Silencing at Work: Sexual Harassment, Workplace Misconduct and NDAs

Abstract

This article draws on qualitative analysis of a small interview study and of oral evidence to the Women and Equalities Select Committee (WESC) of the UK Parliament, to explore and contextualize the use in the United Kingdom of NDAs in settlements to silence allegations and experiences of sexual harassment and workplace misconduct. This demonstrates the central place of NDAs in the failure of individualist individualized workplace protections effectively to combat this behaviour. Several factors emerge as potentially significant to the ubiquity and substantive content of NDAs in settlements: the influence of modern, technologically inflected routinization and sheer paperwork within organizations and lawyering; interactions between the law in practice and long-standing economic, social and cultural means of subordination, including of women and racialized groups. Illustrations comprise, first, employers’ systemic dominance, mediated by lawyers, being enacted and reinforced through frequent imposition of NDAs, including to protect serial wrongdoers, and, secondly, that the shame and career risk attached to people who experience and complain of ill treatment (not alleged wrongdoers) mean claimants themselves often need the protection of NDAs in settlements. Still, suggestions in the evidence of some organizations altering their use of NDAs provides an invaluable opportunity for future research to illuminate the drivers and long-term consequence of this legalized silencing and of what ought to be done about it. Organizational variety also shows that employers are already in a position differently to reconcile the innate tensions the data exposed between individual, employer, and societal interests in whether mistreatment at work is silenced or exposed. This leads to my final argument that, whatever legal reform occurs, employers and their lawyers ought already to take the obvious first step towards changing the culture, of shifting their default position away from routinely using law to silence complaints about sexual harassment and workplace misconduct. If employers are not even willing to take this small initial step, and their lawyers to back them up, this will raise serious questions about current public commitments finally to tackle workplace wrongdoing, and grave doubts that anything much is going to change in this dimension of working life.

1. INTRODUCTION

The #MeToo movement sparked worldwide interest in the use of non-disclosure agreements (NDAs) in the settlement of claims alleging sexual harassment at work. Questions have been raised about how NDAs impose secrecy and silence: shutting down complaint, impeding organizations from addressing workplace mistreatment and imposing lifelong obligations to keep quiet. Of particular concern has been the use of NDAs to hide allegations of serious and criminal behaviour and to provide legal cover for serial wrongdoing. In response, there has been investigative journalism, public debate, policy and political inquiry, discussion of lawyers’ ethical obligations and litigation to enforce NDAs.¹ Calls have emerged in multiple locations either to ban NDAs outright or at least to constrain their use, and reform has followed. Legislative change has generally however been limited, well demonstrated by the UK Government’s law reform proposals.² There have also been soft law interventions, some proposed by Government, some coming from regulators.³ The public debate about sexual harassment and its silencing is, nonetheless, far from over and evidence of continuing widespread use of NDAs shows the need for it to be kept alive.⁴

There certainly remain important gaps in understanding of the use of NDAs and the likely effects of law reform. This results from the shadowiness of the world from which these agreements emerge and the legal obligations to be silent that they create. In this article, I draw on a small interview study and on evidence gathered in response to the #MeToo campaign, to build a more rounded, contextualized understanding of the part NDAs play in society’s overall response to allegations and experiences of sexual harassment and other workplace misconduct. In doing so, I bring into focus the influence on this silencing of employing organisations, law in practice and wider social forces. This enables me to identify areas for further research and to make an initial argument for how organizations and their advisers should change their approach. My geographic focus is the United Kingdom, with some account taken of developments elsewhere. My arguments however have resonance for any systems in which employment relations and protective legislation are increasingly individualized and in which the use of NDAs in settlements is widespread. In terms of disputes, while sexual harassment is the starting point, the evidence and analysis encompass workplace misconduct more generally.⁵

In Section 2, I contextualize NDAs in settlements in relation to earlier socio-legal evidence and analysis about how law, legal process and working life interact to silence complaint and dissent. The relationship between this socio-legal perspective and non-legal theorizations of silencing is also explored. Having set this scene, sections three and four report on qualitative analysis of, first, exploratory interviews with a small sample of UK legal practitioners and, secondly, of oral evidence to the UK House of Commons’ Women and Equalities Select Committee’s Inquiries on Sexual Harassment in the Workplace and on The Use of Non-Disclosure Agreements in Discrimination Cases.⁶ This enables mapping of emergent knowledge in the United Kingdom about the use of NDAs. Section 3 concentrates on qualitative evidence about incidence, content and process, while Section 4 considers what drives the use of NDAs in settlements. In the final section, I relate the qualitative analysis to the broader context sketched at the start and draw out my conclusions for future research and more substantively.

2. SITUATING SEXUAL HARASSMENT, WORKPLACE MISCONDUCT, SETTLEMENTS AND NDAS

A. The Dispute Regulation Context and Silencing

Evidence that has come into the public domain in response to the #MeToo movement is important in part because there is limited existing scholarship about confidentiality obligations at work, in settlements or more generally.⁷ There is, however, growing knowledge and understanding of how people and organizations respond to workplace problems, which is the wider empirical context into which NDAs in settlements most immediately fit.

The concept of sexual harassment emerged in the 1970s in the United States⁸ after which sex discrimination law developed to encompass this behaviour.⁹ Acknowledgement of this phenomenon crossed the Atlantic and in 1986 the landmark case of Porcelli v Strathclyde RC¹⁰ found that sexual harassment could also be unlawful direct sex discrimination under UK law. In subsequent decades, prohibitions on sexual and other forms of harassment, related to identity and not, grew exponentially. For years now, the written UK law that proscribes harassment and other form of poor workplace behaviour has been wide-ranging, stringent and multi-faceted. That has likely both echoed and influenced developments in society, much as EU law has also been significant. A natural expectation was that this legal and social recognition would result in effective prevention and regulation of harassing conduct of sexual and other kinds.

Yet it has become abundantly clear that this has not eventuated.¹¹ A background explanatory factor is that recognition of sexual and other forms of harassment coincided in the United Kingdom with the twin trends of declining collectivism and growing individualization at work, both supported by legal intervention. Further this was against a political, social and economic backdrop in which liberal individualist approaches, often rooted in negative understandings of liberty, gained ascendancy over more collectivist ideologies. Recognition that harassment systemically enacts and sustains gender and other hierarchies at work therefore occurred at a time when individuals were increasingly expected, in both practical and ideological terms, to tackle workplace wrongdoing on their own.

The direct legal means by which that expectation was constructed involved protective workplace laws being legislated in the form of individual rights and within a framework that relied heavily on individual enforcement.¹² To this day, individual statutory rights and individualized enforcement are the bedrock of UK labour and equality law, notwithstanding the existence of some more systemic measures¹³ and provision for a degree of institutional enforcement.¹⁴ This state of affairs, moreover, is echoed in many other countries, albeit with a variety of approaches to the balance between individualism, individualization and collectivism, in law and more broadly.¹⁵

Well before the #MeToo movement gathered pace, however, evidence was mounting that individualist, individualized models of labour and equality law in general, and UK law in particular, have been ineffective to protect working people from wrongdoing, including but not limited to sexual harassment. My 2016 mixed methods study of the operation in practice of the broad range of UK protections against bullying and harassment drew on several strands of evidence to uncover and explain what is happening. Of particular relevance to this inquiry is the centrality of processes by which I found that working people are silenced:

The overall message from the data is that the more management dominates the working environment, the more meaningless in practice individual legal protections are for working people. This follows from the efficiency with which overt claims, complaints and dissent are silenced.¹⁶

Even in organizations that did not overtly shut down complaint, indeed that might even have encouraged people to speak out, my findings suggested greater vulnerability to individual legal conflict escalating in damaging, counter-productive ways. I extracted that recent regulatory trends have baked an incentive into UK workplace regulation for all organizations to mimic the most managerialist ones. If some employers take the opportunities provided by an individualist, individualized legal framework to avoid trouble with their workforce by deploying organizational power to silence dissent, an impetus is created for others to do the same in order to avoid comparative disadvantage. That study made clear that settlement practice is critical to this overall picture and the theorization that can be built from it. The #MeToo movement has in turn highlighted the importance of NDAs in settlements being fully brought into the analysis. These legal instruments need, however, to be viewed against that larger backdrop and evidence about responses to negative workplace experiences.

Individuals’ experiences of problems at work and attempts to do something about them are the counterpoint to the organizational perspective. United Kingdom evidence about what people tend to do, and how it tends to turn out, brings to life the myriad ways in which working people are currently silenced; the reality for vast numbers is not of voice, but of silencing. Large scale studies have repeatedly suggested that most who encounter a non-trivial difficulty at work respond in some way, doing things like seeking advice or information from a range of sources (including managers and trade unions), raising the matter with their employer and trying informally to sort things out. There are equally repeated indications that a very considerable proportion of this group (around half) do not achieve any resolution. Combined with findings that a significant minority take no action whatever in response to workplace issues, this points to an iceberg of problems at work about which complaint is effectively shut down either from the very beginning or when an employee gives up on pursuing the matter.¹⁷ This literature has not disaggregated experience of sexual harassment, but contains indications that a larger proportion of those who encounter sexual harassment take no action at all.¹⁸ Recent #MeToo testimony and research about sexual harassment supports this picture of individual experience and that it takes an even more extreme form in the sexual harassment context, including in the proportion who take no action.¹⁹

The emergent evidence about sexual harassment thereby suggests there is yet more silencing of these experiences than of workplace problems in general. However the general evidence is already stark, with only a small proportion of people who try to do at least something going on to consult the law and lawyers, an even smaller percentage commencing legal proceedings (in the low single figures) and a tiny minority ending up at a contested hearing. Investigations into the experience of litigation then portray this as highly burdensome, challenging and potentially itself harmful, with particularly disturbing findings about the difficulty of pursuing discrimination claims and the independent harm this can do, even where people win.²⁰ For those who get somewhere with their claim, the most likely outcome is settlement and, it now appears, agreeing to an NDA.²¹

That the best that individual enforcement of individual labour and equality rights typically achieves is an agreement to abandon a claim and to keep silent in return for money and possibly other things, like a reference, is arguably the ultimate means by which complaint is shut down. This is also the most dramatic silencing measure in imposing lifelong legal obligations to keep quiet, subject to limited exceptions, and often accompanied by non-disparagement clauses that inhibit any negative comment about either what happened or the employer. Wherever these settlement patterns hold true, they will echo back into the entire system of protective labour and equality laws: if a settlement with an NDA is the best that law and legal process normally offer those who invoke these rights at work, it appears fully rational on that basis alone for complainants to give up early on trying to do something, let alone once additional pressures towards silence are factored in.

My 2016 findings accordingly led me to theorize that the United Kingdom’s superficially protective framework of individual employment and equality rights, while having emancipatory features, ultimately reinforces workplace hierarchy. This theorization also has resonance for many other systems where individual rights and enforcement are central to the regulation of working life. At the structural level, the support to workplace hierarchy results from pitting working people, often on their own, against employers, almost always operating in some sense as collectives. As significantly, the law pushes working people into framing complaints in individualized terms that necessarily fit poorly with the essentially collective nature of work as an activity. There is then limited and diminishing support to pursue redress, including through legal constraints on collectivism at work in trade unions and otherwise, at the same time as law and legal process have been made ever more complicated and intimidating.²² What becomes surprising is that anyone ever does speak out. Viewed in that context it is the ‘icing on the cake’ that the usual endpoint for the relatively unusual instances of successful individual enforcement, rather than shining a light on where things have gone wrong and catalysing systemic change, is a settlement that shrouds what happened in silence.

