Of the 50 organisations reporting the largest gender pay gaps in March 2018, almost half were multi-academy trusts. A startling statistic.
Since organisations were required to publish information on their gender pay gap in 2018, the discussion around gender inequality in the workplace has ramped up a level. It seems fairly astonishing that this is still an issue, and as a data-driven consultancy with equality at the heart of all we do, we felt we should address this topic and see how data can help us figure out what’s going on to cause such a disparity.
The conversation about fair pay is nothing new. From the Dagenham protests way back in 1968, right through to the ramifications of the gig economy, it’s a debate that has raged through the ages. There is still a lot of work to do, but it can only be a positive thing that data is in the public domain and can help uncover what is really going on.
In this blog post we start to unpick this topic. We’ve got definitions to clarify, data to review and explanations of what would need to be done to reduce the gap. So read on and join us on this gender pay gap journey.
So, what exactly is the gender pay gap?
First, let’s be clear about what we mean when we talk about the gender pay gap. This is not about equal pay. Whilst both equal pay and the gender pay gap deal with the difference in pay women receive in the workplace, they are actually two different issues, as defined by the Equality and Human Rights Commission:
- Equal pay means that men and women in the same employment performing equal work must receive equal pay. That’s already a legal obligation as set out in the Equality Act 2010
- The gender pay gap is a measure of the difference between men and women’s median earnings across an organisation or the labour market. It is expressed as a percentage of men’s earnings
So, in a nutshell, this article is about understanding the gap between the median pay men receive compared to the median pay women receive. In Britain, there is a gender pay gap of 18.4% across all professions, meaning the median man earns 18.4% more than the median woman
A lot of well-meaning commentary has been written about the gender pay gap over recent months and from time to time the two issues (gender pay gap and equal pay) have been confused. The maths is often misconstrued and solutions proposed are therefore sometimes invalid.
Demystifying the median
Our friends at Wikipedia tell us that the statistical definition of a median is:
- The value separating the higher half from the lower half of a data population. It may be thought of as the ‘middle’ value
As an example, let’s say there were nine people in the “population” of your organisation. If you lined up these nine people in order of salary, the median is the salary of the person in the middle (i.e. the 5th person). In the example below, where salaries are evenly distributed through the organisation, person 5 earns £40,000. Hence, the median salary, or the middle value, in this organisation is £40,000.
Now let’s look at a second organisation. It has five women and seven men. Here’s how their salaries break down:
The median woman’s salary is £30,000 while the median man’s salary is £35,000. This is a gender pay gap of 14.2% – i.e. the difference in median salaries (£5,000) divided by the median man’s salary (£35,000).
Now let’s overlay their job titles:
Man 1 has the same salary as woman 1 for performing the same cleaning role. Man 2 has the same salary as woman 2 for the same payroll role, and so on up to man and woman 5. However, the two highest paid members of staff are male directors with no female counterpart. So the gender pay gap in this organisation is wholly down to the fact that the two senior members of staff are men. There are no issues with equal pay since two people doing the same job are paid the same, although there may have been issues with equality of access to senior positions.
Now imagine that the organisation has a very good year, and the selfish CEO triples his own salary, but keeps everyone else’s salary the same. What happens to the gender pay gap? Nothing – because the middle man and woman are the same and their salaries haven’t changed.
However, what happens when the CEO takes early retirement and is replaced by a woman? Let’s see:
Since there are six women and six men there is no longer a middle man or woman. Instead, we need to average the salaries of the two middle women (woman 3 and woman 4) and two middle men (man 3 and man 4). So, the median woman’s salary is £32,500, as it is for the median man. The gender pay gap has therefore been removed by changing the gender of the most senior member of staff. Hurrah! The same would have been true if the CEO had remained the same but the sales director left and was replaced by a woman.
Now, what happens when the organisation buys a cleaning robot and decides they no longer need two cleaners? Woman 1 decides to take voluntary redundancy but man 1 stays. Now the women’s pay range looks like this:
Woman 4 is now the “middle” woman and so the median woman’s pay is £35,000 while the median man’s pay remains at £32,500. The gender pay gap is now negative (-7.7% to be precise) because the woman’s median is higher.
Now, the organisation realises they made a terrible mistake and that the robot cleaner is useless. They are making more mess than ever and need to take on three new cleaning staff, and they happen to all be women. The median salary for women is now £27,500 and the pay gap is now 15.3% in men’s favour:
The point is that an organisation can have equal pay for the same roles but a gender pay gap because of the different roles that men and women take up in the organisation. And, increasing the pay of a fat cat or two at the top of the organisation, as inappropriate as it may be, makes no difference to the median gender pay gap. However, changing the roles that are carried out by men and women does.
