What next for pay gap reporting?
First it was gender, with the regulations requiring companies with 250 or more employees to report on their gender pay gap coming into force in April 2017. Then it was executive pay, with the Government announcing in August that it will require listed companies to report on the executive pay gap by publishing the ratio of “CEO pay” to “average pay” (see this blog post for more information).
So what next?
In the run-up to the general election in June, both the Conservatives and Liberal Democrat manifestos promised to introduce ethnicity pay gap reporting, with the Liberal Democrats also proposing sexual orientation pay gap reporting. More recently, in August the Equality and Human Rights Commission published its strategy to reduce pay gaps in Britain; this proposed a consultation on the most effective way to extend reporting requirements to ethnicity and disability pay gaps (on a related note, the EHRC also has some interesting infographics to visualise various pay gaps – see here for ethnicity and here for disability).
It seems that pay gaps related to a number of protected characteristics are susceptible to future reporting obligations. But what would these obligations look like? It will not be as simple as replicating the existing regulations for reporting the gender pay gap for each characteristic; each particular characteristic will bring additional challenges.
- Whose pay should be compared? While gender (for pay gap purposes at least) is treated as binary, other protected characteristics cover a far broader range. For example, the Office of National Statistics provides for 19 ethnic groups. Should the pay gap between each of these 19 groups be reported (which would not only be extremely complicated but, due to the sheer number of different groups, could lead to meaningless results)? Should it be the pay gap between a more limited number of groups, such as the five “main ethnic groups” (white, Asian, black, mixed and “other”) identified by the ONS (still complicated and still potentially meaningless)? Or should ethnicity be treated as binary and the pay gap between white British and all other groups reported (a huge over-simplification which will miss the nuances of pay gaps between different ethnic groups)? A similar issue arises with disability – the EHRC tool gives an indication of the number of potential groups when considering different disabilities.
- How will employees be identified as belonging to a particular group? While many employers will have details of the gender for their employees (e.g. for pensions purposes), there is no obligation for employers to hold details of ethnicity or sexual orientation and employees may be sensitive about providing this information. if employees are asked to self-report the data received may not be accurate; for example, employees who fall within the legal definition may not consider themselves to be disabled.
- How will such data be linked to individuals and therefore to pay? Even where data about protected characteristics is recorded, it is unlikely that it will be held in a way that connects it to a particular individual, so linkage to pay is likely to be impossible without collecting a new set of data connecting protected characteristics to specific individuals. This creates a whole new conundrum from a data protection perspective.
- How will employees be protected from identification by co-workers and the wider public? As the number of groups increases, so too does the risk that there will be very low numbers of employees belonging to certain groups, either within pay bands or in the organisation as a whole, potentially making individual employees easily identifiable.
- How can the results be used meaningfully? While the infographics on the EHRC website are diverting, they are based on the Labour Force Survey data from 1997-2014 (which covered 60,000 households each quarter). It is highly unlikely that employers, even those with thousands of employees, will be able to produce meaningful comparisons – for example, a large pay gap for a particular disability could be driven by there being a single employee in that disability who happens to be an a low-paid position, rather than any other reason.
It is therefore unlikely that existing gender pay gap reporting processes can be copied to meaningfully report the statistics for other pay gaps. Perhaps the focus should be on initiatives to ensure pay gap equality across the board, without the distraction and additional costs of statistical reporting.