All your resources at your fingertips.Learn More
Debbie Grennan, Guildhall Chambers
As long ago as 2011, the Government announced its intention to repeal the specific third party harassment provisions set out in s 40(2)-(4) Equality Act 2010 (EA). Following a consultation process in which the large majority of those who responded opposed the proposed repeal, the Government has nonetheless confirmed that it will go ahead. This article evaluates the stated rationale for the repeal and considers the resulting problems which both employees and employers will face going forward.
The issue of the liability of employers for acts of harassment by individuals other than their employees or agents came to the fore in the infamous case of Burton v De Vere Hotels Ltd  ICR 1. The Employment Appeal Tribunal (EAT) held that an employer could be liable under the Race Relations Act 1976 for exposing members of staff to the racist jokes of Bernard Manning.
In the case of Pearce v Governing Body of Mayfield Secondary School  ICR 937, the House of Lords disapproved Burton, holding that an employer's failure to protect employees from third party harassment did not amount to ‘less favourable treatment on racial grounds' unless the employer's failure to protect was itself discriminatory. Thus, any failure to safeguard the employee must have been done by the employer on the ground of the protected characteristic.
This left an obvious and important gap in the protection offered to employees who had suffered harassment at the hands of third parties.
The issue of compliance of the UK law with EU law (in the context of sex discrimination) was considered in judicial review proceedings in Equal Opportunities Commission v Secretary of State for Trade and Industry  ICR 1234.
The EOC contended that the UK law was defective in two respects which are material to the issue of harassment. First, it was argued that the UK requirement that harassment must be done ‘on the ground of' sex was incompatible with Directive 76/207, which covered harassment which was ‘related to' sex. It was contended that the UK wording imported a causative test which was not permitted by EU law.
Second, it was contended that, in the light of the HL decision in Pearce, the UK law was defective in that it failed to provide for liability for third party harassment, which was required by Directive 2002/73.
Burton J upheld the first contention, holding that the ‘on the ground of' test incorporated a causative element which was not permitted by EU law. The EU test of ‘related to' required only that the harassment be in ‘connection' or ‘association' with sex. It was not necessary that sex be the cause of the harassment.
However, Burton J rejected the second contention, on the basis that it had not been suggested that there was anything in the Directive which required member states to impose vicarious liability on an employer in respect of third party harassment.
It was held that theoretically, an employer could be held liable for failing to take action where there was a continuing course of offensive conduct by a third party, such as a supplier or customer, which the employer was aware of but failed to address. This was on the basis that the employer's own failure could contribute to the creation of a hostile work environment and thus constitute harassment by the employer. However, such cases would be rare, at least while the ‘on the ground of' formulation was in place. Burton J clearly stated that if the ‘related to' formulation was adopted, this would resolve the problem as liability for third party harassment may be achieved through a general harassment provision which was drafted in these terms.
This case must be treated with a degree of caution in relation to the question of compliance of UK law with the Directive in respect of third party harassment. Prior to the hearing of the case, the Government had published a Factsheet which stated that ‘on appropriate facts, the harassment provisions in the [SDA]...... might be interpreted so that where an employer knowingly fails to protect an employee from, for example, repetitive harassment by a customer or supplier, the employer is "subjecting the employee to harassment."' As a result of this, there was not as full an argument before the court as might otherwise have been the case on the issue of UK compliance with EU law.
The judge did find that the existing UK did not reflect the UK Government's interpretation of the legislation as set out in the factsheet. The Government then amended the Sex Discrimination Act 1975 to specifically provide for a claim in respect of third party harassment, but only where the employee could show that the employer had been made aware that there had been such harassment on at least two previous occasions.
None of the other anti-discrimination legislation was amended in the same way. Other strands of discrimination remained subject to position as set out in Pearce. Thus, in Conteh v Parking Partners Ltd  ICR 341, the EAT found that the claim of third party harassment by a Claimant who had been subjected to racist abuse must fail, because the RRA required that the employer's failure to act must itself have been a racially discriminatory act.
The decision not to amend the Race Relations Act 1976 in the same way, despite the existence of the Race Directive 2000/43/EC (which proscribed racial harassment in substantially the same terms as the Equal Treatment Directive), was described as ‘odd' by Underhill J in Sheffield City Council v Norouzi  IRLR 897. In that case (brought under the 1976 Act), it was held that the Council was liable for third party harassment by a child in its care who had racially harassed one of its employees. It had been argued that liability arose in this case under the Race Directive, which could be directly enforced against a public authority. However, this is another case which must be treated with caution. It was decided on the basis of concessions made by the Respondent at which Underhill J expressed some surprise: ‘[we] were initially surprised as to whether the basis of the claim in law was really as clear cut as the council's concession would suggest but ... we see how the case can be put.'
