My heart skipped a beat when I discovered the Royal Statistical Society was publishing a report, out today, which calls for the rules around gender pay gap reporting (companies with 250+ employees are mandated to calculate and publish this data by law) to be refined.
As I wrote on Coffee House last week, the calculations are so crude and void of context, they render the results almost meaningless. They don’t take any like-for-like comparison on job, age, or education into account and don’t even control for full and part-time workers. As a result, the final figures are comparing the CEO of the FTSE 100 company to the junior researcher who just graduated from university.
The flaws in the legislation are so egregious and so obvious, it was simply a matter of time before there were calls to improve the reporting system, or scrap it all together.
Or so I thought.
Unfortunately today’s report from the RSS only serves to double down on the deeply entrenched problems with the pay gap reporting. It makes ten recommendations (in bold below), only one of which – to keep the current reporting threshold at 250 employees – would actually help:
1) Gender pay data should be presented in a clearer and unambiguous format – always in pounds and pence, not percentages.
The RSS argues that presenting the data in pounds and pence (i.e. for every £1 that the median man earns at ITV, the median women earns 87p), and scrapping percentage reporting altogether (i.e. median hourly pay gap at ITV is 12.6 per cent in favour of men) makes the data “clearer, more helpful and more intuitive.”
They are wrong; it does the opposite. Saying that a women, even an average-earning woman, earns 13p less on the pound than a man, contributes to the misleading narrative that women do not receive equal pay for equal work.
Such a statement would be categorically untrue for many women working at ITV; yet they will read it and be made to think they are less valued, and less compensated, than their male colleagues.
2) Improve government guidance to employers, especially over the calculation of the median gender pay gap.
3) Similarly, improve the government guidance to employers over the use of income quartiles.
The guidance around pay gap reporting could be clearer, and RSS are right that many companies have been failing to calculate their pay gaps correctly. (NWN Media Limited was lambasted last year for having a pay gap of 85.2 per cent; only to be discovered, after they revise their data, that it was a pay gap in favour of women.)
But their criticism of the guidance illustrates just how malleable and inconclusive the data is. There are infinite ways to calculate gender pay gaps, which has been unintentionally highlighted by companies who are calculating in differing ways. Small changes, omissions, inclusions, and different kinds of breakdowns can result in varying results.
Indeed, you can come up with almost any pay gap you want, depending on what you control for. The key is to discover which ones matter, and which ones don’t.
4) Provide free online calculators to help increase the accuracy of gender pay gap reporting.
I am both amused and baffled that the team who put together the RSS report appear not to have typed ‘online calculator’ into Google before.
These calculations are nowhere near as difficult to make as the RSS make out. A special, government-funded tool is not required. Men and women can calculate pay gap data, and we can analyse pay gap data – that’s why we know it’s bogus.
5) Ensure that online gender pay calculators have built-in ‘sanity checks’, to ensure accurate reporting and prevent statistically implausible entries.
Imagine if all these resources proposed to tweak flawed pay gap reporting went towards something that actually helped women – say, funding employment tribunal, to hold to account the rare employer taking part in discriminatory activity.
6) The gender pay gap should be calculated by quartile as well, to make it harder for the system to be ‘gamed’.
This is perhaps the most concerning recommendation from the RSS, which argues that employers could ‘game’ the system be ensuring the median man and women are paid the same (resulting in no gender pay gap), with pay discrepancy still taking place throughout the organisation.
This omission from the RSS, that employers will put their pay gap reporting above the well-being of employees, is exactly the concern that has been voiced for years over these reporting measures. They create terrible incentives – like not hiring a large intake of female graduates, for example, or not offering employees salary –salary-sacrificing benefits, which are often taken up by women.
7) Publish each employer’s annual results, side-by-side, to make it easy for trends and progress to be assessed.
The premise of this recommendation is that showing comparative data, year on year, can help the reader gauge if the company is ‘progressing’ on its gender pay gap (that is, is the gap widening or lessening over a period of time).
But a company’s ‘progression’ or ‘lapse’ still won’t tell us anything meaningful.
This year, EasyJet’s pay gap worsened, from 45.5 per cent in 2017 to 47.9 per cent in 2018. But EasyJet actually increased their intake of female pilots; hiring primarily women into cabin crew roles skews their pay gap in favour of men, despite all employees in similar roles being paid the same for their work.
EasyJet could ‘game’ the system by hiring fewer women as cabin crew, and giving those jobs to men instead. That would solve their pay gap woes.
Does anyone think that’s a good outcome for working women?
8) Organisations which employ fewer than 100 women (or men) should be flagged.
The RSS argues that companies that employ a small number of women (or men) should be flagged and forgiven, as they can “appear to have sizeable gender pay gaps even when…they aren’t operating in a discriminatory way.”
First, these figures – which are void of like-for-like comparisons, do not provide us with any compelling evidence that companies with pay gaps are operating in a discriminatory way.
Second, if the RSS are willing to concede that pay gap calculations might be skewed because of gender breakdown, why aren’t they willing to admit they might be skewed because of roles; or age, or education, or countless other factors? Indeed, organisations are made up of individuals from all walks of life, and their salaries are set to reflect their unique backgrounds. Why single out gender?
9) Keep the current reporting threshold at 250 employees.
This is the one positive recommendation of the bunch.
It’s a shame they ruined it by asserting that small companies had to be protected from unjust accusations of “gender-based pay discrimination”, because in a company that’s small, “the difference between their men’s and women’s earnings could be down purely to chance.”
But the gender pay gap is not a measurement of equal pay for equal work. The former can be calculated in a million different ways, is is more or less meaningful based on context. The latter has been illegal since 1970.
10) Improve statistical skills among human resources professionals.
Rather than improving the statistical skills of HR staff, how about we just improve the statistics, which are at best meaningless, and at worst creating poor incentives for employers to hire or retain women in lower-paid roles?
The RSS is right on one account. The pay gap reporting exercise has “provided a vivid demonstration of statistics’ power and potential.” From the lips of politicians to the media headlines, the brutal manipulation of statistics has created a narrative that women are victims in the workplace, being paid less than men and stuck somewhere between their desk and a glass ceiling.
But this couldn’t be further from the truth. 2019 is the best on record to be a women in the workplace. Unfortunately, the RSS’s recommendations will do very little to bring that to light.
Kate Andrews is Associate Director at the Institute of Economic Affairs