Blog purpose

Women entering Oxford have a lower chance of leaving with a first than men do. This blog is an attempt to set this fact in its national and historical context. I’ll then reflect on the institutional response to the gap.

My experience at Oxford has been that I feel little urgency around solving the gender gap: it is a point of discussion rather than a point of action. People really care about it, but nothing much seems to be happening. For the last few years, I’ve been trying to understand why. It’s a real puzzle to me! This blog is my attempt to think about how institution-level targets relate to specific policies and actions.

The language of gender gaps

Men and women sometimes perform differently in assessments. For example, women typically out-perform men in GCSE qualifications. This has been called a gender attainment gap, describing the differential rate of - say - attaining A* grades.

There is growing criticism of the term ‘attainment’, because it centres the issue on the individual. Most people involved in gender gap work now call the phenomenon an ‘awarding’ gap, which emphasises the role of the institution and environment. (Oxford’s documentation mainly uses the term ‘attainment gap’, likely to match the language of the regulator.)

The merit of using ‘awarding’ language is that it drives conversations into a constructive space where people in positions of power ask ‘what could we do?’ The analogy of a pot-plant can be useful here. If a plant isn’t growing in the shade, perhaps you could move it into the sun. But if a plant is just bad at growing, well, that’s unfortunate but it’s not your problem.

Public data: OfS, Oxford

The Office for Students publishes national data on awarding gaps by sex. Over the last ten-or-so years in England, women typically get ‘good’ degrees (firsts or two:ones) at a rate six percentage points higher than men. This margin has stayed fairly constant even as the overall rate of ‘good’ degrees has increased. The gap narrowed to four percentage points during the years of pandemic assessment (2020-21 rates: women 84%, men 80%).

In Oxford, the national picture seems to be inverted, though I have struggled to find directly-comparable data. The 2015-18 average rate of getting a first (rather than the broader OfS measure of a ‘good’ degree) was reported to be over 10 percentage points higher for men than women. The number varied between subjects, with Physics (17.7 percentage points) being singled out in student reporting as having a particularly large gap.

Is a bias towards men getting firsts just a feature of any Russell Group University? In 2015-16 the average Russell Group University awarded firsts to men 1.3 percentage points more frequently than women. The nearest comparator, Cambridge, had a 2015-16 gap of 2 percentage points. Oxford’s numbers seem to be much larger than the typical Russell Group profile.

Like the rest of the sector, the pandemic seems to have been associated with a narrowing of the gap. In my own Division of the University, MPLS, the 2019 gap between women (32%) and men (47%) was 15 percentage points; by 2021 it was 9 percentage points (42% of women and 51% of men getting first-class degrees). In the wider University, the same years saw the gap narrow from 9 percentage points (31% vs 42%) to 6 percentage points (39% vs 45%).

The Oxford gender gap phenomenon is not recent. A 2021 paper by Ahlburg and McCall establishes that the gender gap in the award of firsts is over a century old. Interestingly, the narrowing gap widened once men-only colleges started admitting women in 1974, and has persisted despite recent political pressures such as the 2010 Equality Act.

Conclusions on the public data

Oxford favours men over women in the way it awards firsts, and has done so for at least a century. An unsympathetic reader might judge this pattern of gendered degree outcomes to be an Oxford tradition, especially as ‘good’ degrees are typically awarded more frequently to women nationally.

The recent narrowing of the gap is a good thing, but the current numbers still compare unfavourably to both the sector at large and close comparators like Cambridge. It is plausible that the recent narrowing of the gap describes the pandemic’s effect on the curriculum and assessment, rather than long-lasting changes. This will likely become clearer in the next few years.

Urgency and action

The first thing anyone should say about this data is that the numbers are big and the numbers are bad. The massive role which Oxford graduates play in public life means that the simple fact of this awarding gap has - in its own right - significant implications for Britain. It would be better if degree outcomes were more equitable.

What concrete actions would address the gender gap? Let’s see what answer Oxford’s published policy documents give to this question.

Oxford’s current strategy

In the 2018-24 strategic plan, the third priority is to “set ambitious targets by April 2019 to reduce by 2024 gaps in attainment by gender, ethnic origin and socio-economic background.” 

It was surprisingly hard to discover what this ambitious target turned out to be. The EDI unit of the University has the goal “to reduce the first-class degree attainment gap between women and men from 8.5% to 4.4% by 2025“, which is the closest match to the 2024 deadline I was able to find; perhaps the timelines shifted because of the pandemic.

I was not able to find any local targets (e.g. for Divisions, Departments, or Colleges). It might be that these exist but are not public, though I am not personally aware of any in my College or Department roles.

Targets and lead times

A target alone doesn’t articulate what policies and actions will be used to get there (for that pot plant to grow another inch you need to move it into the sunshine, rather than just have the ambition to move it). The actions for this problem are distinctive because of the time frames involved in HE; for something like degree outcomes, the lead time for an action to come into effect can be very long. 

For example, students are admitted 3-4 years before graduating so actions around admissions should already have happened if the impact is to land in 2025. Students are entitled to study the curriculum they applied for, so really actions on the curriculum should have happened even before students apply (probably including students who defer entry for a year).

