause which carves out (at least) complaints to law enforcement or regulatory authorities. General prohibitions on the repetition of allegations are no longer fit for purpose and should not be used.

Nothing so far has been said about the position of claimants bound by an NDA. Historically some claimants have been forced to enter into an NDA as a quid pro quo for a satisfactory settlement. However, that is a very narrow way to look at the situation. A claimant is not bound to agree to an NDA. Any settlement is optional. In employment claims, it might be said that such option is sometimes illusory, because the cost of a trial would be prohibitive and settlement, however onerous the terms, is the only real choice. Even then, the inclusion of an NDA might result in a better overall settlement than would otherwise be the case. If defendants are unable to obtain confidentiality, the unintended consequence may be that some claimants who would otherwise obtain settlements will no longer do so, or those settlements will be less beneficial in financial terms. What this shows is that any restriction on freedom of contract requires very careful consideration. Hard cases make bad law. A knee jerk reaction to recent headlines would be unfortunate.

Jonathan Cohen QC is a QC at Littleton Chambers specialising in commercial dispute resolution and employment law.

Issue: 
Author details: 
Jonathan Cohen QC

 QC at Littleton Chambers specialising in commercial dispute resolution and employment law.