B. The Societal Context and Silencing

Aside from this socio-legal work situating the use of NDAs in relation to problems and disputes at work more broadly, it points to wider influences on working people to stay quiet. The #MeToo movement illuminated just how powerful those are, reinforcing that the role of non-work, non-legal steers towards silence need to be brought into the socio-legal account of experience with NDAs.

Various societal pressures towards silence are suggested by theorizing about work that does not focus on group identity, except arguably in class terms. First is the sheer power employers have over those who work for them deriving from organizations’ systemic social and economic power, variable as the precise contours of this dominance are and the political ideology within which it nests. While this power is in many ways constructed, maintained and reflected by law, it also reproduces histories, traditions and timeworn patterns of social, economic and political hierarchy that transcend anything that law is able to produce. My own theorization of the functioning of individual labour and equality rights begins and ends with employers using this power to shut down complaint.²³ Second, there is employee reticence to speak out that is rooted in organizational loyalty and sense of belonging. Much as this subjectivity can also be constructed, not least by employer control of the discursive universe of work, the silencing effects of working people having positive feelings about their workplaces and careers should not be underestimated.²⁴ Third, attitudes to employment disputes and disputants create reticence to speak out. This might result quite logically from witnessing how those who make complaints and litigate are viewed and what they go through.²⁵ Rose and Busby’s exploration of how the UK state has shaped working people’s subjectivities about employment problems and complaints, further traced a link between state messaging and CAB clients’ feelings of shame and embarrassment about both having problems at work and taking action in the ET to seek redress.²⁶

Theorization of the gendered and racialized nature of society and working life also points to important causes of silencing at work that again seem stunningly resilient across time, space and political arrangement. The first for consideration is the shame that society attaches to being wronged. This has undoubted salience for women subjected to sexual harassment, including violence, and for racialized groups.²⁷ But shame spreads its tentacles widely, exemplified by Rose and Busby’s account of public discourse inculcating negative judgments for suing your employers. Ahmed commented on those who see the wrongness of sexual harassment against students despite participating in university cultures that make complaint difficult: ‘Guilt; shame; they can leak out, getting everywhere’.²⁸

The tradition of scholarship about how expression can silence, whether overt hate speech or subtler communicative mechanisms, also has potential resonance. Work in this tradition has focussed on how speech silences particular groups, like women and members of ethnic minority communities, but the effects may go further. Ahmed’s investigation into complaints about sexism and racism in universities pointed to the difficulty of complaint helping to maintain the gendered, racialized status quo.²⁹ This dovetails neatly with my socio-legal conclusion above that the operation in practice of the United Kingdom’s current framework of individual labour and equality rights ultimately reinforces existing hierarchies because of the silencing at the heart of how the system functions.

Finally, there are theoretical resources to be mined in feminist analyses of the practical and ideological work done by dividing the public from the private.³⁰ Again this has particular relevance for analysing the secretiveness that surrounds sexual harassment, but is potentially fruitful for more general understanding of why some workplace matters are seen as appropriate for silence and confidentiality and others for openness and debate.

C. The Socio-Legal and the Societal in Silencing

Combining these strands of scholarship provides a range of tools for analysing data about the use of NDAs, including as this varies and changes. Inter-relating insights from socio-legal work and social theory brings investigation of law and legal process, including empirically, into sociological scholarship, which has to date largely ignored this dimension of experience and research.³¹ At the same time, integrating these disciplinary perspectives has potential to deepen generalized theorizing about the role of law in working life, including to produce more finely grained insight into differential experience, and, in turn, how law and society interact.

3. SEXUAL HARASSMENT, WORKPLACE MISCONDUCT, SETTLEMENTS AND NDAS IN PRACTICE

That sets the empirical and theoretical backdrop for qualitative analysis of, first, exploratory interviews I conducted in 2018 with UK-based legal practitioners with extensive experience of workplace disputes, including sexual harassment, and, second, oral evidence from a wide range of witnesses, including again some lawyers, to the UK WESC’s Inquiries on Sexual Harassment at Work in 2018 and on Non-Disclosure Agreements in Discrimination Cases in 2019.³²

I interviewed six lawyers, who together had decades of experience in private practice with litigation and settlement of employment and equality cases, acting for both employers and individuals. Each lawyer was interviewed once, for variable periods, on condition of anonymity. The interviews were recorded, the recordings professionally transcribed and anonymized, with the resultant transcripts then checked by me against the original recordings, except for one interview of which I took notes only. I undertook the analysis using the checked transcripts and my notes.

I have in what follows interwoven qualitative analysis of these interviews and of the oral evidence to the WESC inquiries. The qualitative analysis of the lawyer interviews involved identifying repeated nodes in the transcriptions (or notes) and then combining these to extract over-arching themes in what was said. The oral evidence to the WESC was scrutinized for whether, first, it supported or contradicted those themes and, second, whether additional nodes or themes complicated the picture.³³ In this section, I cover what was conveyed regarding incidence, content and process in the use of NDAs in settlements.

A. The Ubiquity of NDAs in Settlements

Lawyers and others close to employment disputes emphasized that NDAs in settlement agreements are now commonplace, with only some sectoral variety and pockets of change in response to recent developments.³⁴ This comment from one of my interviews captured this: ‘I am trying to think of a settlement agreement that I’ve entered into which doesn’t have [an NDA] in the last 25 years, and I can’t actually think of any.’ The assumption that an NDA would be included was described as so strong that some lawyers, although possibly junior, seemed to think this was a legal requirement and another interviewee observed that: ‘It is just seen as standard. It is not seen as something that individually is negotiated…’ Evidence to the WESC by an individual litigant, Dr Emma Chapman, bore this out when describing a strong reaction to her request for the waiver of existing workplace confidentiality obligations (that she ultimately secured): ‘They came back saying, point blank, absolutely not: “NDAs are an intrinsic part of settlement agreements and we can’t move.”’³⁵ Reports of very rare situations where a settlement is concluded without a confidentiality clause supported the existence of this norm. Examples given were when the strength of a case gave someone ‘leverage’ and if an NDA was pointless because an ET hearing had already been held.³⁶

Subject to a few differences between employers and sectors, the routine inclusion of NDAs in settlement agreements was said to apply across the board. That was, first, in terms of the seriousness of the alleged conduct, exemplified by this comment from an interviewee: ‘I have dealt with people who have experienced indecent assault and rape at work, to the other end of just [being] undermin[ed] constantly.’ Second, there was the range of claims settled subject to NDAs, extending well beyond sexual harassment to ‘racial harassment, disability cases, people who’ve been made ill through the way that they have been treated at work…. Whistleblowers…. It’s the gamut.’

B. Settlement Routines That Encourage and Shape NDAs

Interviewees and witnesses described routine practices that both encourage NDAs and influence their content. First, employers, or their lawyers, were reported to provide the first draft and to draw this up to favour organizations. An interviewee said: ‘They’ll send us the [draft] that helps them most’ and there were descriptions of very one-sided clauses.³⁷ There was some dissension, however, about how often one-sidedness occurs and it may be that this is an area in which practice is shifting. Secondly, it was said that NDAs are invariably accompanied by a non-disparagement clause that obliges signatories to refrain from negative comment.³⁸ Wordings again apparently differ, for example as to whether obligations are imposed mutually and regarding how far they extend. Crucially, however, non-disparagement clauses were said to be as important a legal means to silence as NDAs.³⁹

Third, there is the influence of companies that provide precedent clauses and guidance to law firms.⁴⁰ My interviewees saw them as important in forming assumptions about the content of settlement agreements and therefore as a potential mechanism for change, for example to the one-sidedness of initial drafts. This however begs enquiry about how far such precedents are either impartial with respect to employers’ and employees’ typical concerns, or contain an in-built slant, most likely towards employers, since their lawyers will have the most resources to buy in such services.⁴¹ Still, interviewees pointed out the value of these organizations in avoiding lawyers ‘reinventing the wheel’ each time they give advice, addressing differences in skill between specialist and non-specialist lawyers and helping to establish benchmarks of reasonableness. One interviewee however noted that such precedents may discourage anything being removed, confidentiality clauses included: ‘[I]t’s human nature that when you have got something pre-prepared, you start with how you can add to it. You don’t think about what you take out.’

Fourth, while precedents about exceptions to confidentiality obligations appear influential, standard exclusions came across as both limited and difficult for a lay person to understand.⁴² The complexity of the whistleblowing exception made it hard even for experts to comprehend,⁴³ while the following captures the constraints on possible disclosures generally: ‘You’re [mostly] allowed to speak to your immediate family, as long as they’re going to keep it confidential. You’re allowed to speak to your professional advisors, lawyers, accountants. Usually, you are allowed to speak as required by law.’⁴⁴ Notably for learning about the use, causes and consequences of NDAs, it was confirmed that signatories would not be permitted to speak to researchers, even anonymously and confidentially. It was also said that lawyers would likely be cautious and dissuasive about participation in research.⁴⁵

Finally, although UK law requires employees to receive independent advice about a settlement (or compromise) agreement in which ACAS is not involved, an interviewee depicted the business model of lawyers who give such advice as leading to easy entry into NDAs and potentially onerous ones. The mandatory advice, which can in fact be given by lawyers or other qualified individuals,⁴⁶ is often paid for by the employer, apparently at a normal rate of around £200–500 per individual.⁴⁷ The employer’s lawyer cannot recommend a lawyer but they will suggest ‘a few people in town who do compromise agreements’. At that stage, this interviewee described a mutual interest for lawyers and the individual to make things go smoothly in order to ‘get this done’. Somewhat disturbingly, this was partly attributed to employee lawyers’ financial interests, in that this work stream would go away if they were really challenging. Hence, such employee lawyers were described as ‘not totally quiescent, they will say a few things, but up to the level of the agreed fee, because who is going to pay them after that?’ The expected advice about a proposed NDA was: ‘That is pretty standard; it is just what is done.’

C. Alleged Wrongdoers and NDAs

There were indications in my interview data, backed up by evidence to the WESC, that settlement agreements containing NDAs are not only concluded with people alleging wrongdoing, including harassment, but also with people against whom allegations were made. The existence of this phenomenon is strikingly absent from the public and policy discussion about NDAs such that evidence about it is particularly exiguous.

My earlier work pointed to organizationally imposed confidentiality obligations obstructing the resolution of problems and specifically to people against whom a complaint was made sometimes using confidentiality obligations, for example about internal procedures, ‘to intimidate complainants into silence.’⁴⁸ There are undoubted challenges in ensuring fairness to everyone involved, for example, where there are competing accounts of events. That said, there are indications of confidentiality agreements with alleged wrongdoers leading to several worrying types of silencing. First, Dr Chapman reported that her employer, UCL, entered into a settlement in a way that implied her complaint was, to some degree, unmeritorious, yet confidentiality obligations about internal workplace procedures prevented her from addressing that impression. Dr Chapman related very serious potential career ramifications:

As part of the [employer’s] settlement with him... they emailed my scientific collaboration to say that it had all ended and that there had just been a breakdown in the personal and professional relationship. The email clearly implied that there was equal fault.⁴⁹

As mentioned above, Dr Chapman finally secured a settlement of her consequent ET claim that included a ‘confidentiality waiver’. This enabled her both to speak out and to proceed with her career. The considerable challenge for Dr Chapman to avoid being silenced in this situation points to the urgency of finding out more about the silencing effects and harm done by settlements with alleged wrongdoers.