What is abundantly clear is that prior to the implementation of the EA, there was real uncertainty as to whether employees were protected from third party harassment where the characteristic relied on was anything other than sex.
It is worth remembering that one of the stated purposes of the EA was to introduce, for the first time, a common statutory definition of harassment.
For the first time, the Act provided specific protection against third party harassment which applied to all of the characteristics covered by the general protection from harassment, but only where the employer was aware of two previous occasions of harassment by a third party. It effectively adopted the approach taken in the amended Sex Discrimination Act 1975 by providing for a ‘three strikes' formula. It provides as follows:
(1) An employer (A) must not, in relation to employment by A, harass a person (B) -
(a) who is an employee of A's;
(b) who has applied to A for employment.
(2) The circumstances in which A is to be treated as harassing B under subsection (1) include those where -
(a) a third party harasses B in the course of B's employment; and
(b) A failed to take such steps as would have been reasonably practicable to prevent the third party from doing so.
(3) Subsection (2) does not apply unless A knows that B has been harassed in the course of B's employment on at least two other occasions by a third party; and it does not matter whether the third party is the same or a different person on each occasion.
(4) A third party is a person other than -
(a) A; or
(b) an employee of A's.
Thus, s 40(2)-(4) only provides specific protection from third party harassment in the event that there have been two previous instances of harassment, which is a rather artificial way of providing for liability. However, it is very much arguable that this is better than nothing.
There can be little doubt that uncertainties remain as to the remit of s 40(2)-(4). For example, the test requires employers to ‘know' that an employee has been harassed on at least two previous occasions before they are at risk of being liable for further harassment. It could be argued with some force that an employer did not ‘know' about the earlier acts of harassment until it had carried out an investigation into those matters and established that the harassment alleged had in fact occurred. However, attempting to verify an employee's complaint can often be difficult, especially where it is a case of one word against another. Is it necessary that the employee is able to positively prove two earlier occasions of harassment before the Claimant can be found to have satisfied the three strikes test, or will simply raising a complaint be enough? While the wording of the Act suggests the former interpretation, we await clarity on the point because to date, there have been no appellate decisions on the issue of knowledge. Thus, it may be risky for employers to simply ignore complaints even where it has good grounds to believe they are ill-founded.
A further problem with the provision as currently drafted is that for the duty to take preventative steps to arise, it is necessary that the employer knows that the employee has already been harassed on two previous occasions. It is not necessary that the third party harasser is the same person, nor is there any time limit on when the harassment took place. This could lead to surprising results. For example, an employer would be required to act in the case of an employee who had been harassed on more than one occasion some years ago but not in the case of an employee who was subjected to one incident of serious third party harassment in the last few days and who will facing the same third party during their next working shift. It is arguably unsatisfactory that the employer's obligation to act is triggered simply by the employee's particular history of being harassed rather than the potential risks of third party harassment going forward, particularly in the case of a known harasser.
It is clear the third party harassment provisions could be improved upon. Nonetheless, there can be little doubt that the introduction of these measures has led to employers taking harassment of their staff by third parties much more seriously. One needs only to travel on public transport or visit retail or medical environments to appreciate that many employers have been taking steps to ensure that their employees are not being subjected to harassment by those they come into contact with whilst at work.
In its ‘Plan for Growth' issued in March 2011, the Government announced its intention to consult as to the removal of the ‘unworkable requirement in the EA for businesses to take reasonable steps to prevent persistent harassment of their staff by third parties as they have no direct control over it......'
The reasons given for the proposed appeal were essentially that:
71% of those who responded to the consultation opposed the suggested repeal, with 20% supporting it.
Those who were vociferous in their opposition included the Equality and Human Rights Commission, the Law Society and the Employment Law Bar Association.
Interestingly, even some of those in favour of repeal, such as the EEF, did not ‘regard the measure as a significant step in reducing the regulatory burden on employers.'
It is clear from the responses that incidences of third party harassment are more prevalent than had been indicated by the government. Those who responded gave examples of third party harassment cases they had dealt with or were aware of. The EHRC response indicated that between December 2007 and June 2012, their helpline had received 192 calls relating to harassment, 25 (13%) of which related to third party harassment. The ELA response referred to their members acting for both employers and employees in third party harassment claims. It appears that the majority of such claims had settled without the need for a final determination by a tribunal, quite possibly because the conditions set out in s 40 were sufficiently clear for the parties to take an early view of the merits of their respective cases.
Interestingly, the response provided by the President and Regional Employment Judges of the Employment Tribunals noted that if s 40(2)-(4) was repealed, there was potential for claims to be brought under EU Directives against public sector employees and noted that ‘it is still open to argument that even if the provisions here under consideration are repealed, the requirements of EU law mandate a form of protection for third party harassment, possibly going beyond the limits imposed by section 40.'
It is therefore clear that removing this specific provision does not mean that cases of third party harassment will simply go away.
