This issue of the UK Employment Flash looks at the latest employment law developments and insights about employment law in the U.K., including an exploration of the potential relationship between artificial intelligence and recruitment discrimination, a comparison of “quiet firing” and constructive dismissal, a reminder about legal privilege in communications and an update on incentives and remuneration.
- Investigating the Link Between AI and Recruitment Discrimination
- ‘Quiet Firing’ — A Constructive Dismissal Claim on Employers’ Hands?
- A Timely Reminder About Legal Privilege in Communications
- UK Government Publishes New Guidance on Employment Status
- Tailoring Workplace Policies To Accommodate Menopause
- UK Incentives and Remuneration Update
Investigating the Link Between AI and Recruitment Discrimination
The U.K. privacy authority promises to investigate how the use of AI in recruitment might produce discriminatory results, particularly against neurodiverse people and ethnic minorities.
On July 14, 2022, the U.K.’s privacy authority, the Information Commissioner’s Office (ICO), announced it would study the potential effect of the use of artificial intelligence in recruitment. In particular, the ICO is “investigating concerns over the use of algorithms to sift recruitment applications, which could be negatively impacting employment opportunities of those from diverse backgrounds.”
There are nine protected characteristics under English law with respect to which no individual or group of people should be treated less favorably in relation to their employment (unless an exception or objective justification applies). In its investigation, the ICO will focus on two groups, which fall under the protected characteristics of race (ethnic minorities) and disability (neurodiverse people who meet the statutory definition of “disability”).
While at first AI (and its intended objectivity) can be a helpful tool to make recruitment processes more efficient and fair, from an English employment law perspective, its use includes two main discrimination risks:
- direct discrimination, which generally occurs where, because of a protected characteristic, A treats B less favorably than A treats or would treat others; and
- indirect discrimination, which occurs where A applies a “provision, criterion or practice” (a PCP) to B (who has a protected characteristic) and other people who do not share B’s protected characteristic. However, that PCP puts or would put people with whom B shares the protected characteristic at a disadvantage compared to others, and A cannot show that the PCP is a proportionate means of achieving a legitimate aim.
AI, when used in recruitment processes, might create discrimination risks vis-a-vis ethnic minorities and neurodiverse people in several ways:
- The data set used to build the AI algorithm may be inherently biased, leading to biased results. For example, if the data set reflects historically successful candidates and such candidates have tended to be from a certain ethnicity, this may lead to discriminatory results against individuals who have characteristics (such as names or background) typically associated with ethnic minorities.
- The way in which the AI algorithm has been programmed to select suitable candidates may lead to biased results. For example, analysis has shown that postcodes may be a proxy for ethnicity in cases where certain ethnic groups are predominantly based in certain geographic areas. So if, for instance, an algorithm has been programmed to exclude candidates beyond a certain geographic area, this may lead to discriminatory results.
- Where AI recruitment tools feature face and voice analysis, AI algorithms may not account for the fact that neurodivergent people and ethnic minorities may socialize and communicate differently. For example, some neurodivergent people may have speech difficulties or may communicate differently from neurotypical people (for example, they may not be able to replicate a tone of voice or construct sentences in a particular way). Ethnic minorities may speak with different accents or, if English is not their first language, their spoken or written English may not follow the speaking or writing style of a native English-speaker. If AI algorithms are not trained to analyze diverse forms of speech, writing and facial expressions, the algorithms may produce discriminatory results.
Where a claimant is successful in bringing a discrimination claim against an employer, the claimant may be awarded compensation and, although claimants are required to mitigate their losses, there is no cap on the potential award. Therefore, the financial (and possibly reputational) risks associated with discrimination can be substantial. To reduce the risk of discrimination and those associated risks when using AI, employers may consider a number of steps:
- AI system testing — Before implementation, ensure that the parameters against which the AI algorithm is selecting candidates and the data set upon which the algorithm has been built have been rigorously scrutinized and tested with an aim to eliminate any potential bias.
- Bias Training — Provide bias/discrimination training for employees and developers who create and use AI recruitment tools. This will allow them to understand and address the discrimination risks associated with the algorithm’s programming, the data that is used to train the algorithm, the results produced by the algorithm and any other unexpected biases.