Second, situations were described in which alleged wrongdoers secured the imposition of non-disparagement obligations on colleagues other than the complainant, such that legal requirements not to say anything negative about this individual extended far and wide within an organization.⁵⁰ While it may in practice be difficult for employers to police adherence to such obligations,⁵¹ this remains another means by which even truthful, appropriate comment about an alleged wrongdoer is being legally constrained. Further, this practice has the specific potential consequence of allowing wrongdoers to move from job to job supported by references that cannot lawfully say anything negative about them, let alone mention allegations covered by an NDA.⁵²

D. Particular Fairness Controversies

Evidence about the incidence of unfair, even draconian confidentiality clauses, came across as disputed and uncertain, echoing unclarity about how frequently one-sided initial drafts are presented. Zelda Perkins’ evidence to the WESC portrayed an extremely onerous NDA, the provisions of which did not even allow her to have a copy and left her in fear that she would ‘go to jail’ if she committed a breach.⁵³ Evidence to the WESC⁵⁴ and the following quote from my interview study, however, depicted that particular NDA as unusual: ‘[B]etween mine and my partners’ practice…, [in] 45 years we have never come across a situation where we have been bamboozled into getting our clients to sign into anything like [that agreement], no matter how egregious the facts of the cases…. I just can’t think of any [lawyer] who has said, “Yes, that’s about right. I can’t see anything wrong with that.”… [That agreement] really is an outlier.’

Implicit in this comment is the importance to working people of having expert assistance, pointing to questions about the advice to which individuals have access in negotiating settlements and the NDAs within them. We have seen that the amount and quality of independent advice individuals receive may well be limited, despite being mandatory for non-ACAS settlements. One interviewee put this starkly in terms of the importance of the ‘cash nexus’ to the legal advice a complainant receives:

It is a business relationship… so it is not surprising that lawyers will only push back in circumstances where they can see a beneficial outcome in financial terms, and that includes for themselves, obviously. A partner will say to a junior lawyer who is becoming very ‘gung ho’ about a claim, ‘Who is going to be paying for all this?’’

Another interviewee perceived the danger in ‘the real imbalance of power there, where [as an employer’s lawyer] you think you have, you know, a person with no means, so you can get away with it.’ Dr Chapman’s evidence to the WESC poignantly captured what having good legal representation can mean when she said: ‘[My employer] asked whether I was sure I wanted to proceed… because they would go for me and they would use all their resources to fight the case if I took it to court. Had I not had a lawyer in my ear saying, “You keep going,” I would have been scared; I had everything to lose at home.’⁵⁵

Bulk negotiations of settlement agreements, for example where there is a mass redundancy, are a specific situation in which interviewees described disproportionate confidentiality clauses resulting from inequality of legal arms. This was despite the independent advice requirement and that the settlement might be for only slightly more than the statutory redundancy entitlement. It was described as common for people facing redundancy to enter into life-long obligations to be silent in exchange for just a few hundred pounds.⁵⁶ A lawyer interviewee described how this comes about: ‘[T]he employer might call lawyers who advise the individuals and say ‘This is the standard settlement agreement. We are not going to deviate from it. You can advise them either to agree to this or to litigate. We are not moving. We are not budging.’ And they mean it. So the poor old lawyer has to say to the claimant, “You either take this agreement or you litigate, and if you want to litigate, I need a couple of thousand pounds [on account of my fees].”’⁵⁷ These bulk settlements are another seemingly important dimension of current practice regarding which I have seen limited comment in the existing debate about NDAs in settlement.⁵⁸

Overall this data raises questions about the effectiveness of UK regulatory fairness safeguards in settlement negotiations and agreements, both in general and specifically regarding NDAs. Concern about this is augmented by the absence of any requirement for independent advice in respect of ACAS conciliated settlements when these form a large proportion of settlements⁵⁹ and frequently contain NDAs.⁶⁰ Since the ACAS representative is only a neutral conciliator, the effect is that individuals may well receive no advice at all before agreeing such settlements, including about the confidentiality obligations they contain.⁶¹ Even where independent advice is obligatory, there appears to be no guarantee this will be sufficient to withstand pressure to accept an NDA, ensure non-disclosure obligations are fair or that the individual has a good understanding of what they have signed up to.⁶²

4. CENTRAL THEMES AND RESEARCH DIRECTIONS

A. The Practice Dimension: Causes, Effects and Possible Futures

This data fills in detail about the use of NDAs, emphasizing and helping to explain how commonplace they are despite some indications of variety and change. In doing so, first, it calls attention to how routinization within organizations and in legal practice affects the use of NDAs and what they say. Second, settlements with alleged wrongdoers that contain NDAs emerge as a significant and largely unacknowledged aspect of organizational practice.

The analysis also points to the possible influence of technological and associated social changes that go wider than organizational life and legal practice. There were hints in the data that background factors may be important to the past and potentially future of NDAs as a legal tool for silencing at work. Observations by one lawyer interviewee in different ways raised technological change. First, they attributed the imposition of unreasonable NDAs in bulk redundancy processes to the easy availability and ever-increasing length of precedent documents: ‘That’s an example where I see this standard approach, print off the PLC precedent as a starting point, and then build on it to make it bigger.’ Second, they speculated that people are becoming enured to agreeing without demur to boilerplate legal text, however disproportionate, in order to get access to something. As this interviewee said:

Now, for most people, they are not bothered about what they have to sign to get the [settlement] payment. I think [in] the world around us, we have gotten used to every time we need to do something, [like you] need a new App on your phone, and to access that App you have to click okay to something that could be reams and reams of pages long, no one ever reads them. It’s just what you need to do to get that thing.

A question emerges whether the spread of NDAs, and potentially onerous ones, is one part of a larger picture in which day-to-day experience of technology is normalizing artificial, one-sided contracting.⁶³

Looking at this from the organizational point of view raises the possibility that technological and social changes of this kind are in some sense contributing to opportunism about things that employees might expect to receive in the ordinary course of working life and ‘for free’, like references. Dr Chapman powerfully adverted to this in her evidence when she said: ‘My gut feeling is that the things that are being given in exchange for an NDA should be given freely, and that we are exploiting the vulnerability of these people.’⁶⁴ This also calls to mind bulk settlement agreements in which interviewees described NDAs being ‘bought’ for very little.⁶⁵

A related possibility is that NDAs are entering into circulation in the context of one situation and being recycled for use in others. One scenario is where an employer has previously done a bulk settlement agreement exercise. As it was put to me: ‘[A]re the [employers] getting lawyers involved each time with something new? Or are they pulling something off the shelf?’ Another, more troubling possibility, was raised by reports that agency workers taken on for the Presidents’ Club dinner were required to sign NDAs which prohibited disclosure of anything that happened at the event and that were probably unenforceable. As an interviewee said: ‘I don’t know the full details of what has come out about [the Presidents’ Club scenario]…. What went through my head was, “I wonder if that’s something that [an] agency had on the shelf. Maybe there was something about confidential information and they added in an extra bit.”’ Is there another possible social change in evidence here, whereby onerous working arrangements of dubious legality are circulating widely, often unscrutinized by those subject to them, let alone by lawyers?

B. The Justice Dimension: The Individual/Collective Dilemma and Possible Futures

The other major theme picks up where Section 3 left off, in raising the fairness or otherwise of settlement negotiations, ultimate agreements and NDAs within them. This strain came through in employers’ and lawyers’ assumptions that NDAs form part of settlements, questions about the extent to which first drafts favour employers and controversy over whether standard exceptions to confidentiality obligations go far enough. It culminated in uncertainty about the incidence of oppressive NDAs and worries about systemic imbalances of power and resources in both individual and bulk settlement negotiations, notwithstanding requirements for employees to have independent advice before agreeing to non-ACAS conciliated settlements. A recurrent motif was of conflict or tension between the concerns and interests of individual complainants, their employers and wider society in how workplace allegations and wrongdoing are addressed.

(i) NDAs as Obstacles to Systemic Action—and Glimpsing Change

One lawyer interviewee encapsulated this conflict in describing ambivalent feelings about her own long experience of representing claimants in discrimination cases, although with some optimism for the future because of ripple effects from the #MeToo movement:

[I] went into this area of work because I thought, “I want to be a discrimination lawyer because I want to make a difference in the world.” … [W]hat became increasingly frustrating… was that actually, I was seeing the same things happen in the same places, and on a one-on-one level achieving really good results for people, who had been through an awful lot and wanted rid of it, and wanted a good deal, to move on…. Not only had I not made any difference [more generally], but on top of that, if anything, I was the one signing these confidential settlement agreements…. On one level thinking, 'Rather than doing anything about it, I am kind of complicit, because I can see [the discrimination].'

This captures that NDAs in settlement agreements enable organizations to condone sexual harassment and other workplace wrongdoing and, at the same time, inhibit systemic measures to address such behaviour.⁶⁶

NDAs in settlement agreements were depicted as a mechanism which organizations routinely use to keep known, perhaps serial, wrongdoers in place, often because of their perceived financial value and sheer organizational power.⁶⁷ As the interviewee above poignantly said about past feelings of disillusionment: ‘How many agreements, how many cheques? They are coming from the same banks, and the story is the same. It’s just the name of the individual [complainant] that’s different.’

Dr Chapman’s evidence to the WESC addressed how NDAs, including with wrongdoers, impede specifically universities from addressing systemic problems:

[W]hen justice does not occur… that is kept quiet as well, so it removes all accountability on the institution’s part for the safeguarding of their staff or students.⁶⁸ You cannot solve a problem until you expose a problem, and NDAs are being used such that we cannot actually expose what is happening in universities...’⁶⁹

Another interviewee summed up this theme: ‘Behaviour would be hugely different, I think, if there were not NDAs’, arguing both that individuals would be more careful and that organizations would take effective measures to prevent wrongdoing.⁷⁰

There were nonetheless indications of some change to organizational practice.⁷¹ The hopefulness adverted to above was explained as: ‘[P]eople aren’t accepting that this is how things should be, that it’s just the weather…. These things then, when they come out, are having repercussions.’ One lawyer related that now if there was ‘something really obvious and awful’, then ‘I think you get it sorted without [an NDA]’. Interviewees even reported instances of settlement without an NDA being publicized by the employer, implying that protecting organizational reputations might now require not gagging individual complainants.

It was also said that more employers are seeking advice about improving their processes for addressing complaints. The technical legal point was raised that this may enable employers to defend claims for vicarious liability for discrimination, including harassment. This has been a legal option for decades where an employer can show they took reasonable steps to prevent discriminatory conduct.⁷² The fact that this defence has not much been invoked before now⁷³ begs questions about how likely it is that organizations will make more use of it in future.⁷⁴ Still, one interviewee saw a ‘massive cultural change’ in more organizations taking the stance that they deal properly with allegations and leave individuals against whom these have been made to get their own legal advice. So this might mean saying: ‘If you have done something, you might have to face your individual liability on that, and you will front those costs to defend that, because that’s not by dint of what we have authorised…’

Important as these seams of evidence regarding change are, they were quite sparse⁷⁵ and sectoral, for example in the Civil Service. Comments about change also provided vignettes of the norm from which some organizations might now be deviating. There is the telling contrast with when settlements about even obvious and awful things would automatically have contained an NDA. Equally, that a greater proportion of organizations ‘deal[ing] properly’ with allegations was characterized as a ‘massive cultural change’ is evocative of the prevailing approach.