Despite the extent of the opposition to the proposal, the Government has confirmed that it intends to repeal s 40(2)-(4) on the basis that it is ‘unworkable.' This is now provided for in the Enterprise and Regulatory Reform Bill.
The key issue for consideration is whether the Government's stated reasons for the repeal are well-founded and further, whether such reasons are outweighed by the difficulties that are likely to result going forward.
The first stated reason for repeal is that it is unfair to impose liability on employers where they have no direct control over the actions of third parties. This rationale appears to be based on a fundamental misunderstanding of the provision. There is no automatic liability for third or subsequent acts of harassment by third parties. Liability only arises where the employer knows of two previous incidents of harassment by a third party and then fails to take such steps as are reasonably practicable to prevent further harassment. The requirement of reasonable practicability will take into account all the circumstances of the case, including the extent to which the employer is able to exercise a degree of control over the actions of the harasser. It is the actions of the employer in responding to the harassment which will be in issue. It is therefore difficult to see how it would be unfair to impose liability on an employer who fails to act reasonably.
The second stated reason is that it imposes an unnecessary burden on businesses. As noted below, it is here where the Government's rationale for repeal can be criticised as being inherently contradictory.
The third stated reason is that there are alternative remedies available to victims of third party harassment. In order to evaluate this rationale, it is necessary to consider the alternative claims which might be brought.
A claim for breach of contract is one which can only be pursued in a tribunal if the employee's employment has terminated. If the employee resigns, they could claim constructive dismissal, ie that the employer's failure to take reasonable steps to prevent further harassment amounts to a breach of the implied term of trust and confidence and/or to provide a safe and suitable working environment. However, this hardly amounts to providing them with protection from third party harassment. Requiring someone to leave their employment in order to avoid such harassment is surprising, particularly when many employees wish to retain their jobs and simply want the harassment to stop. Further, compensation for discrimination is uncapped, whereas unfair dismissal compensation is already limited and likely to be further restricted in the near future.
Further, this potential remedy is of no assistance at all to an individual who is not an employee for the purposes of unfair dismissal, but who meets the definition of employee or worker for the purposes of the EA. In the current climate of a-typical working arrangements, this is another notable lacuna.
As a matter of principle, it is clear that the PHA does not provide any specific protection for harassment which is related to a protected characteristic, which can and arguably should have a particular stigma attached to it. In addition, the PHA carries a high threshold as to the type of conduct it covers, with the standard being significantly higher than what is required for harassment to arise under the EA. It also requires the Claimant to prove a ‘course of conduct' on the part of the particular Defendant and may therefore be of little assistance where the employee, perhaps in a public-facing role, suffers repeated harassment not from the same person, but from different individuals. In addition, the PHA provides a remedy only against the third party harasser, who may be without the means to meet any award. Further, in the absence of any remedy against an employer who chooses to ignore the harassment of its employees by a third party, it does nothing to contribute to the aim of seeking to eliminate discrimination in the workplace.
Such claims would have to be brought in the civil courts, leading to lengthier and in all probability more expensive litigation, with employees also having to move into a ‘costs normally follow the event' regime. Compensation may also be inadequate. There is no provision for an award under the PHA to compensate for injury to feelings. Thus, an employee will have to prove psychological or personal injury in order to recover general damages and associated special damages.
Once again, as the Consultation Paper indicates, such a remedy will only be available where the employee sustains psychological or personal injury. It would not enable a claim to be brought where an employee had suffered harassment which, though serious, had not caused such an injury.
As the Consultation papers acknowledges, this applies to the risks to the health and safety of employees, which ‘could include third party harassment.' However, it is far from clear in what circumstances the employer's obligations in respect of health and safety would cover third party harassment, so the potential existence of such a claim is entirely speculative. Further, this would present the same problems associated with requiring Claimants to pursue remedies in the civil courts.
The most likely outcome of repeal is that attempts will be made to mould the remaining provisions of the EA to fit.
It is almost inevitable that Claimants will seek to reply on the general definition of harassment, set out in s 26 EA. It is not entirely clear whether this will be successful but the point is very much arguable.
It is noteworthy that the Government's Consultation paper itself only went so far as to state that ‘it is possible that [the general definition of harassment] covers acts of conduct covered by [Section 40(2)-(4)]....'
The definition of harassment under EA adopts the ‘related to' formula rather than the ‘on the ground of' test upon which cases such as Conteh were decided. According to Burton J in the 2007 judicial review proceedings, this new wording may well be sufficiently broad to encompass a failure to protect against third party harassment. However, the position is far from clear. It is therefore likely that the law will be put back into the state of flux which existed following the decisions in cases such as Conteh and Nourozi.