- Algorithm Audits — Implement regular “algorithm audits,” to evaluate the AI program’s fulfillment of its purpose, efficiency and fairness. In addition, maintain some human intervention in the AI process to monitor the output of such AI, update and refine the algorithm where needed and (if possible) give final approval on recruitment decisions made by the AI. Human involvement is particularly important to ensure that, where needed, reasonable adjustments are made for any disabilities.
- Monitoring of Developments — The legal and regulatory space governing AI continues to evolve, just as the evolution of the technology progresses rapidly. Given the impact (both positive and negative) that AI is having on businesses, employers should track any developments relating to AI generally and specifically in relation to the workplace. Employers can also seek expert advice to help them navigate risks associated with their use of AI.
‘Quiet Firing’ — A Constructive Dismissal Claim on Employers’ Hands?
The discussion on “quiet quitting” and “quiet firing” continues to unfold as employees return to the workplace after the pandemic.
As employees grapple with resetting work and social boundaries post-pandemic, a trend of “quiet quitting” has emerged. The term describes how employees may restrict their efforts at work to address only tasks that they are paid to do. On the flip side of this trend, some employers are engaging in the practice of “quiet firing” — a phrase that describes deliberate efforts by managers to encourage employees to quit. Quiet firing can take many forms, including failing to reward employees for their contributions, neglecting their requests or creating a more hostile work environment so that dissatisfied employees will choose to leave.
Although quiet firing is a newly coined term, the concept of “managing out” employees is not new. The reasons for quiet firing vary, including the implementation of indirect layoffs and employers seeking to avoid liability for dismissing employees. But such employers should be wary of potential pitfalls if they quietly manage out employees. Some employees may feel that their managers’ actions have created an environment where they have no choice but to resign. In such circumstances, there is a risk that employees may bring a constructive dismissal claim against their employers.
Employees claiming constructive dismissal face a difficult burden of proof. To prove a claim, an employee must show that the employer has committed a repudiatory breach of his or her employment contract – i.e., a breach sufficiently serious to warrant the employee resigning. In the context of quiet firing, a singular breach by an employer is less likely to cause a constructive dismissal. Rather, given the passive nature of quiet firing, employees might instead flag a series of breaches that cumulatively add up to breach of employers’ implied duty of trust and confidence. Employees might argue that the final incident in the series of breaches was the “last straw” (even if that final incident is itself insubstantial), which together with previous breaches, amount to a repudiatory breach of contract.
The distinction between managing out employees and constructively dismissing them can be blurred in practice. Employers will be at a disadvantage to defend any unfair dismissal claim (or potential discrimination claim) that might result from a constructive dismissal if managers have not followed the company’s procedures and policies (e.g., refusing to provide feedback to an employee as part of the usual review cycle), change an employee’s duties or title without justification, consistently pass over an employee for promotion or decline salary requests without reason. In any event, employees may bring a claim for breach of contract where they are constructively dismissed (particularly if an employee does not have the qualifying period of service to bring an unfair dismissal claim).
Another (and potentially greater) concern regarding quiet firing is its impact on workforce morale and other employees. The perception that employees are being managed out will likely affect workplace culture, shifting the focus from employees’ work and competencies to internal politics and management style.
Frank conversations about performance or progression, while daunting for some managers, play an important role in employees’ development, general expectation-setting and workforce stability. Having direct and consistent discussions about employee progression is more likely to foster a positive environment where employees can pursue career goals (or alternatively find opportunities with another employer). While managing out employees may appear to be an attractive cost-saving measure compared to a formal redundancy plan or dismissal process, companies should not underestimate the corollary impact on workplace culture, employee motivation and the ability to attract new recruits in the future.
A Timely Reminder About Legal Privilege in Communications
A recent case highlights the importance of protecting the privileged status of internal communications that may prevent their disclosure in court.
In the recent case of Sommer v. Swiss Re Corporate Solutions Services, the employment tribunal found that the claimant had been discriminated against by a senior manager, in part because she had not been allowed to work from home when experiencing pain and discomfort during pregnancy. Around the same time, a male colleague had been allowed to work from home for health-related reasons. While the case relates primarily to discrimination and harassment, it also provides a helpful reminder of the basis for legal privilege.
