A latest paper on the recently published draft regulations on Gender Pay Gap Reporting.
Download the full report here: Gender Pay Paper or read below.
GENDER PAY GAP REPORTING: MAKING MEANING OF MANDATE
PHASE 3 CONSULTING 18.03.16
Getting Gender Pay Gap reporting right is an opportunity – and we have a new mandate for sizable organisations to do so just confirmed.
There are a few reasons to be interested. Let’s start with compliance as a basic for the need-to-know. Or you might find figures interesting – how in the case of pay these can be presented or how your HR system is going to arrive at those figures. And the meaning of course lies in the fact that reporting pay gaps opens eyes and offers confidence when it comes to a question about our working world that strikes to the heart of the ability for optimum organisational productivity, ethics and social opportunity. Mandate, maths or meaning, I hope you will share my interest therefore.
A Gap in the Pay or a Gulf in the Reporting?
Prior to the publication of Draft Regulations in February, I had feared that we could be headed for gulfs, let alone gaps. I feared that the differences may not reflect the pay itself, but in the differences that reporting methods, tactics and uncertainties could create. Post-publication I take a degree of heart and move on to look at why as professionals we would be wise to look ahead to getting the best out of a new “must do”.
In this discussion paper, I outline the requirements of the draft regulations, explore some of the more interesting implications for professionals in HR and in systems alike and offer some advice to get ahead.
- To be able to explain your data and why it matters – understand the different metrics and why gender pay rules need to include or exclude those factors
- To talk the language of your colleagues on gender pay and its analysis – see the perspective of a systems expert or reporting analyst and being able to translate outcomes into impacts your leaders need to know
- To decipher your own and comparator results – be in a position to find benchmarks and understand whether differences in pay are significant
- To make short shrift of the reporting job itself and save time – prepare well such that this reporting mandate takes the least time for the best result
- To anticipate results and be ready to make the best of what follows – get ahead so that you are ready with communications when first results are produced
- And to reassure yourself that you know what it looks like you are going to have to do – have a speed read of the rules!
- Draft regulations have now been published requiring larger organisations to publish Gender Pay Gap information – read here what is required
- Whilst the regulations are to be applauded for a relative clarity and simplicity, inevitably there are particular issues worth care – look out for the pay period in question, of terms and conditions, reporting method and your own job roles
- HR professionals and accountable leaders are wise to work closely with analysts and payroll experts to be sure of confident analysis of potential results – issues for each to watch out for are noted here
- Get ahead with predictions as to results – be ready for your analysis to create opportunity for both internal and external communications
- Reporting can apparently suffice without comment for compliance and without consequence for non-compliance – read beyond the rules in this regard
- Consider some suggestions to get ahead and make short shrift of work and opportunity of outcome – find following recommended actions
- The complex question of gender pay difference is not readily boiled down to simple metrics – but it is a relevant issue that is hard to ignore with justification, wherever motivations lie
We knew that in the course of this government we would see compulsory reporting about the Gender Pay Gap in larger organisations. Draft Regulations have been long-awaited, with considerable uncertainty therefore about who, what, when and how questions – and the why left somewhat to our own imagination, interpretations or perhaps inevitabilities, given the low take-up of voluntary equal pay auditing. In February 2016, the Government Equalities Office published the Mandatory Gender Pay Gap Reporting regulations in draft and final consultation closed on 11 March 2016.
The draft regulations are short and relatively simple. The timescale is tight for legislation to be passed, which may indicate that significant change is unlikely when put before Parliament, and many questions have indeed been answered. Organisations in the private and voluntary sectors, employing more than 250 people are covered – but the government intends new rules soon to apply to the public sector too.
Since the Equal Pay Act 1970 (EPA), the question of comparing men and women’s pay has of course been real, but it is only now that compliance requires publication obligation in the absence of any apparent cause for concern (a claim). Whilst the pay gap in this country has narrowed, more recently it’s stubbornly stuck at a 2015 figure of 9.4% for the full time comparison and 19.2%, which is the gap more commonly debated, including those who work part time.
These are lowest ever values and the reasons for the gap that remains these days is unclear. What is clear, however, is that the government thinks voluntary is not working and there is more to be done.
What do the new rules say?