Dr Chapman’s evidence of what she has been able to do because UCL waived confidentiality obligations also suggests that the normal position remains that complainants are stopped from speaking out:

I campaign in a group called The 1752 Group to stop this kind of thing happening, and I am able to speak to Universities UK, for example, and University and Colleges Union very frankly, and give them evidence… I am one of the few people who can actually say, no, this does actually happen; it is not just hyperbole; here are these in black and white. That has been very powerful in getting UUK, for example, to convene a taskforce on promoting professional guidelines…⁷⁶

The sobering point is that actions of this kind are not possible for the many whose silence is legally compelled.

(ii) NDAs as Protective of Complainants—and Glimpsing Change

Still, it came across equally powerfully that NDAs can be protective of individuals who bring and persist in complaints to the point of settlement, including who have been very seriously wronged and harmed. This protection was depicted as having many facets. Yet, all can be related to workplace sexual harassment and other misconduct being framed, including by law and legal process, as an individual employment matter rather than a societal concern.⁷⁷ The practical effect is that the current system pits the safeguarding of individuals, against the wider societal aim of combating workplace wrongdoing, with employers the apparent beneficiaries of the silencing that results. Yet any employer benefit is of dubious value given the systemic negative effects of secretiveness about complaints for workplace culture and behaviour and the attendant failure effectively to address workplace misconduct.

The settlement sum increases with an NDA

Most starkly, it was made clear that the money paid on settlement increases if there is an NDA: ‘You know what, because [confidentiality] is so valuable to the employer, they’re willing to pay extra for it.’ Related is that the possibility of concluding an NDA may be how internal opposition is overcome, for example, from the person accused or a board member opposed in principle to settling such claims: ‘[Y]ou may have more moderate voices on the board or in management, who are trying to persuade that person. They may use confidentiality as a way to get that kind of aggressive litigious person onside to settle with the claimant.’

The corollary is that without an NDA, complainants have less ‘leverage’ in settlement negotiations.⁷⁸ Indeed, the wholly logical but somewhat ironic point was made that the legal limits on restricting public interest disclosures (or ‘whistleblowing’) can reduce a complainant’s negotiating power precisely because employers have less scope legally to buy silence.

An important part of the landscape is therefore that the settlement sums paid to complainants may be more, possibly significantly more, than complainants would have received if they had gone to court. Indeed, the claimant might have lost if the claim had proceeded. If anything, however, this raises even more questions about how silence is being transacted over⁷⁹: it is not just that it is being traded for things that employees should arguably be given ‘for free’, but that silencing has become such an organizational reflex that lifelong confidentiality is being purchased with scant regard even to the viability of the underlying legal complaint, let alone to the overall benefits and costs for employers, individuals and society.

There are multiple good reasons for complainants to settle, and with an NDA

The next clear theme was of there being myriad reasons, irrespective of the size of any settlement payment, for complainants to want to settle and to do so subject to an NDA. Interviewees depicted those reasons as explaining why in their experience the vast majority of their clients either actively desire settlement and confidentiality or become resigned to this.

I was vividly told of many considerations favouring settlement:⁸⁰ the financial pressure over time of being out of work; anger being replaced by a desire to move on; that complainants have already ‘gone through the mill’ before they get to lawyers, then receive warnings like: ‘The road you are going to travel on is going to be long and hard and stressful and costly… The truth may out, or it may not.’ The following arguably encapsulates the dynamic complainants find themselves in: “This awful thing happened to me. What I really want is for it not to have happened to me. But I don’t want the fact that it happened to me to make things worse.” In a telling phrase, one interviewee described individuals viewing settlement, even with an NDA, as at least providing ‘some form of justice’.

Equally, there were cogent arguments for wanting NDAs: ‘It’s the worst episode of their life, and they actually want the whole thing shut up.’ Striking accounts were given of complainants carrying deep feelings of shame and being unable to talk about what happened even with family and friends:

[W]ith many people, things that happen at work which they can’t control can be quite shameful, can really undermine their self-confidence and self-esteem, and they don’t want to be discussing it with everybody, not even their own partners.

The indignity and humiliation specifically of being discriminated against were described as making it hard for those experiencing this to admit it to themselves, let alone to anyone else.⁸¹ The point was also made that shame might be experienced differentially by members of different ethnic, religious and cultural groups. Another interviewee stressed how different people are, relating the potential impact of sexual harassment which some might not see as serious: ‘It’s like a trauma in the workplace. That’s real. That’s not invented. That’s not false. For someone to come and even tell a lawyer about what they’ve gone through, it’s an emotional thing to unpack that shame.’

Finally, there were powerful comments about complainants’ concerns for their careers, for example wanting to be known for what they had done in their work, ‘not for having brought a complaint’. The quite recent creation of a searchable database of ET judgments has augmented this pressure on complainants to settle with an NDA, since it means prospective employers (and others) can easily check if an applicant has sued a previous employer. As an interviewee said, potential litigants need to think hard about what they want to have ‘out there in the public domain for perpetuity on Google’.⁸²

NDAs shield individuals from the burdens and risks of speaking out

This theme moves beyond evidence about why complainants want, or at least come to want, settlement with an NDA, to ways that NDAs really do protect individuals from the burdens and risks of speaking out. These ranged from the observation that, much as winning a claim can be empowering, it may well leave a claimant struggling to find another job because of potential employers thinking “Oh, you’re a bit difficult, you’re a bit too feisty.” There were also strong statements, echoing other research findings about what claimants are put through in the course of litigation: ‘[T]ypical tactics are to destroy the credibility of the individual victim.’⁸³

Views about the impact of changes in the use of NDAs that interviewees had either observed, or that they anticipated, reinforced this theme. Interviewees foresaw that employers would be more likely aggressively to litigate if NDAs were not readily available. The logic behind this was encapsulated as follows:

Yes, there’s a real danger… if we don’t have the protection of these types of clauses, that employers think there is nothing to lose, let’s go for it, and [they] might be well advised to do so. “Let’s clear our name and besmirch this other person.” The fear of that is also, not even if [employers] do it, [but] if they [only] start down that road, then this individual may well back out and come away with nothing.

On this stark vision, neither would wronged individuals achieve the ‘form of justice’⁸⁴ that settlement with an NDA provides nor would there be any societal gain. Rather the phenomenon that was documented in my earlier project, and encapsulated by an interviewee here as employers using ‘their might and their resources’ to silence complaint and ‘scare away’ complainants, would be pursued in yet more ways. It is also noticeable that this speculation about what an NDA-less future would be like, contradicts earlier evidence of cultural change towards dealing properly with allegations, in rather positing organizational readiness to engage in yet more aggressive repression of complaint. This is further cause to be sceptical of how much change is really happening.

It was nonetheless pointed out that requiring settlements to be public would serve the wider interests of justice so far as it led to unmeritorious complaints being deterred. The broader evidence about workplace disputes introduced in Section 2, however, suggests that alterations to employers’ benefit/loss calculus regarding settlement would likely have implications for all sorts of claims. If employers come to regard themselves as better advised to use more of their resources to ‘scare away’ complainants, this would likely be the case independently of the merits of any claim.

By the same token, these different, to some degree contradictory observations,⁸⁵ emphasize how little is really known about organizational reactions to recent NDA controversies, let alone about what will happen as the debate moves on. It is also noticeable that nothing was said by my interviewees about the possibility of practice simply morphing into adding other tools, legal and of other kinds, to employers’ armoury for silencing complaint, which did come up in the WESC oral evidence.⁸⁶

(iii) Probing the Long-Term Effects of NDAs

A final theme involves hints in the data that a longer-term analysis of the effect of NDAs in settlements might uncover important consequences for individuals, organizations and society of which we are currently ignorant. This makes a compelling case for extending the time frame for assessing the impact of these agreements.

As regards individuals, lawyer interviewees simply did not know how clients felt over time about having signed an NDA because they did not stay in touch. One lawyer raised a possible analogy with the long-term consequences of being silenced for child abuse survivors. It was also mentioned that calls sometimes came in from individuals expressing regret at not being able to speak out.

Evidence to the WESC highlighted the ongoing impact NDAs can have.⁸⁷ An interviewee raised this in considering why people have joined the #MeToo movement, noticing that at times it seemed easier to disclose an experience on social media (rather than to loved ones) and in order to help other survivors. Finally, I was told:

[T]here is another set of motivations, which is around a kind of almost cathartic, being able to let go of that horrible secret, and to feel like you will be accepted for it. It’s kind of breaking a taboo, which is a bit like [the] ‘80s, ‘90s when people were coming out about their sexuality. That actually that thing you felt you would be potentially stigmatised for, you can say. You can be the person you really are with what happened to you known, understood and accepted about you. You can let go of it after all.

This interviewee added that for legal obligations to be added to so many powerful social mechanisms that silence people ‘is pretty huge.’ This points to quite profound questions about the existence, nature and extent of individual consequences as time passes and the broader social effects of those being hidden from view.⁸⁸

From an organizational perspective there is the equal and opposite question of how silence about allegations and related internal processes affects the working environment as time passes, and how these effects echo back into society. Dr Chapman’s evidence to the WESC was again illuminating about long term difficulties for employers:

I think… [UCL] realised that maybe it was good for them as well to get me to speak out, because I do not believe that universities are happy with how these things are going; they cost a huge amount of money and time…. I think [universities] need help, and by speaking out I am almost helping them…. The reason I have a good working relationship with UCL is that it is made up of individuals who want to do the right thing and who I like very much, but their hands are tied by discriminatory policies that were written 20 years ago.⁸⁹

Notably this implicates the neglected phenomenon of settlements being made that shield an alleged, indeed maybe known, wrongdoer, and specifically of NDAs being entered into with such people.

5. CONCLUSIONS

Relating my qualitative analysis of lawyer interviews and WESC oral evidence to the broader theoretical and empirical background sketched earlier leads to several conclusions. First, the evidence about NDAs strengthens the argument that silencing of complaint is central to the functioning in practice of individualist, individualized models of employment and equality law in general, and UK law in particular. It follows that the current use of NDAs is critical to the failure of this model of law effectively to combat workplace misconduct and to seeing how it ends up overall supporting established workplace hierarchy. The evidence was eloquent as to quite how far law and legal process have normalized NDAs in settlements and expanded their substantive content.

This adds granularity to understanding of the role of law and legal process in shutting down allegations and experiences of sexual harassment and other workplace misconduct, while my second set of conclusions concern interactions between legal influences on silencing and broader social, economic and political forces. Lawyering itself emerged as potentially in the grip of technological and social developments that are encouraging the spread of oppressive boilerplate legal obligations, with multiplying, lengthening NDAs part of this larger phenomenon. The proliferation of NDAs in response to individual discrimination complaints evokes Ahmed’s depiction of the ineffectual development of race equality action plans in universities. Both processes create vast ‘paper trails’ by which organizations avoid tackling enduring inequality.⁹⁰ It is, however, conspicuous that this echo exists despite the action plans responding to the 2000 race equality duty. That was a pioneering equality law measure in moving the focus away from individuals and towards catalyzing organizations to pursue systemic equality goals. That there is this parallel not only raises that modern, technologically inflected forms of routinization and sheer paperwork may be significant to repeated failures of law and organizational life effectively to address inequality, but that this phenomenon might be resistant or impervious to innovation in the design of the surrounding law.

The vision into the world of NDAs also provided fine-grained detail of employers’ social and economic dominance and its mediation by lawyers. This was present in accounts of NDAs being insisted on for all settlements, from the entirely trivial to the extremely serious, and of employers leveraging their superior legal resources and ‘repeat player’ experience in disputes and negotiations. Equally, the structural weakness of complainants was clear from how protective NDAs were described as being for them. The corollary was the riskiness the data suggested of individuals pursuing allegations owing to, for example, employers’ adversarialism in the face of complaint and career damage from being known to have made a claim. Equally, evidence about complainants’ own determination to preserve their careers spoke to their vulnerability when making decisions about settlement and the discursive pressures on them, not least as mediated by the state through the creation of a publically available register of ET cases.

