So here they are, out yesterday, a strange parallel universe where months last 30.44 days and years 365.25, and where you don’t include pay for periods of leave except when you do.

In past blogs here we have criticised Government Regulations and statutory Guidance as too vague, leaving employers unclear whether they are caught by (or complying with) the legislation or not. That is not a criticism which can be levelled at the draft GPG Regulations. These are, for the most part, exceptionally, indeed fantastically and pointlessly, prescriptive. I say that not out of any disregard for the Regulations’ very worthy aims, but because they aim at millimetric accuracy in circumstances where there are no viable means of checking it, no real legal sanction for getting it wrong and, where give or take a few percent, it doesn’t matter legally anyway – at risk of labouring the point, the existence of a pay or bonus gap does not in any way prove (nor necessarily even suggest) any unlawful pay discrimination.

So here we go with some key points from a first read of the draft Regulations:

  • By Reg 1, the comparison is to be made between male and female “full pay relevant employees”, meaning those who are not earning less during the relevant pay period (basically, April each year) because they are “on leave”. “Leave” includes family-friendly and sickness absences, so if you are on nil or reduced pay on those grounds, you drop out of the averaging process. That is sensible, since more women than men take family leave and women take on average some 50% more sickness absence than men, so including those payments would distort the pay gap even if basic salaries were exactly the same. On the other hand, the definition of pay includes the premiums paid to night shift workers. They are predominantly men, so the scope for a perceived inequality without a real one is still alive and well.
  • However, by Reg 3, the “ordinary pay” to be included in the calculation includes leave pay, which it can’t if the definition in Reg 1 is correct UNLESS your leave pay is the same as your normal pay, in which case why not just use ordinary pay from the start and omit the carve-out for leave payments altogether? This means that before you can exclude someone from your averages because they were on leave, you’ll need to know whether they were on full pay or not, an entirely gratuitous over-complication.
  • Ordinary pay does not include “remuneration provided otherwise than in money” (Reg 3(2)(d)) and so appears not to address benefits provided in the form of goods or services. On the face of it, therefore, a car allowance counts as pay, but the provision of a car does not.
  • The Regulations require reporting of mean and median percentage gaps in respect of pay and bonus separately. However, the definition of pay for that purpose includes bonus but essentially only if it is paid in April. If you pay it in some other month, then it drops out of the pay calculation altogether. For the bonus gap calculations, the sum is annualised and so it doesn’t matter when it is paid.
  • If the Government’s aim was transparency, it would surely have been easier to have regular pay and allowances measured men as against women plus bonuses and other variable figures done separately, annualised to give an average. Using bonuses either once or twice depending on precisely when they are received does not seem to shed the light on how the genders are paid that the Government must have hoped for. Take this example: A male employee earns a basic £50,000 and a bonus of £10,000, while a female colleague makes a basic £60,000 but no bonus. If his bonus is paid in April there will be no obvious pay gap, even though his salary is adrift of hers by 20%. If his bonus is paid in some other month that gap will suddenly appear in her favour, while the bonus gap will go in the opposite direction, even though ultimately they are paid the same overall. Quite what lessons the casual reviewer of GPG reports is supposed to take from this is unclear.
  • Then we come to a key component of any assessment of average hourly rates, i.e., knowing the number of hours actually worked. The average hourly rate of an executive working 9 to 5 is very different from one on exactly the same salary who does 8 to 7. Veering suddenly away from the minute precision of the rest of the calculations, Reg 7 tells us that when you don’t really have normal working hours, the employer should use “a number which fairly represents the number of working hours in a week having regard to” the average hours worked by that employee and others in the same role. It is hard to see any employer doing more than sticking a moistened finger in the wind on that one, absolutely not bearing in mind in any way that the longer the deemed average day at senior (male) levels, the lower the hourly rate and the smaller the corporate pay gap will become.
  • By Reg 7(4) the averaging process for weekly hours excludes weeks in which no hours were worked, but seemingly not those in which one hour or more was done. So if this calculation is carried out scrupulously (which surely no-one will), an employee who took four days’ paid leave in that week will have in it an hourly rate five times that of his colleagues without actually having been paid any more than those who weren’t away at all.
  • But my favourite piece of legislative over-engineering is Reg 13(1). This sets out how to calculate the percentage of male full-pay relevant employees in a quartile pay band (let us say 60% for the sake of argument) and then adds a whole new paragraph to enable you to work out separately that this means that 40% are women. What, really?

Right at the end of the draft Regulations is a sad little apology for not getting the balance between detail and proportionality remotely right here. Within the next 5 years, says Reg 16(2), the Secretary of State must review these rules “and assess… the extent to which [their objectives] could be achieved with a system that imposes less regulation“.