Last week the government issued its first official guidance on ethnicity pay gap reporting.  Somewhat unusually among workplace guidance, it is prospectively a very useful read. To its immediate credit, for example, it accepts right up front that there can be many legitimate reasons for disparities in average pay between ethnic minority groups.  “It should not be assumed” [as the Court of Popular Opinion has done in relation to gender pay gaps] “that any disparities are necessarily a result of discrimination”, it says.

The guidance is designed to develop a consistent approach allowing employers to take “meaningful action while remaining proportionate and without adding undue burdens on business“.  This sounds fab, though anything which mirrors the Byzantine calculation processes strictly required under the Gender Pay Gap Regulations can scarcely be said not to add undue burdens.  The guidance shies away from replicating them in full.  It is not clear whether this is because bigger employers will already have done those numbers for all their staff to comply with the GPG Rules or because others are best not put off ethnicity pay reporting altogether by sight of the full horror of those calculations if done by the book.  Instead it provides a somewhat simpler set of pay determination principles, though these still run to a very discouraging 20-plus pages. Little secret is made of the complexity of the calculations suggested – these should even be “checked with analysts“, says the Introduction. No undue burdens, then.

More than pay, this guidance focuses on the real messages to be taken by employers from the data to be revealed – lower pay among a particular ethnic group may be because that group disproportionately occupies the more junior positions in the organisation, not because of any like-for-like differences in pay in those positions.  The question which that should flag up is then whether the employer is doing enough to provide adequate progression opportunities for people in that group.  Why do they only reach so far up the organisation?  What is stopping members of that group occupying a proportionate number of more senior roles?  Why do they mostly apply for the junior roles in the first place?  Does that group suffer from particularly high turnover?  In the end, as the gender pay equivalent should have done, the new guidance suggests a much more holistic and wide-ranging set of questions around internal and external factors which the employer can sensibly ask itself than simply around equality of pay. 

All good so far, but how do you actually go about this?  After all, in gender reporting your available comparator groups are pretty limited, but ethnicity is clouded by multiple issues of parentage, heritage and self-identification. So which ethnicity groups should you use?  The guidance cautions strongly against simply dividing your workforce up into whites and others, since the aggregation of all ethnic minority groups into one will obscure particular disparities.  According to the Introduction, some minority groups earn on average more than their white counterparts and some less, and those differences will be lost if they are just all lumped in together.  However, none of this is mandatory and so if you have a workforce which is substantially made up of two or three different ethnicities only, you will not materially devalue your conclusions by focusing on those groups and not having a separate calculation for very much more minority minorities, if you see what I mean (which may also run you into difficulties in maintaining confidentiality in any case).  The groups used are left to the discretion of the employer, but with the caveat that you will need to be able to defend that choice if challenged – while you don’t have to do this in the first place, it will have to look like a reasonable stab at a genuine enquiry if you do.

Whatever the groups used, you will obviously need to ask about ethnicity as this may not be information you already hold.  To be safest, the guidance recommends initially seeking responses in the categories used in the 2021 Census (2022 in Scotland), making sure that you include a “prefer not to say” option.  If a material number tick that box or simply ignore your request altogether, that category should be separately reported upon. Remember that the answers given will count as special category data under the GDPR and so must be kept secure and with circulation limited so far as practicable.  With confidentiality/GDPR in mind, the guidance recommends a minimum category size (i.e. below which there is an unnecessary risk that individuals can be identified) of 5 to 20 for reports used internally only, and of at least 50 if they are to be published externally.

Having obtained that individual ethnicity information, the guidance recommends that for the sake of doing the comparison it should then be distilled into five aggregated groups – Asian, black, mixed, white and “other” — so there is actually then a degree of lumping together after all, but at least on an informed basis. The government has no immediate plans to introduce mandatory ethnicity pay reporting, so why bother with any of this?  And if you do it anyway, why be that fussed about doing it right?  The answer to the first is that in due time there is a likelihood that this will become law, so best to get into the habit now. The answer to the second is that while your calculations do not have to be as millimetrically precise as under the GPG Rules and you do have some discretion as to the groups you choose, you will lose both the PR benefits of being seen to make a genuine effort on this and (at least as importantly) all the valuable learnings which an employer can gain from the exercise.  Half-cock reporting will produce half-cock outcomes and no-one will gain anything for their time.