David Kernohan and Louis Coiffait take a look at the higher education sector data behind the gender pay gap headlines.
This article is more than 8 years old

It’s a depressing reality that, on average, women are paid less than men. Yet it is still real issue, and higher education is far from immune. Why that remains the case – and what best to do about – are tricky questions.
There are a number of issues conflated within the headline figures: the straight-up legally actionable (since 1970) inequality of different rates for the same job; the differences in career trajectories that exist between men and women and how these are compensated for in pay and promotion; the lingering suspicion that “woman’s work” (often a great deal of gruelling emotional and organisational labour) is not appropriately valued; and the tendency for women to be over-represented in part-time and lower-paid roles while men are more commonly found in better remunerated jobs.
Understanding the scope and nature of the issue is not going to solve it overnight, but it is a valuable first step. Sunlight, as the saying goes, is a great disinfectant – you can be sure that organisations shown to be underpaying women will be looking to rectify the issue in the face of significant popular disdain.
Since the 5th of April 2017 it has been a statutory requirement for all employers with 250 or more staff to publish data about six different pay calculations every Spring:
Public sector bodies have a slightly earlier deadline and must comply with the public sector equality duty, but essentially comply in the same way as charities and businesses.
That information must then be published on the employer’s website and by the government. The data collection and reporting is overseen by the Government Equalities Office (which is part of the Home Office), where you can search for individual employers or download the latest data in one go. According to the Home Office it takes up to an hour for an employer’s online submission to be published, making the data almost live. However, as this is new, self-reported data, it’s currently very “dirty”, requiring significant cleaning-up before analysis.
Employers can also choose to publish an accompanying narrative, including challenges they face, successes they’ve had, and planned actions to close gaps.
Failure to report the data is an “unlawful act’ and the Equality and Human Rights Commission (EHRC) can take enforcement action, with penalties including an unlimited fine (see s34 of the Equality Act 2006). They may start an investigation if they suspect an employer is hiding a pay gap. Over 10,000 employers submitted data, with about a tenth doing so on the last day. The Commission said that 1,557 employers missed the deadline and that they will start by writing to each of them for an explanation within 28 days. Terms of reference will then be issued (publicly) for any enforcement processes. The reputational risk to employers that don’t comply, or those that have large gaps, is high.
It’s worth noting that HESA already collects a wealth of information about higher education sector staff, including pay, but it is not currently open data, and is not offered at an institutional level without payment.
As above, this is very early days, and the data is pretty dirty. In particular, coding to SIC (Standard Industrial Classifications) codes is very unreliable. Despite the existence of specific codes (85421 and 85422) for undergraduate and postgraduate tuition, these are not reliably used. Some institutions do not use any code at all – others a miscellaneous variety of labels seemingly applied at random.
To generate our list of higher education providers we had to rely on human judgement – using searches to the two identified codes, alongside general searches on the terms “university” and “college”, plus cross checking with existing lists. As it stands, we’d be surprised if we captured everything (we’re notably low on institutions based in Wales and Scotland) but we did get most of it, using the data as it stood on the afternoon of April 6th, 2018.
One interesting artefact of this manual process was that we could pick up related organisations (everything from Students’ Unions, to traded services, to Oxbridge colleges). We’ve omitted these from our main visualisations, but they are there and can be added back (or viewed seperately) if you choose via the “groups” drop down. The size of employers is a hugely important factor in pay differential, so you can filter by size as well.
There are six visualisations, accessible via the tabs at the top as follows:
As always with our Tableau visualisations, you’d be best enabling full screen mode (the icon at the bottom right).
Across the whole sector, men are over-represented in the top quartile, while women are significantly over-represented in the bottom quartile. This pattern holds for all groupings apart from the students’ unions we have data for.
The Russell Group and other pre-92 institutions are likely to have a greater mean difference in hourly pay than other groupings – probably due in part to them being larger employers. However, there are no clear grouping predictors for the median difference – institutional size and structure may play more of a role here.
Only four institutions have more men than women in the lower quartile of pay, whereas 30 have more women than men in the top quartile of pay. It is heartening that some women are beginning to reach higher levels of pay (though there is still a way to go), but we should be seriously concerned about the numbers of women amongst the lower paid of the sector.
According to our analysis the median HE-sector pay gap is 16.32%, compared to a gap of 9.7% across all employers. Whatever way you cut it, that is not a good look.
Official guidance about what employers can do about gender pay gaps is provided by the Advisory, Conciliation and Arbitration Service (ACAS). In HE, Advance HE, unions, and employer associations also have a wealth of guidance available, with some details of the latter provided in UCEA’s article on Wonkhe.
Some of the solutions include using training and institutional processes to tackle all forms of bias at different career stages, promoting flexible working practices, as well as tackling any pregnancy or maternity pay discrimination. Shared parental leave is another area that could lead to improvements.
The top-level pay gap data, and much of the accompanying commentary, has been relatively blunt. However, this new reporting requirement has clearly helped to raise the profile of this issue and should encourage all employers to understand and tackle any pay gaps. There is an opportunity for individual HEIs, as well as the sector as a whole, to demonstrate leadership, not only with their own employment practices – but in how they conduct teaching and research. Educational institutions have a unique capability to address gender issues, including pay. Those institutions named and shamed in the press for large gaps have signalled that they recognise the issue, the question now is whether they can show progress by the time the data comes out next year.

David Kernohan | Data | 9/11/26

David Kernohan | Data | 6/11/26

David Kernohan | Data | 16/09/26

David Kernohan | Data | 18/08/26

David Kernohan | Data | 13/08/26

David Kernohan | Data | 9/07/26