In the course of disclosure in the claim, some internal communications between colleagues at the respondent were disclosed. These included a number of emails from the senior manager in question to the human resources team revealing that he wanted the claimant to be removed from his team. The emails discussed ways of doing this, such as exploring a performance plan or making her role redundant. The content of the emails presented a poor understanding of the law governing unfair dismissal and, because the emails were not correspondence with legal counsel, they were not designated as privileged and were therefore disclosable as evidence.
As a reminder, there are two main types of privilege: (i) litigation privilege and (ii) legal advice privilege. Litigation privilege relates to confidential communications between a client and any third party, created for the sole or main purpose of litigation that has already commenced or is in reasonable contemplation, with the aim of obtaining information or advice about such litigation. Conversely, legal advice privilege attaches to confidential communications between clients and their lawyers, used for the purpose of giving or receiving legal advice. In the Sommer case, litigation privilege did not apply to the emails that were at issue in the claim because litigation was not probable at the point the emails were sent (even though it may have been a possibility and a claim did eventually arise). Legal advice privilege did not apply either, as no lawyers were engaged in the correspondence on the email chains.
They key takeaway is that emails between the businesses and their HR teams will not be legally privileged. Parties should take extra care when seeking internal advice on a tricky employment issue if legal counsel are not part of this correspondence and not giving legal advice. An email to HR flagging an issue with an employee is not necessarily problematic. However, an email sent with the aim of finding a way to dismiss an employee could be used in evidence if a claim is subsequently brought. HR departments should inform legal counsel of any issues as soon as they are aware of them and involve legal counsel in discussions with managers that could potentially result in a dispute.
UK Government Publishes New Guidance on Employment Status
The government has published new guidance to help HR advisers identify the employment status of staff. It serves as a relevant refresher of key employment concepts and applies them in a gig economy context.
In July 2022, the U.K. government published guidance that clarifies the employment rights and protections to which personnel are entitled — including pay, leave and working conditions. According to government ministers, the guide brings together in one place case law on worker status, allowing businesses and workers to access information and improve their understanding of the concept.
Under English law, employment law defines three main categories of individuals: an employee, a worker, and an independent contractor. This categorization is typically described as an individual’s “worker status” and determines, among other things, the employment rights to which an individual is entitled. Correct categorization depends on several factors, which are not always clear-cut. This has led to complex litigation, particularly in recent years. With the rise of the “gig economy” and accompanying emergence of diverse, hybrid models of working, the concept of “worker status” has become increasingly disputed. The guide’s language and case studies are useful for understanding worker status in this new context.
Despite this initiative to clarify worker status under English employment law, the U.K. government, in response to its 2018 public consultation on employment status, has confirmed that it does not intend to reform the U.K.’s separate tax rules with respect to worker status. Under English tax law, similar worker status rubrics exist to determine the tax treatment that should apply to an individual’s working arrangement, but the tax regime contains only two categories of working individuals (employees and self-employed individuals), and the criteria for determining the worker status of each is different from the criteria used in an employment law context. Practitioners and businesses widely acknowledge that the two worker status frameworks (and the interaction between them) are confusing and difficult to navigate and have suggested that legislative reform and alignment between employment and tax law may help to clarify worker status.
However, for now, the U.K. government maintains that “the benefits of creating a new framework for employment status are currently outweighed by the risk associated with legislative reform. Whilst such reform could help bring clarity in the long term, it might create cost and uncertainty for businesses in the short term, at a time where they are focusing on recovering from the pandemic.” While the new government guidance on worker status is useful from an employment law perspective, the extent to which it will help reduce disputes is yet to be seen.
Tailoring Workplace Policies To Accommodate Menopause
A U.K. Parliamentary Committee published a report on the treatment of menopause in the workplace and advocates for broad reforms to workplace policies.
In 2021, the Women and Equalities Select Committee opened an inquiry into the treatment of menopause at work. One of the inquiry’s aims was to “[examine] the extent of discrimination faced by menopausal people in the workplace.” In July 2022, the committee issued its resulting report — a 56-page analysis that details a wide array of possible reforms, ranging from the modest to the more ambitious.