Further, how employers exert their power to maintain established hierarchies, not least in identity terms, was illuminated by descriptions of organizations using NDAs to shield alleged wrongdoers, often senior and even when known and serial, and to accept the side effect that wider organizational causes and effects of workplace misconduct would go unaddressed.⁹¹ Indications specifically of these transactions fortifying gendered, racialized and other identity-based hierarchy came from the mere fact that sexual harassment and other discriminatory conduct were central to accounts of the allegations routinely covered by NDAs, much as other types of misconduct were also implicated. In addition, repeated observations about feelings of shame, particularly regarding discriminatory conduct, were arresting. These feelings were not described amongst those against whom allegations were made,⁹² nor even in organizational representatives, but amongst complainants. This echoes empirical findings alluded to earlier⁹³ and theorizing across times and places of the role of shame, and related emotions like humiliation, in producing and maintaining gender, racial and class stratification.⁹⁴ The suggestion is that widespread use of NDAs to silence is producing a toxic interaction between long-standing social, economic and cultural means by which groups are subordinated and the operation in practice of legal measures supposed to combat this.⁹⁵

The hypothesis that I extract from this examination of the use of NDAs is of a developing symbiosis between legal and non-legal mechanisms to silence complaint at work that amplifies employer power and reinforces gendered, racialized and broader inequalities. This makes an urgent case for further enquiry into where NDAs in settlements, and related confidentiality obligations, fit into society’s overall response to experiences and allegations of sexual harassment and other workplace misconduct, and into what ought, from a normative and practical perspective, to be done with law and other tools. Crucially, this would not only illuminate the place of law in these processes, but also deepen understanding of the non-legal forces at play.

Notwithstanding this assessment of the current position, my third conclusion is about the allusions to variety and change. There were some indications that the ubiquity and extent of NDAs in settlement agreements for many employers itself represents a change,⁹⁶ as well as references to organizations or sectors that already eschew NDAs and suggestions that others are, or may be, altering their approach. It is, however, critical not to assume that all change is in the direction of less silencing. Rather it is highly predictable that employers will instead find other tools, including legally, to keep shutting down complaint. In the US context, for example, mandatory arbitration provisions in standard form employment contracts, incorporating NDAs and non-disparagement clauses, are a powerful legal mechanism of this type. The failure mostly of recent US reforms about NDAs to proscribe such contractual clauses, including owing to federal legal constraints on doing so, is a good example of how even apparently progressive legal reform may be wholly consistent with widespread continued silencing by legal means.⁹⁷

Even so, variety and fluidity in how organizations and individuals are behaving following #MeToo provides scope to excavate the drivers and outcomes of different approaches. In research terms, my analysis points to the particular importance of using this period of change to illuminate the legal and non-legal influences on what different organizations are doing and, on this basis to work out how the innate tensions between individual, employer and societal interests in silence or exposure can be reconciled. Regarding the legal dimension, the evidence made clear that investigating the part of lawyers and their professional regulators in the production of change and stasis is crucial, but also the value of this being nested in appreciation of the wider social, economic and cultural influences at play. Additionally, the consequences of NDAs for individuals, employers and society need to be uncovered well after any individual complaint has ‘gone away’. In truth we have vanishingly little knowledge about how NDAs are experienced over time, amidst increasing indications that the long term effects can be significant.⁹⁸ A specific hope would be that recent changes following #MeToo would make it possible more systematically to bring the testimony of those subject to NDAs into future research, the methodological challenges of doing so notwithstanding.⁹⁹

Fourth, I want to conclude by considering if initial conclusions can be reached about what ought to be done absent additional research and irrespective of the regulatory landscape. In particular, the question is if there is a way to sidestep the dilemma my analysis has revealed as inherent to the current state of affairs. This is that, from this starting point, introducing strong legal restrictions on the use of NDAs in settlements would add yet another burden to the already disproportionate load carried by the few individuals who reach the stage of settling a complaint, by requiring them at the final stage to put collective, societal interests in the exposure of workplace wrongdoing above their own. At the same time, limited legislative and regulatory measures appear equally unpromising for effecting change, for not striking at the basic legal and social structures that I argue are interacting so powerfully to silence complaint.

Still, even accepting that organizations which are reconsidering their use of NDAs are probably islands in the sea of those still routinely using them, and that there remain all sorts of other legal resources that may perform a similar function to NDAs, the indications of change suggest that even now, it is possible to do things differently. And if this is true for any organizations, it is true for all, law firms and barristers’ chambers included.

It is important here that, just as the data repeatedly adverted to divergence between societal, collective concerns in workplace wrongdoing being exposed and the interests of individual complainants, employers and their lawyers, it also strongly suggested that it is organizations’ concerns which are pivotal to how NDAs are being used. It follows that individuals’ undeniable interests in settlements with NDAs are significantly created by organizations’ choosing to exert their power in particular ways, often mediated by legal advice.¹⁰⁰ The indications of variety in turn suggest that some organizations are opting to do things differently, with some lawyers supporting that different stance.¹⁰¹ The ineluctable conclusion is that there are other ways for employers to analyse their interests, for example to take account of the reputational risks from insisting on confidentiality, to factor in the longer term damage that widespread use of NDAs may do to workplaces, and, less instrumentally, to pay attention to the ethics and morality of silencing complaint at work.

The idea this leads me to float is that organizations, backed up by their lawyers, should make a simple change. This is to move away from a default position of routinely using NDAs and other legal tools, in settlements or elsewhere, to silence allegations and experiences of sexual harassment and other workplace misconduct.¹⁰² This would be a limited shift, given all the social¹⁰³ and legal factors that already stop people speaking out. Certainly the law would continue in multiple ways to require extreme caution in making allegations, for example the law of defamation. Agreement in a given situation to some kind of legalized silencing, including via an NDA in a settlement, would also remain open.

However slight a change on the surface, however, for organizations and lawyers to oppose routine legalized silencing of complaint would be a shift with potentially far-reaching implications. That is a further reason to find out more about workplaces which are already either doing or contemplating change in this direction, in the learning this would deliver about how to make that work.¹⁰⁴

Most fundamentally, my evidence and analysis point to this being a useful starting place for challenging the culture, inside and outside workplaces, that constructs complaint about, and experience of, sexual harassment and other negative workplace conduct as too problematic to talk about, needing to be closed down and any wider implications ignored. The routine shunting of allegations of such conduct into quasi-private, individualized processes, hedged around with confidentiality obligations and culminating in life long legal obligations to be silent, is central to the maintenance of that culture. The mere assumption that organizational exchanges about these matters should occur in private ascribes shameful, negative connotations to whatever occurred and to it being spoken of. Ahmed’s study in universities also drew this link: ‘A complaint is made confidential as soon as it is lodged, so all of this happens behind closed doors; a complaint as a secret, a source of shame, what keeps you apart from others’.¹⁰⁵ However much it has come to seem uncontroversial to organizations, encouraged by their lawyers, in turn to take every chance to give that secrecy the force of law, the problematic attitudes that underlie this approach need to be questioned all the way down. I argue that removing organizational defaults into legalized silencing, whether with NDAs in settlements or in other ways, is the obvious first step towards disrupting the set of norms, beliefs and behaviours that maintain this taken for granted approach, with direct consequences at work and the potential for these to ramify more widely.¹⁰⁶

The longer term hope would be that this change would stimulate thought, creativity and experimentation about how allegations and experiences of sexual harassment and other workplace misconduct can safely, appropriately be aired, learned from and productively responded to, not only individually, but systemically. Indeed, this includes where allegations are determined to be partially or wholly unfounded. The role of lawyers in what currently happens suggests they would have a critical role to play,¹⁰⁷ not least in applying their technical ingenuity to make such change work in legal terms, recognizing the undoubted challenges, including of ensuring fairness towards everyone.¹⁰⁸

Specific possibilities include innovation in the means available to both individuals and groups to raise issues, allowing this to be done on behalf of others and providing support for those against whom allegations are made. Equally, there could be wider participation in the organizational response, especially to orient responses to addressing systemic problems that individual issues raise. This would need to involve not only trade unions, but also other entities like civil society actors and internal representative structures, in order to provide for workplaces which are not unionized. Regulators and sectoral, industry level bodies are additional potential partners in such change, with, for example, potential to build on the reputational and other damage that widespread use of NDAs can do to argue for change. An example is Can’t Buy My Silence’s campaign for universities to sign a pledge not to use NDAs in cases of sexual misconduct and bullying, backed up by Michelle Donelan, current UK Minister for Higher and Further Education.¹⁰⁹ Organizations could also themselves be active in finding out where misconduct is occurring and in taking the initiative to address this despite no complaint being made. Various technologies which are in use, including third party reporting schemes and staff surveys, might all help to put this into practice. At all stages, the assumption should be against confidentiality obligations being imposed. Even where it is found that an NDAs is needed, consideration could be given to other ways of airing issues, for example by the use of anonymity¹¹⁰ rather than confidentiality.

Ultimately, we have learned from the #MeToo movement and its aftermath that employers, individuals within them and their lawyers, have agency and choice in the approach they take to sexual harassment and other workplace misconduct. This indicates that, whatever else happens, there is the capacity through a coalition of effort, inside and outside organizations, to disrupt the legal and wider forces behind the acceptance of widespread legalized silencing of allegations and experiences of sexual harassment and other workplace misconduct. Further, taking this path holds the promise of finding new ways of responding that fairly balance the interests of everyone and of beginning to construct more equal working lives. If employers are not willing to take even this small initial step, and their lawyers to back them up, this will raise serious questions about the many current public commitments finally to tackle sexual harassment and related misconduct at work, as well as grave doubts that anything much is going to change in this dimension of working life

Thanks to the interviewees who kindly gave their time and shared their knowledge, to participants at an Industrial Law Society evening meeting in November 2021 for their insights, to Jeremias Adams-Prassl, Nicole Busby, Ruth Dukes and Emily Grabham for comments on an earlier draft, and to Saphié Ashtiany and Kate Malleson for many conversations about this work and for assistance in carrying out the interviews. Remaining errors are my own. All noted urls last visited on 21 February 2022.

Footnotes

¹ 

The following are some indicative examples, but note that several address the use of NDAs in the context of a larger inquiry into sexual harassment: M. Garrahan, ‘Harvey Weinstein: how lawyers kept a lid on sexual harassment claims’ Financial Times, 23 October 2017; R. Moorhead, ‘AO, AO, AO: Weinstein’s men and the long arm of the law’, https://lawyerwatch.wordpress.com/2017/10/24/ao-ao-ao-weinsteins-men-and-the-long-arm-of-the-law/, 24 October 2017 & ‘Professional Ethics and NDAs: Contracts as Lies and Abuse?’ in P.S. Davies & M. Raczynska (eds), Contents of Commercial Contracts, Terms Affecting Freedoms (Oxford: Hart Publishing, 2020); Equality and Human Rights Commission (EHRC), Turning the Tables: Ending Sexual Harassment at Work (March 2018); M. Townsend, ‘Ex-Weinstein assistant calls for ban on contracts to silence harassment victims’ The Guardian, 26 August 2018; ABC & Others v Telegraph Media Group Ltd [2018] EWCA Civ 2329 (injunction granted to stop publications that would reveal information allegedly in breach of settlement NDAs about workplace harassment. The identity of the alleged harasser was subsequently revealed in Parliament.); J. Kantor & M. Twohey, She Said: Breaking the Sexual Harassment Story that Helped Ignite a Movement (London: Bloomsbury Circus, 2019); Linklaters LLP & another v Mellish [2019] EWHC 177 (QB) (injunction granted to stop ex-employee disclosing information about gender equality allegedly in breach of a confidentiality clause in their contract of employment); BBC Radio 4 Programme, ‘Contracts of Silence’, 20 March 2019, https://www.bbc.co.uk/programmes/m0001hwr; WESC, Sexual Harassment in the Workplace (House of Commons, July 2018) (S/H inquiry) & The use of non-disclosure agreements in discrimination cases (House of Commons, June 2019) (NDA inquiry); Australian Human Rights Commission, Respect@Work: Sexual Harassment National Inquiry Report (2020); A. Mohdin, ‘BBC may waive gagging clauses of previous equal pay settlements’ Guardian, 18 January 2020.