Assuming that legislation matches the government’s draft unaltered, here are the facts:
- Mandatory reporting applies to organisations in private and voluntary sectors with 250 or more employees working in Great Britain on 30 April 2017
- The government states an intention that public sector bodies will also need to comply, but exactly how and when is subject to more consultation
- We all have until April 2018 to publish information – but that information must relate to the pay period including 30 April 2017
- There are 5 headline figures which must be published, which are the mean and median pay gaps, bonus pay gap and proportion of men and women getting bonuses, and finally the numbers of men and women in each quartile pay band.
- We’ve got to put those figures on our websites and submit them to the government portal, but there is no requirement to add detail or comment
- Anticipate doing so each year, with 12 months each time to get your figures out there
Please read on for a bit more about what’s covered and what not and how to work it all out.
A Closer Look and Some Good Sense
I like the clarity of the draft and applaud its brevity. I am no mathematician but the combination of mean and median reporting solves one problem and without adding any extra work. The mean is the layman’s “average” but it gets terribly distorted if you have particularly high and low earners. The Office of National Statistics (ONS) prefers to use the median, which at school we called the “middle number” – so we’ve got to do both.
If you like maths, consider that the mode is well catered for because we have to say how many employees come in each band – this helps to identify any really striking issues with gender pay in pockets of an organisation.
It’s essential to understand that figures must relate to that pay period which is used in the organisation that includes the precise date of 30 April 2017. A pragmatism about reference to what is effectively a payroll run has to be a good thing, as does the attempt to avoid really tricky seasons – or indeed too vast an amount of accumulated data.
It is probably also a good move to separate out bonus pay – and this with a reasonable definition on pay elements covered and not. Bonuses have an enormous skew factor. Whether they are double-counted could prove to be a moot point, but I think right. Bonuses are set to be a contentious area ongoing, so to err in favour of more information rather than less could be wise.
Finally, let’s note the link made between the work of the ONS in the ASHE on gender pay in terms of method so that we have the best hope that meaningful use can be made of what we all do with the new rules. There is an obvious point here in that pay values must be taken as hourly rates, which of course allows the full and part timer comparisons. These should be rates “derived” from total pay and hours worked in the period.
Keep Reading and Ask Some Questions
So what is pay and what is not, say HR and Finance?
Luckily we are not bereft of guidance here. Your average pay rates have got to include basic pay and leave pay, with regular allowances and bonus pay. Overtime, expenses and deductions – including those adjusts made under salary sacrifice – are out. So far so good.
I note the closeness of these clauses to recent case law outcomes on holiday and overtime and the idea of what is paid as “normal”. It looks like we need to assess pay elements in terms of their regularity – or at least that this is the intent. So what are we going to do when we start to see overtime attracting holiday (because it’s becoming “normal”)? It’s out – and with it the associated holiday – but if it is ruled to be normal because of holiday pay law, will it become normal because of gender pay?
What of sick pay? The ONS like that their data is “unaffected by absence”. Gender pay gaps have got to include sick pay (and other leave pay), as paid in that pay period. I think this doesn’t create any extra difficulty for you, but it does potentially mean that results become affected by absence. Leave pay could well be lower than regular pay. It is perhaps a shame that in drafting about gender equality note a reference to maternity but not to paternity or parental leave too, which is generously attributed to a need for brevity.
And on oddities, I wonder if the pay period that includes 30 April 2017 will include a certain May Day Bank Holiday (weeklies?). If Bank Holidays raise for you an “uh-oh” and this applies, then please note. A wider point of good sense is to give the month of April a quick moment’s thought for any working or pay practices which are going to warrant attention with regard to what your men and women do.
I am a little bit confused by annual bonuses paid in April. If they are “pay for a different pay period” (and I think not) then they are excluded; if they are not then they are in and, if sizable, further increasing their exaggerated effect. I invite comment. What’s to make of it?
Speed read! It’s April 2017’s pay. Gross pay. Most regular payments that are made, but not overtime or benefits in kind.
Deriving the Answers – Over to the Analysts
Above we’ve looked at concepts as to what comprises pay, allowing you to work out how that applies within your own pay structures. There provides the objects for your reports. I trust it’s not too tricky to work out means, medians and quartiles – or to count “numbers of”.
Now comes the question of derived hourly rates.
Making the issue of gender pay real, we have to bear in mind that the fact that regulations are littered with the word “employee” is well justified, so keep reporting identified on a by-person basis. But I would be including pay element identifiers to the bitter end, such that data can afterwards be cut different ways for analysis. Someone ought to be making sense of the outcomes.