The gender pay gap in education
Right, back to the matter at hand. We can see the overall 18.4% pay gap is vast and in favour of men. Armed with this knowledge, we wondered what this gap looks like within the education sector, Mime’s primary area of focus.
We first took a look at statistics on the gender pay gap published by educational establishments. Academy trusts and schools with more than 250 employees were required to publish their gender pay gap data by April 2018. And when the data was first published, education establishments did not come out well. The education sector had one of the worst median gender pay gaps in the UK at 19.7%, and half of the worst 50 organisations were academies or multi-academy trusts.
Before we move on in it’s worth reflecting on the fact that the reporting was for organisations with over 250 staff. This leaves out smaller settings, especially nurseries and primary schools. So, the main educational establishments that published gender pay gap data are likely to be Multi-Academy Trusts (MATs) who represent a number of schools and will have a range of central roles outside of their schools.
What other data can help?
All state-funded schools, including academies, are required to submit information to the DfE on their workforce. We wanted to see if we could use this data to provide a crude indication of the types of roles men and women have in schools, and from there establish how this affects the gender pay gap.
The data tells us the number of people in England in different roles, split by gender and by school phase. From here we established ten different job roles in England’s schools and researched the typical hourly salary for each role. We recognise this is a simplification of the range of roles and salaries in schools, but it’s what the publicly available data permits.
To set the scene, there are 498,000 teachers in schools in England. Of these 76% are women and 24% are men. On top of this, there are 824,000 people working in teaching assistant and other support roles.
Ninety three percent of teaching assistants are women (that’s over 350k), 90% of admin staff are women, 87% of auxiliary staff (school maintenance/reception/ catering etc.) are women and so on. You can see where we are going with this; the sector is heavily weighted towards women, however, the weighting drops the higher up the pay scale you go. For example, only 66% of secondary school teachers are women compared with 87% at primary school/nursery teacher level. However, when looking at heads, deputies and assistants the female take up is a lot lower – 69% of these leadership roles are occupied by women.
The chart below shows the full list of school roles available in the data, their average hourly salary and the proportion of the role filled by women or men.
We can see that as we move up the pay scale, generally there is a higher proportion of men in each role. However, the data shows that there are still twice as many female heads in schools than male heads. The chart below shows the actual numbers of women and men in each role, rather the than the percentages.
So now back to the issue of the median role and salary. For the purposes of this we are going to slightly simplify the data and group the roles into four main groups – heads, deputies and assistants, teachers and support staff – and take the average salary in each.
Each person in the chart below represents one percent of the people working in our schools. We have highlighted in orange the “middle” person – i.e. the median value or 50th percentile. This median “person” is in a support role. In fact, support roles make up a total of 63% of the workforce.
The people “pyramid” chart below shows how we calculate the median pay for women and men separately. Because women make up the majority of the school workforce, the women’s section of the pyramid looks similar to the overall picture above; 66% of women are in support roles and the median woman is in a support role. However, the men’s part of the pyramid looks different. Only 43% of the male workforce are in support roles and the median man is a teacher.
*Note that the average wages displayed here are not the wages we use to calculate the gender pay gap as the median man is a lower-than-average paid teacher whilst the median woman is in a higher-than-average paid support role. To calculate the median, we took the position in the distribution of support roles or teachers our median woman and man were in and layered this on top of the distribution of wages. More details can be found in the footnotes .
Teachers on average are paid a lot more per hour than support staff, and hence there is a significant gender pay gap in our simplified model. The average pay for our median male teacher is £13.41 per hour whereas for our median female support role it is £9.63. So, based on these figures, the gender pay gap would be 28% in favour of men.
So where does this pay gap come from?
The gender pay gap is based on the median role and is therefore concerned with the distribution of roles within the female cohort and the male cohort. Simply put, women are far more likely to occupy lower paying support roles than men. The data shows that the female median is weighed down by the sheer volume of women in the teaching assistant, admin and other support roles. The extent to which women are represented within these roles appears to be the main driving force behind the gender pay gap, not some underlying unspoken factor.
There you have it. The rate of pay for each role, and disproportionate gender weighting within the lower paid roles, unwittingly causes the gender pay gap.
Removing the pay gap
So, what would need to happen to our people pyramid for the gender pay gap to be removed or, in other words, for men and women to have the same median role? This is the bit that is often misreported.
Far less than half of women are in teaching roles or leadership positions. Because of this, moving women up from a teaching role to a leadership position makes no difference to the median salary. We could promote female teachers all we like, but the median role will not change – it would still be a support role.