² 

See Department for Business Energy and Industrial Strategy (BEIS), Confidentiality Clauses, Response to the Government consultation on proposals to prevent misuse in situations of workplace harassment or discrimination (Crown Copyright, July 2019), pp 8–16. The proposals, which have not yet been implemented, are to prevent NDAs in settlements prohibiting disclosure to ‘the police, regulated health and care or legal professionals’, to require NDAs to spell out when disclosure is permitted, to extend the requirement for independent advice to cover exceptions to non-disclosure obligations and, finally, for enforcement of these rules.

³ 

Solicitors Regulation Authority (SRA), Warning Notice, The use of non-disclosure agreements (first published March 2018 and updated Nov 2020); The Law Society, Practice Note on non-disclosure agreements and confidentiality clauses in an employment law context (Jan 2019 and updated Dec 2019); Equality and Human Rights Commission, The use of confidentiality agreements in discrimination cases (Oct 2019) & Sexual Harassment and Harassment at Work: Technical Guidance (2020); ACAS, Non Disclosure Agreements Guidance (ACAS, Feb 2020). See ibid., BEIS, 11-14 regarding EHRC and ACAS guidance on the wording of NDAs and updating of the SRA Guidance.

⁴ 

In July 2021 the Government Equalities Office (GEO) announced various reforms to the law on sexual harassment in Consultation on sexual harassment in the workplace: government response (Updated 21 July 2021), with linked research publications, L. Adams, L. Hilger, E. Moselen, T. Basi, O. Gooding & J. Hull, IFF Research, 2020 Sexual Harassment Survey (GEO, 2021) and L. Adams, L. Hilger, E. Moselen, N. Morrice, O. Gooding, and A. Karadia IFF Research, Literature review of sexual harassment in the workplace (GEO, 2021). See also the global campaign launched by Zelda Perkins and Julie MacFarlane, Can’t Buy My Silence https://cantbuymysilence.com/ and, in particular, legislative developments in Prince Edward Island, Canada, the Republic of Ireland and the United Kingdom (in the form of a Private Members Bill).

⁵ 

That the misconduct concerned extends beyond harassment is exemplified by concern about the silencing of women regarding ill treatment related to pregnancy and maternity. See the wide framing of the WESC, NDA inquiry (n.1), the NDA campaign by Pregnant then Screwed https://pregnantthenscrewed.com/campaigns/ and L. Bates ‘Pregnant then screwed: how gagging contracts are used to silence sacked mothers’ The Guardian 22 January 2019.

⁶ 

(n.1).

⁷ 

General work about civil justice to some degree considers confidentiality in ADR processes and settlements, in particular bringing attention to problems with privatizing justice. The classic early US analysis is O. Fiss, Against Settlement (1984) 93 Yale Law Journal 1073 and in the United Kingdom see the work of H. Genn, including Hard Bargaining: out of court settlement in personal injury actions (Oxford: Clarendon Press, 1987); Judging Civil Justice, 2008 Hamlyn Lectures (Cambridge: CUP, 2010); & ‘What is Civil Justice for – Reform, ADR, and Access to Justice’ (2012) 24 Yale Journal of Law and the Humanities 397. Socio-legal investigations specifically into the operation in practice of workplace employment and equality rights also touch on confidentiality obligations in settlements, including my own work on which this study builds, L. Barmes, Bullying and Behavioural Conflict at Work, The Duality of Individual Rights (Oxford: OUP, 2016), as well as E. Rose & N. Busby, ‘Power in Employment Disputes’ (2017) Journal of Law and Society 674 & N. Busby & M. McDermont, ‘Fighting with the Wind: Clients’ Experiences and Perceptions of the Employment Tribunal’ (2020) 49 ILJ 159. There is also relevant research about other equality law systems, eg, D. Allen ‘Against Settlement? Owen Fiss, ADR and Australian Discrimination Law (2009) 10 International Journal of Discrimination and the Law 191 and ‘In defence of settlement: Resolving discrimination complaints by agreement’ (2014) 14 International Journal of Discrimination and the Law 199 on confidential conciliation processes in Australia.

⁸ 

L. Farley, Sexual Shakedown: The Sexual Harassment of Women on the Job (New York: McGraw Hill, 1978).

⁹ 

C. MacKinnon, Sexual Harassment of Working Women: A Case of Sex Discrimination (New Haven, CT: Yale University Press, 1979) and subsequent legal developments.

¹⁰ 

[1986] ICR 564 (CSIH).

¹¹ 

On workplace sexual harassment specifically, see: TUC, Still just a bit of banter? Sexual harassment in the workplace in 2016 (TUC, 2016); Sexual Harassment in the Workplace (2017, ComRes for the BBC) & EHRC (n.1). And see further from the TUC: Not part of the job: Young workers’ experience of third party harassment: polling and survey findings (TUC, 2018) & Sexual harassment of LGBT people in the workplace (TUC, 2019). Adams et al. (n.4), Literature Review for the GEO compares these and other studies at ch.2, while Adams et al. (n.4), 2020 Sexual Harassment Survey, also for the GEO, touches on workplace sexual harassment at ch.6. Note, however, the exclusion of sex-based harassment from the definitions of sexual harassment in all these surveys, despite overlaps in principle and practice with what the studies coded as sexual harassment, and that sex-based harassment is squarely within the Equality Act 2010 protections from discriminatory harassment. For discussion of socio-legal, sociological and psychological studies about poor workplace behaviour, including sexual harassment, that were ignored by the Literature Review see Barmes (n.7), 11–25 & 53–54.

¹² 

The most well-known (but not only) indirect legal means is the many legal constraints placed on collectivism at work, particularly in trade unions.

¹³ 

A prominent example is the Public Sector Equality Duty contained in Equality Act 2010, s.149. Further, there is some provision for institutional enforcement of this measure by the EHRC.

¹⁴ 

This is an area in which there have been developments in recent years. In particular, the Immigration Act 2016 created the post of Director of Labour Market Enforcement and changed the name and remit of the Gangmasters and Labour Abuse Authority. Further, Establishing a Single Enforcement Body for Employment Rights (BEIS, June 2021) explains the Government’s plans for a new single enforcement body, notably emphasizing the challenge of getting complainants to speak out at 4, 7, 16 and especially 20, where it was recognized that tackling barriers to workers coming forward needs to be a primary goal in designing the new body and its approach.

¹⁵ 

See Allen (n.7) on the individual dimension of Australian equality protections, and, L. Barmes, ‘Collective Bargaining under the Fair Work Act in UK (and European) Perspective: Ideology, Individualisation and the State’, chapter 11 in S. McCrystal, A. Forsyth & B. Creighton, Collective Bargaining Under the Fair Work Act 2009 (Annandale, NSW: Federation Press, 2018) on recent shifts in Australia in the balance between individual and collective workplace measures.

¹⁶ 

L. Barmes (n.7), 216. See further chapters 8, 9 & 10 on the mediating effects of legal advice and process, including regarding settlement, and on silencing more generally,

¹⁷ 

See further, ibid., 29–37 & 53–54. It is very surprising that Adams et al. (n.4), Literature Review for the GEO, ignored the wealth of socio-legal and sociological evidence (including Government commissioned) about how people respond to problems at work and their experiences when they complain and litigate, despite this being a key topic for the review.

¹⁸ 

See R. Fevre, D. Lewis, A. Robinson & T. Jones, Trouble at Work (London & NY: Bloomsbury, 2012, 136: ‘[R]espondents with problems that involved sexual harassment were less likely to attempt to put the problem in writing, go to a formal meeting or discuss the problem with their employer (55 per versus 74 per cent with other problems).’

¹⁹ 

See n.11 in general, but especially TUC (2016), 18–23, Comres (2017), Table 321, TUC (2019), 26-29 & also Adams et al. (n.4) 2020 Sexual Harassment Survey, 68 & 76–78. See also analogous patterns in: TUC (2018), 14–16 about harassment of young people at work by third parties, but not limited to sexual harassment; EHRC (n.1), surveyed a self-selected group who had been sexually harassed, and employers, and reported that this ‘uncovered the shocking and stark reality of individuals whose careers and mental and physical health have been damaged by corrosive cultures which silence individuals and normalise harassment. We also found a lack of consistent, effective action on the part of too many employers’; & WESC, NDA inquiry (n.1), HC 1720, 19 December 2018, Q34. Considering higher education specifically see S. Ahmed, What is the Use? On the Uses of Use (Croydon: Duke University Press, 2019), 157–162, 181, 183–184 & 189–190 about multiple ways that university students and staff complaints of sexual harassment are stopped, S. Ahmed, Complaint! (Durham, NC: Duke University Press, 2021) more generally, & D. Fernando & A. Prasad, ‘Sex-based Harassment and Organizational Silencing: How Women are Led to Reluctant Acquiescence in Academia’ (2019) 72 Human Relations 1565.

²⁰ 

See J. Aston, D. Hill, and N.D. Tackey, The Experience of Claimants in Race Discrimination Employment Tribunal Claims (DTI, ERRS 55, 2006); A. Denvir, A. Broughton, J. Gifford, and D. Hill, The Experiences of Sexual Orientation and Religion or Belief Employment Tribunal Claimants (ACAS, 2007); Barmes (n.7), 44–47 & 227–236; Busby & McDermont (n.7).

²¹ 

See BEIS, Survey of Employment Tribunal Applications (SETA) (July 2020) for the latest in this series of studies. This concerns only complaints that get to the stage of an ET claim. As such it leaves out all the situations where either nothing is done or the complaint falls always earlier, including by settlement. It remains interesting to see that the reported outcomes were 58% settled (40% via ACAS and 18% privately), 7% claimant successful at hearing, 5% default judgment in favour of claimant, leaving 13% withdrawn, 7% unsuccessful at hearing and 7 % ‘dismissed/other’.

²² 

See further Barmes (n.7), particularly 250–259.

²³ 

Ibid. The reality of contracting for work is equally rooted in the lived reality of employer power as it interacts with law and legal process, today evoked in disputes over both contractual rights and access to statutory ones, since the latter is dependent on the legal categorization of work arrangements.

²⁴ 

ibid., 255 and see further Fevre et al. (n.18), especially 141, 162, 169, 190–171. See also Ahmed (2021 (n.19), 97, and especially Ch.5, but there regarding the failure of bystanders and managers to speak up or take action in response to complaints in universities. She explores here, for example, ideas of collegiality having this effect, 196–201, and further that collegiality can function as a means of protecting ‘some and not others or even some from others’ (p.201), and specifically of protecting white people.