Representation of women among academic staff is peculiarly difficult to change rapidly at Oxford, as staff turnover is so low. It would likely take decades to substantially rebalance the 23% female representation among Academic & Research staff in the MPLS Division, for example.

Quicker timeframes exist for actions like routine improvement (that coursework guidance isn’t quite working - could I restructure it?) or systematic staff development (e.g. formal, scholarly training in how to write an exam question).

Where are the strategies? What are they?

Overall, I found it hard to discover what the strategies were and what specific actions have already been taken.

The 2020-21 Equality Report states “The University is implementing a programme of enhancements to promote flexible and inclusive teaching and is making progress towards its commitments to eliminate awarding gaps under the Access and Participation and Strategic Plans.“ But it has been hard to discover what this programme of enhancements consists of, or how extensive its adoption has been.

The Strategic Plan is clear on the ambition to reduce the gender awarding gap, but does not provide detail on how to accomplish this (which seems reasonable for high-level document).

The Access and Participation Plan has no outcomes targets relating to gender, though it does for Black students (Target 4) and disabled students (Target 5).

The Access and Participation Plan details how the University proposes to address these non-gender gaps, which gives a flavour of what a gender target might look like. It describes objects like the resources produced by the Central Teaching and Learning team (such as a toolkit to support racially inclusive teaching) and coordinating College-Department support for disabled students, as well as packages of financial support. These actions are not directly aligned with solving the gender awarding gap, but may address it incidentally.

Reflections

Overall it is difficult for me to discern from the public documentation what specific actions are being taken and who is accountable for narrowing the gender gap at the proposed rate. It’s very hard for me to work out whether this is because I am ignorant of work being done behind the scenes, or instead whether there just isn’t much happening.

Ignorance is a very plausible explanation! I am not involved with examining in my role, for example, and it might be that mandatory training has been developed for examiners on how to set and mark questions in ways which allow women to succeed equitably. There may be working groups and communities of practice which I haven’t discovered because of the way my job started in the same year as the pandemic (if you’re out there, please get in touch!).

I also have an imperfect view of the position in other subjects. Chemistry is on the point of changing its curriculum (which means changes won’t happen before 2025 in this subject), but other disciplines may have made changes already. The 4.4% target may reflect the gradual turnover of curriculum reviews across the different disciplines within the University. If there are strong principles of equitable course design threaded through the review process, it might be that the institutional gap wanes as more subjects establish inclusive curriculums.

Concerns

The lack of visible local targets concerns me. It would mean very little to a woman in Physics that the whole University meets its target when her experience is ‘balanced out’ by someone in Sociology. This is a risk with any institution-level target.

It’s also not clear to me that the University has a diagnosis of why the gender gap exists. Is it because of admissions? Teaching? Assessment? Some specific mixture of the three? Articulating this diagnosis would likely stimulate and coordinate action. It might also serve to clarify the accountability between the colleges and departments, which seems like a barrier to any kind of change within the University (“we’d obviously love the gender gap to be smaller in our subject, but really it’s a college issue around admissions”).

Paired with this diagnosis issue is a paralysis about knowing what actions have a reasonable chance of success. If someone sees the problem and concludes that - say - it’s a problem to do with the format of exam questions, what do they do? What specific thing should they try? I am not aware of a scholarly ‘menu’ of actions, and Oxford’s lack of systematic capacity in disciplinary SoTL expertise (T&S career pathways etc.) makes constructing one difficult.

Suggested actions

So I feel I’m back to the puzzle which started me writing this blog. I think staff really want to end the gender gap, but it’s also clear that the gender gap endures. There is an appetite for change, but somehow this appetite remains unsatisfied. What’s the hold-up? I think the hold-up is the specific step from discussion to action.

What does a useful suggestion look like in this context? I have three proposals:

1. Clear, local targets - for example at the Department level - would be useful to build momentum and accountability within existing collegial structures. “By 2025, Chemistry aims to have an awarding gap of 4.4%” is the kind of statement which means something to colleagues and external examiners.

2. Specific communities of practice - perhaps at the Division level - would be useful to build coherent strategies sympathetic to the structures of each discipline. To hear what Maths is doing about the awarding gap is probably a useful thing for a Physicist, and to have someone to talk to about plans and progress would be a step forward.

3. Central support - perhaps in the form of an annual conference - to talk about patterns in the data and case studies from different areas would help build wider networks as well as share practice between interested educators. To see - from time to time - what History is doing is also valuable for a Physicist.

Again, these things may be happening already. If they are, I’d sincerely love to hear more.

Conclusion

Change takes time, but the current structures seem to be betraying ~hundreds of women each year. Whether it’s through our admissions or our teaching or our assessments, something about Oxford is going really wrong here: our education significantly discriminates on the basis of gender. This is bad! The scale of the gender gap needs to be addressed with an urgency that has been absent for a century.

Hitting the 2030 target of a zero awarding gap probably means that the right structures need to be put in place in the next couple of years. That means that if we seize it, this could be a moment of really positive change.