We have spotted a worrying trend in gender pay gap reports calculated internally by organisations over the last 24 months. It concerns the distribution of men and women throughout the pay range of an organisation. The gender pay gap legislation is crystal-clear regarding how a list should be split to find the gender balance in each of the four quarters, but by using the word quartile, confusion and incorrect application have crept in.
So what?
Making assumptions on incomplete or wrong data can lead to wasted time, wasted money and poor plans which can hinder progress on your pay gap journey. Even if you are confident enough to navigate the legislation, it pays to seek third party verification from an expert… So, what is the difference between a quarter and a quartile?
Quarter vs Quartile
Imagine a pizza sliced into four equal chunks. The initial quarter is the whole first slice whereas the first quartile marks the end of the first slice and the beginning of the second. Ok, but that’s just semantics, right? They are broadly describing where the first bit ends and the second starts. Well yes, and no – the next two paragraphs assume that calculations are being done using statistical quartiles instead of the Government’s Gender Pay Gap Reporting methodology.
When is a quarter the same as a statistical quartile?
Below is a list of eight numbers, the arrows show the quartile 1, 2 & 3 values. You can see that these arrows separate the list into four even quarters (purple bars) with two numbers in each. Perfect, the quartile boundaries separate the list as requested by the legislation.
What’s the potential problem?
In gender pay gap reporting the individual values are adjusted hourly rates of pay. These can be worked out using complicated calculations or the employer can declare a set hourly rate as the fairest reflection of an employee’s pay. Imagine a cleaning company who apart from head office staff and regional management, have a core team of cleaning staff with a few tiers of experience. Each of these tiers of staff may get exactly the same pay regardless of gender. If the hourly rate for these employees coincides with a quartile boundary then the final list isn’t four equally sized chunks anymore. By changing the values in our last example, we can demonstrate this below.
Now we see that the first chunk has three values and the second has only one. Whilst statistically correct, the list is no longer split into four equal chunks and therefore does not conform to the legislation.
What data does this misinterpretation of the legislation affect?
This error only affects the proportions of male and female full-pay relevant employees in the four quartile pay bands. And here’s a fundamental reason why employers may be making this error – the phrase quartile pay bands suggests that this part of the methodology is calculated using statistical quartiles. However, the legislation simply asks that the list of full-pay relevant employees is divided into four equal quarters with any remainder being spread evenly throughout the entire range.
How do we know that some employers are calculating this incorrectly?
A list of any number of values may be split into four equal chunks with a maximum of three remaining values. Therefore, the maximum difference in size between any two of the four chunks should only ever be one. Where employers have been detailed enough to include the numbers of males and females in each quarter, and that quarter-difference is greater than one, we can categorically state that the wrong methodology has been applied. There are also other clues in the accompanying narrative, but not as explicit as the quarter difference described here.
Real-world example
ABC Limited has 747 employees of whom 705 are full-pay relevant employees. Their mean pay gap is 20.0% and their median pay gap is 31.2%. Whilst the selection of quarter/quartile methodology has no bearing on these pay gap figures, it has an impact on the powerful insights that the pay range analysis shows. Their overall ratio is 55:45 male/female; if they are striving for a zero median pay gap then the gender ratio in all quarters of their pay range should have this ratio. The two charts show the correct (quarters) and incorrect (quartile) methods of calculating these key metrics. In very simple terms, the correct interpretation directs us to look at the distribution in the upper and lower quarters whereas the incorrect interpretation would have us perform extra work and examine all but the upper mid quarter. Also, errors can be amplified with lower headcounts and less balanced workforces.
ABC Limited has 747 employees of whom 705 are full-pay relevant employees. Their mean pay gap is 20.0% and their median pay gap is 31.2%. Whilst the selection of quarter/quartile methodology has no bearing on these pay gap figures, it has an impact on the powerful insights that the pay range analysis shows. Their overall ratio is 55:45 male/female; if they are striving for a zero median pay gap then the gender ratio in all quarters of their pay range should have this ratio. The two charts show the correct (quarters) and incorrect (quartile) methods of calculating these key metrics. In very simple terms, the correct interpretation directs us to look at the distribution in the upper and lower quarters whereas the incorrect interpretation would have us perform extra work and examine all but the upper mid quarter. Also, errors can be amplified with lower headcounts and less balanced workforces.
Three steps to avoid making this mistake in you gender pay gap report
- Don’t use the word Quartile – to use the pizza analogy, you should talk about the slices not the cuts.
- Split your completed list of adjusted hourly rates of pay into four equal chunks. Remember, you can have up to three extra people, so increase the quarters using the spares as follows: 1 spare = lower-mid quarter, 2 spare = lower-mid and upper-mid quarters, 3 spare = all but the upper quarter.
- If you do have identical hourly rates of pay that span more than one quarter, then don’t forget to apply the final bit of legislation magic to ensure that you have a fair gender balance in each of the affected quarters.
What should I do if I made a mistake in my gender pay gap report?
If you think you’ve mixed up quarters and quartiles in a past submitted report, don’t panic. It’s best to be transparent about your mistakes and rectify them, so that progress in years to come can be accurately traced back. You can resubmit previous reports to correct mistakes like this.
There are many other areas of the legislation that are easy to misinterpret, even if you are confident, it always pays to seek third party verification.
Get in touch if you would like to speak to a pay gap specialist.