²⁵ 

Barmes (n.7), 231–236.

²⁶ 

Rose & Busby (n.7), 692, and in particular: ‘The shaping of worker subjectivities was also demonstrated by some participants in the way that they appeared to have internalized a deep concern about the legitimacy and appropriateness of challenging employers and pursuing their employment rights. This manifested as a sense of shame or embarrassment for experiencing conflict with one’s employer or asserting one’s rights against an employer.’

²⁷ 

See: Fernando & Prasad (n.19); K. Weiss, ‘Too Ashamed to Report: Deconstructing the Shame of Sexual Victimization’ (2010) 5 Feminist Criminology 286; TUC (n.11) (2016), 9, 11, 21, 23–25; & Essa v Laing Ltd [2004] EWCA Civ 2 for a case law account of the shame evoked by a single racist statement at work: ‘[The claimant] was picked to carry the Welsh flag before boxing for Wales… but became upset whilst carrying it, asking himself what right he had to be representing Wales. He was overcome by similar feeling[s] during the fight.... He says it was "the way he spoke to me, it was the way he treated me. I’ll take it to my grave…" His sense of rejection as a Welshman has so distressed him that he intends to leave Wales to take up professional boxing in England’.

²⁸ 

S. Ahmed, Living a Feminist Life (Durham, NC: Duke University Press, 2017), 141.

²⁹ 

ibid., 6, 34–36, particularly at 35 ‘Connections can be what we have to struggle for, because there is so much silence about sexism; sexism makes it costly for women to speak about sexism. Because, after all, to name something as sexist is not only to name something that happens as part of a wider system…, but it is also to give an account of that something as being wrong and unjustifiable’ and more generally, Ahmed (2021) (n.19).

³⁰ 

For example, S. Fredman, Women and the Law (Oxford: OUP, 1998), 16–17 & N. Lacey, ‘Theory into Practice: Pornography and the Public/Private Dichotomy’ in id., Unspeakable Subjects: Feminist Essays in Legal and Social Theory (Oxford: Hart, 1998).

³¹ 

The absence in Adams et al. (n.4) Literature Review for GEO of any consideration of highly relevant socio-legal evidence and analyses is a very recent case in point, particularly given its concern to illuminate the legal reform debate.

³² 

(n.1).

³³ 

References, therefore, to interviewees are to those who spoke to me. Specific references to the WESC oral evidence are set out in the notes. The WESC references are not necessarily exhaustive of every time a theme arose.

³⁴ 

See on organizations curtailing the use of NDAs in settlement agreements: regarding the Civil Service and ‘their Arms’ Length Bodies’, Cabinet Office Guidance on Settlement Agreements, Special Severance Payments on Termination of Employment and Confidentiality Clauses (Crown Copyright, July 2019) & note immediately below for UCL. See also on experience of the Civil Service approach in practice, WESC (n.1) NDA inquiry, HC1720, 13 February 2019, Q170–172, Q183–Q187, Q208–Q213, 6 March 2019, Q301; on the House of Commons, ibid., Q194–195, Q200–201; on the BBC, ibid., Q249–Q250 (covering also non disparagement clauses), Q257–Q261, Q264–Q265, Q385 (on Deloitte's approach to settlement agreements regarding complaints of bullying and harassment).

³⁵ 

ibid., NDA inquiry, 20 March 2019, Q543. Note, however, the announcement by Dr Chapman’s then employer, UCL, on 17 April 2019 that it ‘no longer uses confidentiality clauses or “NDAs” in settlement agreements with individuals who have complained of sexual misconduct, harassment or bullying as a matter of course’ (and on other measures to address this conduct): https://www.ucl.ac.uk/news/2019/apr/ucl-statement-non-disclosure-agreements. In contrast, see the following evidence about the ubiquity of NDAs in settlement agreements and across a wide range of complaints: S/H inquiry, HC 725, 28 March 2018, Q91, 25 April 2018, 246, 260; NDA inquiry, HC 1720, 19 December 2018: Q4, Q6, Q13; 23 January 2019, Q85, Q93, Q107; 13 March 2019, Q301, Q303–Q304; Adams et al. (n.4), 2020 Sexual Harassment Survey, 85–86 & 95.

³⁶ 

ibid., NDA inquiry, 23 January 2019, Q155.

³⁷ 

See further WESC (n.1) NDA inquiry, HC 1720: 19 December 2018, Q11; 23 January 2019, Q136, (on confidentiality obligations on employers generally being looser because they apply across the organization), Q156 (noting indications post #MeToo of some initial drafts being ‘fairer’ and containing fewer ‘more extreme’ clauses).

³⁸ 

ibid., 19 December 2018, Q6–Q9 & 23 January 2019, Q107.

³⁹ 

See further on the importance of these clauses ibid., Q11.

⁴⁰ 

For example, Practical Law, https://uk.practicallaw.thomsonreuters.com/Browse/Home/About/AboutUsAboutPracticalLaw.

⁴¹ 

I am grateful to Jeremias Prassl and an anonymous commenter for pointing this out to me.

⁴² 

This may be a specific example of one-sidedness in widely circulating standard precedents. Regarding WESC (n.1), see the discussion of exceptions to NDAs, S/H inquiry, HC725, 28 March 2018, Q172–174, including at Q174 that initial drafts do not always include standard exceptions.

⁴³ 

See on the complexity of the whistleblowing exception, WESC (n.1): NDA inquiry, HC 1720, 23 January 2019, Q106 & the Law Society, Practice Note on Non-disclosure agreements and confidentiality clauses in an employment law context (Dec 2019), Section 4, culminating in the statement that: ‘Whistleblowing in the public interest is a complex matter. It’s often legally uncertain whether a person can talk about how they have been treated because they made the disclosure…. Parties who may wish to blow the whistle will often need professional help to navigate this issue.’

⁴⁴ 

See further the somewhat longer, but still quite limited, list of exceptions referred to as ‘common’ in Section 3 of the Law Society Practice Note, ibid. And see, for example, WESC (n.1) NDA inquiry, HC 1720, 19 December 2018: Q16, from Emma Webster, Joint Chief Executive and Senior Solicitor, Your Employment Settlement Service, indicating that an exception is not always included to enable signatories to speak to medical professionals and further, Q17–Q18 on exceptions being very limited.

⁴⁵ 

See for additional evidence to this effect ibid., Q18. However S. Ahmed (2021), n.19, 15 emphasized her interviewees’ need anyway for anonymity ie not only because some were subject to NDAs. See BEIS (n.2) for planned legal reform to specific required exceptions.

⁴⁶ 

See Employment Rights Act 1996, ss.203 (3) and (3A).

⁴⁷ 

See further WESC (n.1): S/H inquiry, HC 725, 25 April 2018, Q270; NDA inquiry, 19 December 2018, Q42, 23 January 2019, Q150–Q151. See BEIS (n.2) for planned legal reform to require independent advice in non-ACAS settlements to cover explanation of the exceptions to an NDA.

⁴⁸ 

Barmes (n.7), 195.

⁴⁹ 

WESC, NDA inquiry (n.1), HC 1720, 20 March 2019, Q538.

⁵⁰ 

See also Ahmed (2021) (n.19), 97–99, about universities, discussing evidence of the silencing that comes both from alleged wrongdoers making use of, or threatening to make use of, confidentiality obligations owed to them, and the spreading out of confidentiality obligations to alleged wrongdoers to constrain others, Ahmed herself included.

⁵¹ 

See on the difficulty of enforcing these clauses, ibid.: 19 December 2018: Q10; 23 January 2019, Q136, but about NDAs themselves, and relating looser obligations being put on employers given that the confidentiality requirement applies to the whole organization and employers have limited control over compliance.

⁵² 

cf. the Financial Conduct Authority ‘regulatory reference’ and other requirements to establish someone as a ‘fit and proper’ person to hold certain posts in that sector: https://www.fca.org.uk/firms/approved-persons/fitness-propriety.

⁵³ 

See WESC (n.1) S/H inquiry, HC 725, 28 March 2018, Q89. Also see further on fear stopping signatories from breaching NDAs, irrespective of whether they are enforceable or would be enforced, NDA inquiry, HC 1720 19 December 2018, Q20; 23 January 2019, Q137.

⁵⁴ 

ibid.: S/H inquiry, Q71; NDA inquiry, 23 January 2019, Q92, but cf. Q87 suggesting some law firms were not that shocked by the clauses in Ms Perkins’ NDA.

⁵⁵ 

ibid., NDA inquiry, 20 March 2019, Q566. See further on concerns about power imbalance and access to legal representation: S/H inquiry, 25 April 2018, Q238, Q241, Q265; NDA inquiry, 19 December 2018, Q11, Q42, 23 January 2019, Q87, Q101, Q147, Q150–151, 6 March 2019, Q306.

⁵⁶ 

See ibid., NDA inquiry, 19 December 2018 discussing NDAs as part of enhanced redundancy agreements: Q3 (where pregnant women are unfairly made redundant), Q7, Q27, Q47; 23 January 2019, Q107.

⁵⁷ 

See also ibid., NDA inquiry, 19 December 2018: Q21 on the pressure even on trade union officials and lawyers to accept NDAs in settlements lest the claimant end up with nothing.

⁵⁸ 

See however, ibid., Q27 on this phenomenon.

⁵⁹ 

See the most recent SETA (n.21) finding that 40% of ET claims were settled via ACAS.

⁶⁰ 

See Rose & Busby (n.7), 697, finding that frequent inclusion of confidentiality clauses in such settlements contributed to dissatisfaction amongst CAB claimants with the limited outcomes available from conciliation, ‘further reducing the ability of workers to demonstrate to themselves and others that their side of the dispute has merit’.

⁶¹ 

The concerns identified here are reinforced by Busby & McDermont (n.7), 182–184 on how claimants experience ACAS conciliation, concluding that: ‘In reality, the claimants in our research often lacked the knowledge and skill required to reach an informed settlement…’

⁶² 

See, however, BEIS (n.2) for planned legal reform to require independent advice in non-ACAS settlements to cover explanation of the exceptions to an NDA.

⁶³ 

S. Zuboff, The Age of Surveillance Capitalism (London: Profile Books, 2019), 7, 48–50 & 235–237, although ironically the aspect of such agreements (evocatively termed ‘click wrap’) on which Zuboff’s analysis centred are those giving corporations legal entitlements to make use of information about the other party, whereas the concern here is the legal imposition of silence.

⁶⁴ 

WESC (n.1) NDA inquiry, HC 1720, 20 March 2019, Q574.

⁶⁵ 

See also ibid., 6 March 2019, Q306 on teachers agreeing to NDAs in order to secure a reference.

⁶⁶ 

See further ibid., 19 December 2018: Q27, Q51.

⁶⁷ 

ibid., 23 January 2019, Q107 (Julie Morris, Partner and Head of Personal Legal Services, Slater and Gordon, although seeing this less often than 10 years ago because of a culture change), but cf Q93 (Jane Mann, Partner and Head of Employment Group, Fox Williams LLP, that in her experience this was always a relatively small category of confidentiality agreements) & again at Q146 mentioning potential problems with employers not imposing NDAs on complainants, for example where an investigation has exonerated the person complained against (in answer to a question about UCL’s decision not to use NDAs in sexual harassment etc settlements as a matter of course, further details at n.35). It is worth remembering, however, all the other legal and social pressures to be silent, not least the law of defamation.