Combining data about salaried employees and those paid on an hourly basis is not in theory too much to ask, nor anything we are not used to dealing with. But what happens if individuals work on differing hourly rates within a pay period? Because the intention has to be to arrive at a by-person answer, rather than by-job or by-pay-rate, it follows that hourly rates need weighting and averaging. It follows that the number of hours paid does matter, so rule out anything too desperately simple. And “overtime” is out. Question variable, flexible and zero hours pay.
A second worry for systems experts when it comes to putting data out there often starts with that of data cleanse (or possibly “help!”?).
I doubt you have too much to linger on here, with an assumption that that payroll reconciliations for HMRC’s benefit are taking care of the accuracy of what is currently going on in payroll. If you do have reason to believe that your error rate is going to cause a big headache in this context, then I suggest there are more pressing issues to focus on sorting before a worry for organisational data with a 12 month delivery expectation.
You may wish to spare a thought to your audience however. Does the presentation of your pay data require some rationalisation to be a good job done? Is what is paid and why, buried within your reports, decipherable? If you will be merging data from different systems and payrolls, more to think on there. Note we are all clear that it is apparently single figures that matter, but I am hoping very much that those single figures stand up to digging deeper.
Speed read! Not too much for the analysts you are not used to, but please talk to those accountable for the answers about method – and see above for the “what’s in” question.
Getting to Grips with Outcomes
No Comment, No Consequence?
A key question answered by the publication in February is whether there would be the requirement to offer commentary behind figures – or indeed the extent to which detailed breakdowns were to be obligatory. That answer is no.
Please envisage – if drafts are unchallenged in this respect, and I’d imagine it will be popular – that it is just 5 simple metrics in question. No need to profile by job type, pay grade, location, age…..
…..For the purposes of the bald reporting that is. If you have no need to offer comment, then consider if you will wish to do so. What will be the consequences of publishing your gap data?
If it’s a success story, you’ll surely wish to promote. If not so, you will no doubt wish to make certain you have explanations ready (some clues below). I think that the safest presumption is that comment is called for. In the best case, comment is an opportunity to differentiate, to attract and to celebrate. Be ready to do so both internally and externally.
Comment with confidence becomes possible if:
- You understand and can defend how figures are arrived at, and:
- You have explored how those figures relate to pay period anomalies, pay policy and terms, your organisational activities, and:
- You can relate that local context to comparators you deem to be appropriate
Remarkable about the draft regulations, well-partnered with ONS approaches, is a relatively loose link to the EPA itself. This is a real problem with the single metric answer, without breaking down to detail. There is no attempt within new rules themselves to link those single figure outcomes to the concepts of “like work” or “equal value” claims, so this is going to be a step you need to make yourself in the analysis. It must remain true that a defensible position is one where job evaluation supports pay outcomes and pay difference is explained therefore by factors other than direct discrimination. But it is not a link that the regulations themselves are going to help you to make on their own.
Note also that the GEO clarify that no further civil sanctions are to be taken against non-compliance at this stage, beyond publishing “examples” and those compliance levels are to be monitored. Yet I trust no need to make the “why bother” case any more explicit than it is.
Speed read! No comment, no consequence is not realistic. If your pay gap looks set to be of significance, be ready to answer to that significance.
Is Our Gender Pay Gap Big or Little?
Accelerating to a report in front of us with those figures on it, how do we understand whether that is going to be a good answer or a bad one?
Here are some suggestions:
- A start point are the national figures. Note the headlines above. Understand the “part time factor” which explains that enormous difference between a figure of 9-10% and one of 19-20%. More women than men work part time. Part time jobs have in general lower hourly rates. These factors are so significant that, even though female part timers actually get paid more on average than male part timers, when part and full time employment is compared, the overall picture is that women earn less by the hour than men.
- Narrow to your own industry – and you will be in a better position than me to suggest how – and whether to wait until first data sets are published by the government. Note we don’t yet know how this will be done exactly. You’ll wish to find some closer-to-home benchmarks if your narrative is to be a convincing one, as we all know that men and women at work is extremely industry-sensitive (“market segregation”).
- Look to best practice recommendations for treating variances. The Equality and Human Rights Commission recommend that 5% variations are of significance, so this might be a safe value by which to judge parameters of acceptability. There is a little rule of thumb of 4/5 (or 80%) to answer to significant variations on questions that are yes/no values. This could work if you wish to look at pay elements on a side-by-side basis, which is rather lost in translation from Equal Pay law, where each term has to be applied equally, to the new reporting regulations.