Or, if we moved all male heads and deputies/assistants back into teaching roles, what would happen to the median role for men? Again nothing, the median man would remain a teacher.
Instead, we would need to shift women from the lower paid support roles into teaching roles to move the women’s median role to match the man’s median role. In practice it wouldn’t be possible to create the new roles and not fill the gaps that are left but, for the purposes of this exercise, let’s just pretend you could.
One way would be to take roughly 70% of all female teaching assistants and make them into teachers (that’s about 257,000 teachers, so it may take some time…). This would look as follows:
Or alternatively, what would need to happen for the male median role to drop down? Well, if just under 50% of male teachers (equating to 49,400 men) became teaching assistants, the median man’s role based on this data would become a support role, and the gender pay gap would be eliminated. This is shown below:
What this means is that the education establishments reporting extremely high gender pay gaps are likely to be those where a large proportion of their posts are support roles and where the vast majority of people in these roles are female. It’s less likely to be, as is often reported, because male teachers or middle managers are promoted to senior leadership roles more frequently than women, even where that is happening.
What is causing the imbalance of men and women in lower paid roles?
At this stage, it would be easy to say our work is done, unplug our laptops and pop out for a cuppa. However, the information at hand poses more questions than it answers. Just why are women not leaping forward and breaking those glass ceilings down? Are they being shut out from the more high profile opportunities or just shying away from progression to focus on their own childrearing? Is it that men are ‘taking our jobs’ or that women in education are more comfortable within less pressured roles? Or is it that working within a school setting as a teaching assistant supports a more desirable work-life-balance, especially amongst those with young children? So many questions, that in all honesty, the data can’t help us to answer.
However, others have dug into the issue and found a variety of reasons at play:
- Stereotyping – Unconscious stereotyping might be negatively impacting women. There could be assumptions that women don’t want to be promoted or take on more responsibilities; this could be more acute where women have additional caring responsibilities
- The Motherhood Penalty – Sociologists developed this term to describe working mothers who face disadvantages within the workplace due to the notion that their competence has been reduced relative to childless co-workers
- Part-time work – This also overlaps neatly with the Motherhood penalty. Lower paying jobs may be more likely to offer part-time working where overtime is not an expectation, and therefore appeal to the primary child carer. Recent research has suggested the lack of part-time opportunities may be a particular problem in the retention of classroom teachers
- Reputation – Maybe women with young children are actually put off of committing to higher roles within education, particularly in the current climate: reputation, hours worked, stagnating pay and intense scrutiny from Ofsted. These may not be perceived as desirable attributes to encourage working mothers (in particular) back into the middle or higher echelons of the education workforce
Whatever the reason, or combination of reasons, it’s worth remembering that if we are wholly focussed on the median gender pay gap as a measure of gender inequality, we are looking at a narrow measure that would be dealt with by encouraging more female teaching assistants into teaching roles, or more men into support positions. But since this may be the only option for women in some cases, it is taking us away from a potentially more pertinent issue which doesn’t affect the gender pay gap; specifically, why female classroom teachers are far less likely than male ones to progress into senior leadership positions.
A final word
Our discussion here is based on imperfect data but hopefully helps to shed a light on why the gender pay gap, a statistical measure, is not always the best way of thinking of gender inequality in the workplace.
Contributors: Diane Bartholomew, Steve Preston & Tim Bailey
Enjoyed this article? Don’t forget to share using the links below
If you’d like to discuss this article further, please get in touch with us through our contact form.
Note that this article was updated on 20/03/19
 Note that strictly speaking the pay gap should be based on pay per hour, but for the purposes of this example, we are using annual salaries.
 This would affect the mean gender pay gap, but that’s not the focus of this article.
 We aren’t completely sure why the data suggests that secondary school teachers earn more per hour than primary on average – is this to do with the positions of extra responsibility in secondary schools (which are normally larger than primary) that are below deputy/assistant head level? However, this doesn’t actually affect the calculation of the median salary.
 Our median male teacher was 7/47th of the way up as can be seen in figure 10, which is roughly the 15th percentile of male teachers. Looking at the distribution of wages from the workforce data, the 15th percentile falls within the lower end of the income bracket £25,000-£29,999. We converted this to an hourly wage (using a 37 hour week) which gave our median man a pay of £13.41 per hour which is below the £18.8 average wage for all teachers. A similar process was taken with the wage data for our median woman who is located 50/66th of the way up the scale for support roles. We took teaching assistant salaries as a proxy for all support staff as they are paid similarly and found that 50/66th of the distribution in wages is £9.63 per hour which is slightly higher than the £8.41 average for all support roles.