⁶⁸ 

See further on the experience by university students of sexual harassment, and complaints about this, including this being kept out of the public eye: Ahmed (2017) (n.28), 139–142, observing at 139 ‘Complaints about sexual harassment are not made public as a way of protecting the organization from damage. Even if the complaints are successful, even if a contract is terminated (which is rare) or someone leaves rather than face a tribunal, it can be as if what happened never happened. No one is allowed to speak of it; no one speaks of it’; and further, including noting the use of NDAs, Ahmed (2021) (n.19).

⁶⁹ 

See WESC (n.1) NDA inquiry HC 1720, 20 March 2019, Q545. Also for investigative journalism about university use of NDAs, R. Croxford, ‘UK Universities face ‘gagging orders’ criticism’, 17 April 2019, BBC News https://www.bbc.co.uk/news/education-47936662 & ‘Sexual assault claims ‘gagged’ by UK universities’, 10 February 2020, BBC The Next Episode https://www.bbc.co.uk/news/uk-51447615.

⁷⁰ 

ibid., 19 December 2018, Q34, Q36.

⁷¹ 

ibid., 23 January 2019, Q143, Q156–Q158 (the latter also on change amongst lawyers and their regulator).

⁷² 

See now Equality Act 2010, s.109.

⁷³ 

See further WESC (n.1), S/H inquiry HC725, 28 March 2018, Q183 & 6 June 2018, Q481.

⁷⁴ 

It is interesting to compare the position in the United States, where Edelman and her collaborators, have shown that, first, the US Courts (not legislature) crafted a similar employers’ defence, secondly, that this followed employers’ widespread adoption of measures on the advice of HR practitioners to avoid legal liability and, thirdly, that claims are increasingly, denied on this basis, and despite these HR practices being ineffective to address workplace harassment. Edelman depicts these ‘symbolic structures’ as critical to the inadequacy of US law in this area. See L.B. Edelman, ‘How HR and Judges Made It Almost Impossible for Victims of Sexual Harassment to Win in Court’, Harvard Business Review, 22 August 2018, L.B. Edelman, ‘The #MeToo Movement, Symbolic Structures and the Limits of Law’ in A. Noel & D. Oppenheimer (eds), The Global #MeToo Movement (2020, Full Court Press) & L.B. Edelman & J. Cabrera, ‘Sex Based Harassment and Symbolic Compliance’ (2020) 15 Annual Review of Law and Social Science 361.

⁷⁵ 

See further WESC (n.1) NDA inquiry, HC 1720: 19 December 2018, Q5 & 23 January 2019, Q107.

⁷⁶ 

ibid., 20 March 2019, Q554.

⁷⁷ 

ibid., WESC, 19 December 2018: Q22, Q25, Q45.

⁷⁸ 

ibid., 23 January 2019, Q135. See also n.35 on UCL’s change of policy regarding the use of NDAs.

⁷⁹ 

See also Ahmed (2021), 99–100 on various in-kind benefits offered to staff and students to stop complaints, and the significance of this practice for institutional culture.

⁸⁰ 

ibid., 19 December 2018, HC1720, Q45 and Q46 specifically about the experiences of pregnant women, emphasizing that they have no choice but to sign an NDA & 23 January 2019, Q107, Q135, Q140.

⁸¹ 

See ibid., 19 December 2018, Q5, the stark suggestion also from Joeli Brearley, Founder, Pregnant Then Screwed, that those who sign NDAs are the lucky ones with others left with only humiliation: ‘Of the 390,000 women who encounter some form of pregnancy or maternity discrimination, I would guess that less than 1% sign a non-disclosure agreement. They are the lucky ones; they are the women who are getting some form of compensation. The majority of women do not have the privilege to do anything about this; they just walk away with humiliation and no compensation whatsoever.’

⁸² 

See Z. Adams, A. Adams-Prassl & J. Adams-Prassl, ‘Online Tribunal Judgments and the Limits of Open Justice’ 2021 Legal Studies 1, particularly on harms to working people of ET judgments now being easily accessible, including from employer use of AI in the context of both recruitment and disputes. But cf. K. Wiggins & J. Browning, ‘How Corporate Britain Hides Thousands of Discrimination Cases’ Bloomberg, 6 September 2019 mining the ET judgments database to expose that large numbers of withdrawn sex discrimination claims denote widespread settlement of such claims with NDAs, many by household name employers.

⁸³ 

See n.20 and the text to this.

⁸⁴ 

Or as another interviewee put it, something that a complainant ‘may perceive as justice’. Equally, the point was made that taking away some claimants’ option of signing an NDA would merely reduce their settlements because no one would be interested in what happened to them. This is all too plausible given what is known about the fate of reports of sexual harassment etc.

⁸⁵ 

See further for a range of views on this topic, including regarding settlements not declining where settlements do not include NDAs, WESC (n.1): S/H inquiry, HC725, 28 March 2018, Q163–Q164; NDA inquiry HC 1720, 19 December 2018, Q38–Q40, Q53, 23 January 2019, Q87, Q142–Q143, 13 February 2019, Q256, 6 March 2019, Q321–Q322, Q331–Q332, Q390.

⁸⁶ 

ibid., NDA inquiry, 23 January 2019, Q136–Q137 on the use already of libel and defamation laws to silence complainants & 6 March 2019, Q334 on lawyers finding other means in settlements to ensure silence, for example with ‘beefed up’ non-derogation clauses.

⁸⁷ 

ibid., NDA inquiry, 19 December 2018 Q45, 23 January 2019, Q87 & see Wiggins & Browning (n.82).

⁸⁸ 

Compare also Ahmed (2019) (n.19), 162–163, 181–182 & 189–190, and more generally Ahmed (2021) (n.19), on long term effects on individuals and organizations of both experiences of sexual harassment and how complaints were treated in universities,

⁸⁹ 

WESC (n.1), NDA inquiry, 20 March 2019, Q572.

⁹⁰ 

Ahmed (2017) (n.28), 90 & 103–105.

⁹¹ 

See further on this interaction WESC (n.1), NDA inquiry, HC1720, 19 December 2018, Q27.

⁹² 

See ibid., Q41 for this bleak assessment from Emma Webster, Joint Chief Executive and Senior Solicitor, Your Employment Settlement Service: ‘There is no shame attached to pregnancy and maternity discrimination; it is an acceptable form of discrimination as far as most businesses are concerned. Going on to Twitter and talking about it will not harm their reputation that much, because there is… an awful lot of feeling amongst an awful lot of people that, if you take a year off, you are fair game….’

⁹³ 

See ns.27 & 28 above and the text to them.

⁹⁴ 

For some examples, see F. Fanon, ‘The Fact of Blackness’, in The Post-Colonial Studies Reader, B. Ashcroft, G. Griffiths and H. Tiffin (eds) (1952; repr., London: Routledge, 2003), 323–26; B Skeggs, Formations of Class and Gender, Becoming Respectable (London, California & New Delhi: Sage Publishing, 1998); & L Hayes, Stories of Care, A Labour of Law, Gender and Class at Work (London: Palgrave MacMillan, 2017).

⁹⁵ 

See WESC (n.1), NDA inquiry, HC1720, 23 January 2019, Q87 in which Baroness Helena Kennedy QC characterized NDAs as part of the bigger pattern of women being silenced about the ways in which they are abused.

⁹⁶ 

See WESC (n.1), S/H inquiry, HC 725, Q171 where Max Winthrop, Chair of the Employment Law Committee of the Law Society, perceiving the origins of NDAs in commercial legal practice, commented: ‘I am not at all happy with this creep of the idea of NDAs…. I do not see how something as broad as that is appropriate in an employer-employee relationship’; & NDA inquiry, 23 January 2019, Q85, where Kiran Daurka, Discrimination Law Association and Partner, Leigh Day said: ‘What has happened over time is that [NDAs] have become more onerous. It may have started with confidentiality around the terms of the settlement. It has now expanded to a fact of the settlement agreement. Over the last maybe six or seven years, it is almost standard to see in all circumstances surrounding the termination of your employment or leading up to the settlement and the impact…’; Q87; & 6 March 2019, Q303–Q304.

⁹⁷ 

See on the US legislative response: J. Ence, ‘I Like You When You are Silent: The Future of NDAs and Mandatory Arbitration in the Era of #MeToo’ (2019) Journal of Dispute Resolution 165; M.A. Weston, ‘Buying Secrecy: Non-Disclosure Agreements, Arbitration, and Professional Ethics in the #MeToo Era’, (2021) University of Illinois Law Review 507, 533–536. Legislative developments and debate are ongoing in the United States, including at the federal level, See, for example, the campaigns by People Parity Project regarding the use of ‘coercive contracts’, which encompass, amongst other things, forced arbitration clauses (as well as NDAs in settlements and employment contracts), https://www.peoplesparity.org/coercivecontracts/, and about the use by the legal profession specifically of such contractual devices, https://www.peoplesparity.org/firms/

⁹⁸ 

WESC, The use of non-disclosure agreements in discrimination cases (House of Commons, June 2019), p.17 para [14].

⁹⁹ 

See in this regard the database of anonymous testimonies of the Can’t Buy My Silence Campaign (n.4) for such accounts.

¹⁰⁰ 

See for a fascinating meditation on individuals’ needs for NDAs, and the complicity of organizations, people within them and their lawyers in constructing that reality, WESC (n.1) NDA inquiry, HC 1720 19 December 2018, Q22, plus 23 January 2019, Q107 (that if complaints became more acceptable, complainants would have less need for NDAs to protect their employment prospects), Q160 (that change can be driven by employers) & 6 March 2019, Q301 (that complainants asking for NDAs comes from ‘a culture in which there is no discussion of discrimination or harassment that has taken place, and the blanket use of confidentiality agreements just adds to that culture’).

¹⁰¹ 

ibid., 6 March 2019, Q715, Q726 & Q728 about SRA and EHRC witnesses perceiving change in lawyer behaviour etc regarding NDAs.

¹⁰² 

ibid., Q726 by an EHRC witness on the need to focus on ‘standard practice’, rather than the use of NDAs in ‘egregious cases’, and suggesting an appetite for change in the direction of this suggestion.

¹⁰³ 

ibid., 13 February 2019, Q256 on organizations who do not settle with NDAs finding complainants still do not speak out for other reasons.

¹⁰⁴ 

ibid., 13 February 2019: Q238 on the process of change in the Civil Service; Q251–252, Q285–Q286 on Brighton & Hove Council’s developing approach.

¹⁰⁵ 

Ahmed (2021) (n.19), 40.

¹⁰⁶ 

See Weiss (n.27), 305: ‘[P]art of an agenda to reduce the shame of sexual victimization must start with… having frank public conversations about sexual violations and sexuality more generally…. Policies regarding rape victimization need to help victims deal with the sexual component of their crimes without perpetuating the idea that such situations are private matters or too embarrassing to talk about.’

¹⁰⁷ 

See, for example, WESC (n.1), NDA inquiry, HC 1720: 19 December 2018, Q28; 23 January 2019, Q143, Q158.

¹⁰⁸ 

ibid., Q146, Q161, Q166. Another set of technical challenges to address would be those regarding data processing, although it is surely not beyond the wit of lawyers to work out how this could be done.

¹⁰⁹ 

See further https://www.gov.uk/government/news/universities-pledge-to-end-use-of-non-disclosure-agreements and https://cantbuymysilence.com/universities-pledge/, including for the impressive list of universities which have already signed up.

¹¹⁰ 

See Adams et al. (n.82) arguing for anonymization of the public database of ET judgments, given the actual and potential harms to claimants of information about their claims being in the public domain. This conclusion further testifies to the urgent need for a change of organizational and broader culture regarding allegations of workplace misconduct.