Ethnicity pay gap reporting: how would it work?

Both the Conservatives and the Liberal Democrats have pledged to introduce ethnicity pay gap reporting if they come to power on 9 June. But how would this policy work in practice? Rob Moss examines pay gap reporting for ethnic minorities.

At first glance, the idea of emulating the new gender pay gap reporting requirements to create a similar tool to measure employers' ethnic diversity seems to make sense. Take the regulations used for gender and apply them to race in much the same way. One protected characteristic is simply swapped for another.

However, it's one thing to compare men's pay to that of women's; it's another to compare white British employees' pay with the pay of other ethnic groups.

The Office for National Statistics (ONS) groups individuals into 19 ethnic groups. Will employers be expected to compare their white British pay with that of the other 18 ethnic groups?

Mark Crail, head of salary surveys and benchmarking at XpertHR, says: "Where it is possible to calculate the gender pay gap as the difference between two variable figures, doing the same calculations on so many variables would be a mathematical nightmare - if it were even possible.

"Doing this properly would often mean reporting not on the ethnic pay gap, but on many pay gaps between people of different ethnic origins. The sheer number of options could also mean that it becomes difficult to report any meaningful results if the number of individuals from specific ethnic backgrounds is small."


The manifesto pledges follow a recommendation from Baroness McGregor-Smith in February for Government to legislate for businesses and public bodies with more than 50 employees to publish workforce data broken down by race and pay band.

The ONS often refers to five "main ethnic groups": white, Asian, black, mixed and "other". If an incoming Government decided to compare the pay in these groups, that would simplify matters to some degree. Nevertheless, you would still be comparing white with four other groups.

The binary option, the one most similar to a gender pay gap between male and female, would be to compare white British with all other groups, essentially the widely used term, "black and minority ethnic" (BME).

It is what the ONS calls the ethnic minority group, which accounted for just under 20% of the population in the 2011 census. So, what would a comparison between white British and BME actually tell us?

"Whereas gender (in the eyes of the law at least) is binary, the attribute of race sits in a far broader range," says Denise Keating, chief executive of the Employers Network for Equality and Inclusion (ENEI).

"Simply splitting the workforce down to white or BME is obviously flawed, as nationality must be considered, and sub-groups within both groups will have very unique cultural differences. Even the Government's own census options for race can't be considered perfect."

If the aim of ethnic pay gap reporting is to highlight race inequality, the binary option is not going to work. By gathering so many minorities into one "BME" group, you are grouping people together who may face different levels of ethnic disadvantage. This would skew any ethnic pay gap and make it of questionable value.

Melanie Stancliffe, partner at law firm Irwin Mitchell, suggests comparing each ethnic group with the overall mean, rather than each group with one another. "The number of groups will need to be limited if the data itself is to have any meaningful impact."

"The devil will be in the detail," she explains. "But, as for gender pay, it will require businesses to look at their pay bands and remuneration packages to ensure they are not treating employees differently for unlawful reasons."

She adds: "A fair and transparent approach has to be the right way forward."

Gathering data

While the idea of ethnicity pay gap reporting clearly has admirable intentions, there is also the question of where the data comes from.

"Whilst employers will usually hold gender details of employees for insurance and pensions purposes, there is no current requirement to hold details on ethnicity," points out Keating. "Any attempt to gather that data from an existing workforce could be viewed suspiciously."

Crail agrees: "Employee ethnicity data is provided voluntarily, so it will be incomplete. The strength of gender pay gap reporting is that, because it is based on virtually all employees, it gives an accurate picture of the structural inequalities within an organisation.

"If you only have data on, say, 75% of employees, you can't really do that analysis with any confidence."

Stancliffe adds that while the introduction of gender pay gaps reporting means the infrastructure for analysis is there, ethnic diversity data is not universally collected. "Employees would have to confirm they identify with an ethnic group, which some may prefer to avoid," she says.

Crail adds: "It could be the case that those employees who are most concerned about discrimination at work are the least comfortable at telling their employer about their ethnic origin. This would distort the figures, potentially quite significantly."

Writing in her review into race in the workplace, Baroness McGregor-Smith acknowledges this issue: "When a new employer asks for your bank details, you know this is so you can be paid and so you fill it in. The same is not true when considering ethnic origin and employers should consider being clearer about why they are collecting the data."

Indeed a further recommendation is for employers to consider taking positive action to improve reporting rates among their workforce. This should include clearly explaining how supplying ethnicity data will assist the company in increasing diversity overall.

Employees' willingness to declare their ethnicity to an employer for the purposes of pay gap reporting might be further reduced if they also fear being identified.

Dr Jill Miller, diversity and inclusion adviser at the CIPD, says there are concerns around confidentiality in ethnicity pay gap reporting: "McGregor-Smith's recommendation is for listed companies and organisations with more than 50 employees to publish this data. However, employees risk being identified if a particular ethnic group is underrepresented in a certain pay band, especially given the ONS uses 19 categories of ethnicity.

"In contrast, if higher-level categories of ethnicity were used there is the danger of the real issues becoming hidden, as we know from research that different ethnic groups face different challenges between industries."

Keating highlights: "It is one thing to be the only woman at a particular level in the organisational hierarchy, and quite another to be, for instance, the only person from a Bangladeshi origin in the whole organisation. This is an issue that will have to be thought through in far more detail than the current arrangements for gender."

The Data Protection Act holds that information about someone's racial or ethnic origin is sensitive personal data; their gender is not. "Where there are relatively few individuals from a particular ethnic background," explains Crail, "it is going to become very hard to maintain those individuals' privacy in any public reporting of the data.

"Employers will also need to know that they are staying within the much tighter legal rules that apply to anyone processing this sort of sensitive personal data."

Detail of what the Conservatives and Liberal Democrats - who also want pay gap reporting for sexual orientation - have planned may well emerge in the run-up to the election, but it is clear that a drag-and-drop approach to tweaking the gender pay gap reporting rules is unlikely to work.

The challenge will be in creating an ethnicity pay gap measure that is detailed and universal enough to convey meaning, while being both an exercise that is not overly onerous and cumbersome for employers, and something that employees want to participate in.

"The Conservative's manifesto pledge to introduce ethnicity pay gap reporting contrasts with Labour's pledge to introduce equal pay audit requirements," says Keating. "Whilst both policies are thin on details, evidence suggests that the reputational risks of public reporting tend to focus the minds of senior executives more than internal reporting.

"However, we are as yet unable to judge the impacts - if any - of the nascent gender pay gap reporting requirements."

Fit for purpose

A Tory or Lib Dem government would inevitably consult with the HR community to ensure that race pay gap reporting proposals were "fit for purpose". If they were not, Miller fears that "pay reporting could end up being seen as a burdensome tick box exercise that's another cost of doing business, rather than a driver of workplace, economic and societal change".

She adds: "It needs to be made clear that just encouraging data reporting is not sufficient to drive the real change we need to see, but rather a way of highlighting which gaps exist where.

"As with gender pay gap reporting, what's important isn't where you start, but having a narrative about where you are trying to get to, how you will do it and how long it will take."

Crail believes that it is already clear that gender pay gap reporting is a relatively blunt tool: "It looks at one aspect of the problem while completely overlooking other important issues such as unequal pay and, significantly, the sheer lack of either men or women in some organisations.

"It is quite possible to employ virtually no women but to have no gender pay gap because the relatively few women in the organisation are in reasonably well paid jobs. These problems would be magnified several times over in reporting on ethnic pay gaps."