- Add value to voluntary comment, whether internal or to publish, with analysis of the influence of market forces affecting pay locally or in the industry, peculiar pay elements unduly distorting, legacies of TUPE and red-circling or values to be found within the pay banding and quartile data pointing to predominance of one job type and gender choice. Discuss the relevant pay period and season, if that is appropriate.
How to Get Ahead
Here are some suggestions as to how to make sure that the words mandatory, reporting, regulation don’t put you and your colleagues off opportunity for a confident approach:
- Work out if you qualify for October 2016 reporting as mandatory and if you do, don’t be fooled by the April 2018 reference – pay data has to come from April 2017. If you are in the public sector, fore-warned is fore-armed. Remember that, if you are on the brink of 250, have a think if growth means that by April 2017 you could be there.
- Watch out for final regulations and how much proves to be varied. Look for guidance on interpretation of the questions above amongst for sure others. This is particularly key if you’ve complex questions around flexible hours working and/or are already getting stuck with what’s afoot on holiday pay and overtime debate – or if bonuses are a big feature.
- Don’t leave it until the deadline to anticipate your organisational results. Predicting report results positions you to respond with analysis and expectation. Work out what you think the figures are going to be. Aside from the substantive issues, is this a potential PR coup or PR employee relations disaster?
- Make sure of consensus between those accountable in the leadership teams, HR and payroll or analyst colleagues, of how the data is to be prepared in relation to roles, personnel and pay elements. Don’t leave it to the mercy of a trusted data analyst to decide on questions of organisational policy and the results that ensue by accident in systems report methodology.
- Document methods proposed and actual to enable you to explain, correct if an iterative approach becomes called for, analyse and cut the data in different ways, and to repeat next year based on like-for-like comparison.
- Check on – ahead of time – any likely “funnies”. Are there peculiarities of terms and conditions, job roles or timings? Remember it is the pay period including 30 April 2017 that matters.
- Don’t worry for the purposes of publication in so far as mandatory about breakdown by parameters such as location, job role, formal pay structure, age or other personal characteristic other than gender. Do worry about how you might want to break down data for the purposes of your own internal analysis. If it doesn’t go in the reports, it can’t come out.
- Take the no comment, no consequence drafting with a pinch of salt and consider the opportunity and requisite for both internal and external narrative comment. Reflecting on reasons that are widely suggested for gender pay differences is of universal value, but do look to what is organisationally contextual and different too. There will be structural, social and statistical clues.
I am glad that these regulations, albeit draft as yet, have finally arrived with us. I welcome a relative clarity and simplicity. I don’t envisage the reporting gulf that there could have been and I do anticipate the opportunity for advancing understandings within our organisations about gender pay difference, without too much pain.
That said, to fail to anticipate ahead could be to miss a trick and to take an undue risk in the results, when a little thinking solves that problem. It’s not too arduous – and it could just prove interesting – so why not?
You will have your own view of the Gender Pay Gap and therefore the right and best things to do with results could prove to be. Are these just numbers? Are these choices? Does unfairness persist? Can employers fix the problem to any significant extent at all in the absence of new solutions to how on earth to juggle work and family and to make that pay? Is the stubborn percentage shifting little because, as things are, there is no further to go?
Whatever your personal view, I encourage you to take an interest in what is happening here in the GEO regulatory answers. With the following I agree for sure “There is no single measure that adequately deals with the complex issue of differences between men and women’s pay.”
 The Equality Act 2010 (Gender Pay Gap Information) Regulations 2016 – with report for consultation on this draft available from the GEO and published on 12 February 2016
 Figures from the Office of National Statistics, reporting in November 2015 on 2014 data
 Don’t worry about quartiles and pay bands being difficult. The regulations use “band” to mean the number of employees falling in the each 25% (quartile) of pay rates, rather than any term related to pay structures, or indeed particularly clever sums.
 Bonuses have to be included in the average pay figures for mean and median, but also reported as a separate figure, with the proportion of those affected
 Annual Survey of Hours and Earnings – an annual report published by the ONS (really interesting to read!) and including gender pay figures, thus offering comparable methods to use and trend values
 These are my summary words – read the draft regulations for the exact expression, but don’t expect more than one or two more lines
 The word “normal” isn’t in the regulations – but I spot what seems to be a similar intention and a felt-fair approach in the new rules
 41% of women workers are part time but only 11% of men (Labour Force Survey data, 2015)
 ONS in the ASHE (see above), November 2015 preliminary report