Indeed, there is a risk that the law will return to the position as it was following the Mayfield case. While the third party aspect of Mayfield is obiter, it is nonetheless a view from the House of Lords. The essential question will be, if an employer puts, for example, a black employee in a position where they have been subjected to incidents of racial abuse by third parties and fails to take adequate steps to seek to prevent further racial abuse, can the inference be drawn that this failure was related to their race? There are arguments either way. It is strongly arguable that this lack of certainty is simply unacceptable.
Further, in respect of those claims which can potentially be pursued under EU law, whenever such cases are brought, the tribunals are likely to be faced with much more complex arguments on EU law. This will go hand in hand with the attempts to fit the facts into other forms of protection from discrimination. In contrast, the current provision provides at least some clarity for employers and employees.
This uncertainty will have practical considerations for employers in particular. While it has to be accepted that there are aspects of s 40(2)-(4) which are unsatisfactory, they do have the virtue of spelling out clearly when an employer's duty to take reasonable steps arises. As noted above, if the repeal goes ahead, Claimants will inevitably mould their claims in more artificial ways in order to seek an effective remedy in the employment tribunals, by relying on s 26 and/or EU rights. The only other alternative will be to move into the civil court arena. In either event, costs, complexity and delays are bound to increase. It is also arguable that lack of certainty may hinder the parties from reaching settlement, something which the relatively clear current provisions appear to encourage.
Many of the responses to the Consultation referred to the deterrent effect of s 40(2)-(4), noting that employers are more likely to take their responsibilities towards their employees seriously if there is a possibility that they may be taken to an employment tribunal under a specific legal provision. The Government rejected this contention, stating that it did not accept the implication that employers can only be encouraged to do the right thing through the threat of legislation. However, time and time again, it has been apparent that unless and until there is clear legislation in place to deal with a particular problem, the problem continues, sometimes for some considerable time after the relevant legislation is in place.
Further, the Government's response does not appear to have dealt with the argument that the very existence of this provision may have the practical consequence of raising employers' awareness of third party harassment as a specific issue which may affect customer facing employees in particular, which the business needs to address. It is noteworthy that the introduction of this provision was met with a huge amount of publicity, with employers introducing specific training to deal with this.
The Government's response thus fails to acknowledge in any way the positive impact caused by s 40(2)-(4) having created a climate whereby employers take greater responsibility for ensuring staff are not subjected to third party harassment. The relatively small number of these claims may be explained by the fact that its very existence has led to many employers implementing clear procedures for responding to such forms of harassment. The danger is that if the deterrent effect of this specific provision is lost, there may be a return to the old ways.
As noted above, removal of the protection from third party harassment from the EA may well be contrary to EU law, at least in relation to the protected characteristics of sex and race. If EU law is found to offer protection against third party harassment related to sex or race, would this not of itself lead to a deficiency in the structure and scope of the anti-discrimination legislation? We could be left with a situation where employees who are harassed for a reason related to these characteristics have protection which is not available to an employees who are harassed due, for example, to their disability. It is difficult to see any justification for such differing levels of protection. Further, why should an employee's right to protected from third party harassment depend on whether they are employed by a public or private sector employer and can therefore directly enforce the applicable EU directives? It is strongly arguable that protection from third party harassment has never been more important, given the increasing prevalence of secondments, for example. Should it not be a matter of policy that the law provides protection in such circumstances, as well as public/customer-facing environments?
An important focus of the Government's impact assessment has been on the direct costs of the current provision for employers. The Government suggests that the repeal will save businesses £0.3m per year. However, the same impact assessment also estimates the transitional familiarisation costs to employers to be between £1.9 - 4.3m, with its best estimate being £2.78m. Further, it is estimated that repeal would lead to more instances of workplace discrimination, which would involve additional costs to employers of £0 - 0.87m per annum. There is then the cost occasioned by individuals seeking alternative and more expensive forms of redress. It is therefore clear that the saving of cost alone cannot possibly justify this repeal.
Further, what is not at all clear is the extent to which any cost/benefit analysis has been addressed from the perspective of whether the provision is worth keeping precisely because relatively few claims are being brought (and most of which seem to be settling prior to hearing). This may well indicate that the current provision is working and preventing incidents of third party harassment from occurring.
In addition, the obvious question is how one factors into the ‘costs to business' arguments the practical effects of failing to deal directly with third party harassment? What about cases of job burnout and employees leaving their employment or that sector entirely? What about loss of productivity through increased staff absence and stress-related ill-health? In reality, these are likely to dwarf the cost-saving sums put forward by the Government.
One might well take the view that the government's arguments are internally contradictory. How can the Government argue with any force that the legislation is both useless and unused, yet at the same time represent such a burden or hazard to business that repeal is justified? The inevitable result of repeal will inevitably be a weakening of the protection available to Claimants, uncertainty for both businesses and employees and increased costs overall, the very things our employment law should be seeking to avoid.
Authoritative analysis of the rules governing termination of employment provides coverage of the...