The report cites how government legislation and employer policies shape employee experiences and posits that neither has done enough to effectively support those experiencing menopause. Menopause can affect women in serious ways (including through impacts on mood, sleep, memory and concentration), and the report documents that such symptoms have driven some to leave work, and more have felt unable to seek workplace adjustments due to fear of colleagues’ reactions. At the same time, pre-pandemic data found that women over the age of 50 were the fastest growing group in the workforce. Currently, around 4.5 million women aged 50–64 in the U.K. are employed, and women are staying in the workplace longer. As such, employers have strong economic incentives to revise their policies to account for the effects of menopause, alleviate exit rates among women in this age group and prevent the erosion of an experienced, highly skilled talent pool.
Tweaking existing policies is a helpful step: Proposing additional sick leave for menopausal employees who require it, making sure that absence policies specifically cover menopause-related sickness, adjusting the absence threshold for performance reviews/disciplinary action and offering flexible work more widely are some examples. Also, to address the issue in the workplace culture, companies should ensure menopause (a) is widely understood and taken seriously and (b) discussed on par with other health issues — launching support groups, regularly surveying workforce needs, training supervisors and improving communication with and between HR, well-being and diversity/equity/inclusion functions are all good practice.
The report further finds that the U.K.’s current set of laws offers insufficient protection. Though health and safety laws and the Equality Act 2010 impose a responsibility on employers to shield workers from certain risks, neither explicitly addresses menopause. As menopausal symptoms tend to be intermittent in nature, they may fail to meet the bar required to prove an “impairment” in the context of disability discrimination. Similarly, menopause is an issue that lies at the intersection of sex and age — relying exclusively on either axis fails to capture the dual aspect of menopausal discrimination. The report recommends that the U.K. government appoint a menopause ambassador to work with business and union stakeholders, craft “model” policies and share guidance with employers.
The committee also considered the case for making menopause a stand-alone protected characteristic in the Equality Act, to mirror the way in which sex-specific conditions such as pregnancy and maternity are treated. This option is made more compelling by the fact that menopause applies to more women than those who experience pregnancy or maternity. Notably, the concept of intersectional discrimination has been previously proposed and then rejected before implementing legislation was drafted, but the report also considers the concept of allowing “combined” or “dual” discrimination claims, whereby successful claimants would have protections if they can show that their less favorable treatment relates to a combination of characteristics.
There are doubts as to whether the government will take action on these proposals, but it may draft improved guidance to give employers a deeper understanding of the existing legal position. For employers, there is a clear, arguable case for reforming their workplace policies: to better attract and retain employees who are experiencing or have experienced menopause, build a reputation as an accommodating employer and reduce the risk of menopause-related discrimination claims.
UK Incentives and Remuneration Update
On September 23, 2022, the chancellor of the Exchequer delivered what has been termed his “mini-budget.” In the weeks since, though the government has rolled back (or at least paused) certain of the key measures that had been announced, the measures that remain likely to come into force will be important to U.K. employers. Changes to certain U.K. tax-advantaged share schemes have also recently been announced. These measures and changes will impact how U.K. companies engage with and recruit, retain and incentivize their workforce.
Income Tax and National Insurance Contributions (NIC) Rates
Although the headline policy to abolish the additional rate of income tax (currently 45%) has been paused, starting on November 6, 2022, the rate of employers’ and employees’ NICs is still expected to decrease by 1.25%. This development comes after rates were temporarily raised (by 1.25%) in April 2022, with NIC rates returning to 2021/2022 levels in April 2023, at which point officials planned to introduce a separate Health and Social Care Levy tax (also of 1.25%). However, legislation has been put before Parliament to withdraw the Health and Social Care Levy tax. These changes will, in effect, return the NIC rates and social security levies generally to their pre-April 2022 positions. The mini-budget statement also announced a reduction in the basic rate of income tax from 20% to 19% starting in April 2023. However, the current chancellor has now ruled out this reduction.
The chancellor of the Exchequer will deliver the U.K. government’s autumn statement on November 17, 2022, and we expect further changes at that time.
Tax-Advantaged Share Schemes
The mini-budget announced certain changes to the Company Share Option Plan (CSOP) regime. The CSOP is a form of U.K. tax-advantaged share scheme providing for the grant of share options that may benefit from income tax relief on exercise, provided that the options meet certain prescribed requirements, including that they are granted with an exercise price equal to the market value of the underlying shares at the time of grant (but ignoring any restrictions applicable to such shares). The CSOP legislation is more restrictive, and the potential tax benefit offered thereunder substantially more limited, than under the more flexible Enterprise Management Incentive (EMI) legislation (although CSOPs can be utilized by companies that are too large to offer EMI options). Starting April 6, 2023, the maximum value of shares over which a participant can hold CSOP options has increased from £30,000 to £60,000, and certain of the requirements that shares subject to the CSOP options must meet have been removed, making the CSOP potentially more attractive to a broader range of companies and participants than previously.
On October 14, 2022, HM Revenue and Customs (HMRC) released a long-awaited update to its guidance on the use of discretion regarding EMI options. Prior to this update, HMRC’s position and published guidance were generally accepted to be unclear and potentially incorrect in some cases. The exercise of discretion can, in certain circumstances, lead to companies inadvertently changing a fundamental term of an EMI option, thereby causing its release and the grant of a new option. This new option may not qualify for EMI treatment and, even if it did, the relevant tax benefits would not be as great if the market value of the relevant shares had increased since the initial grant. The most important and useful aspect of the updated guidance is the distinction drawn by HMRC between specified event (i.e., exit-only) and time-based EMI options. For specified event EMI options, discretion to accelerate awards will typically be allowable on the basis that “when” the option is exercisable is not changed (even though the timetable has). In contrast, acceleration of time-based EMI options would typically amount to a change of a fundamental term, which would in turn cause a release and regrant of the option, as detailed above. Where EMI options contain features of both time-based and specified event EMI options, HMRC has confirmed that discretion to accelerate based on the occurrence of a specified event will be permitted. In all cases, the EMI plan rules must provide for such discretion. While only guidance, these updates give much-needed certainty on the issue and greater flexibility to directors and scheme operators to waive or accelerate certain vesting conditions (in particular, with regard to performance vesting).
In Employment Related Securities Bulletin 46, published on October 14, 2022, HMRC stated that it had previously provided advice to taxpayers that it now believes contradicts its updated guidance and that in some cases such advice may have resulted in the overpayment of tax on the exercise of options. While HMRC is attempting to contact taxpayers that it believes received such advice, any companies that have operated EMI schemes in the past and withheld tax in connection with such advice from HMRC may wish to consider whether tax may have been overpaid.
Another tax-advantaged share scheme subject to potential changes is the Save As You Earn (SAYE) plan. The SAYE plan operates on an “all-employee” basis (unlike the CSOP and EMI plans, which are discretionary). Participants are granted share options with an exercise price set at a discount of up to 20% of the market value of the underlying shares at the time of grant, with the exercise price funded by the participant’s savings, made on a monthly basis from post-tax salary via payroll. The SAYE plan legislation includes a bonus rate mechanism under which participants may in certain circumstances be entitled to a tax-free bonus. The bonus rate is currently 0%, and has remained at this level since December 2014. The U.K. government recently announced a review of the bonus rate mechanism to address the plan’s “extreme complex[ity],” and Employment Related Securities Bulletin 43 confirms that while authorities conduct that review, the bonus rate will remain at 0%. Practitioners anticipate that the review will facilitate a potential increase in the SAYE plan bonus rate in line with the recent (and further anticipated) changes in U.K. interest rates, in order to ensure SAYE plans remain a relevant savings plan option for participants and an effective incentive plan for employers.
A recent parliamentary debate on SAYE and Share Incentive Plans (another “all employee” tax-advantaged share scheme) again noted the increasing gap between how tax-advantaged share schemes operate and the modern reality of U.K. employers and employees. As previously highlighted in our September 28, 2020, client alert “A New Focus on UK Tax-Advantaged Share Schemes” and the November 2020 UK Employment Flash, elements of certain tax-advantaged share schemes discourage or prevent wider participation, resulting in long-standing calls for more flexibility in the permitted terms and requirements of such arrangements so that they are available and attractive to a wider range of companies, industry sectors and participants. The announced changes to the CSOP and SAYE regimes and the long-awaited clarification on the use of discretion in EMI schemes show a commitment by the government to ensuring the effectiveness and continued relevance of tax-advantaged share schemes. Employers and plan participants should monitor potential further review and reform efforts.
For further information, please contact:
Helena J. Derbyshire, Partner, Skadden