Sex and the REF

Does research assessment discriminate against female academics? asks Barbara Graziosi

March 20, 2014

Source: Getty/Alamy montage

Heads of department often see the worst side of their colleagues. Perhaps, I considered, women prefer to devote their energy to research and teaching

In the Iliad, at the height of battle, Zeus needs to thunder three times to get any attention. He may be the most powerful god, the ultimate patriarch on Mount Olympus, but nobody listens to him until he thunders three times in a row.

The same is true of patriarchal order in academia. Most of the time, I do not even notice it. But last year, Durham University made a serious and impressive effort to tackle gender inequality, and that meant confronting some unpleasant facts. First came a salary review: professorial women were rounded up in a meeting room, offered unappealing sandwiches and told that we earn less than our male colleagues. We were then encouraged to apply for better salaries, asked for suggestions on how to change institutional culture and offered a training course to strengthen confidence and networks. I made a few distracted comments but quickly went back to my own research (on Zeus, as it happens).

The second thunderbolt illuminated the problem of leadership. A governance review made it plain that there are no female pro vice-chancellors at Durham and (at the time) only one female head of department. Again, the issue did not capture my attention because I could imagine the reasons. Administration can be unpleasant. Heads of department often see the worst side of their colleagues – their pettiness, rivalry, inflexibility and, at times, sheer laziness. Huge pressure on resources and unwieldy administration also create problems. Perhaps, I considered, women simply prefer to devote their energy to research and teaching.

What finally shook me from my torpor was a detail about the research excellence framework: it is always the third sign that proves a pattern. While the data have not yet been published, it seems that the proportion of women submitted to the REF was lower than the proportion of female academics working at Durham. This should not have surprised me: similar statistics are true, at a national level, of the research assessment exercise carried out in 2008. And yet I was shocked. Of course, the lower return of women can be explained in part by reference to the other two problems. That academic men have more money and more power may influence decisions about research quality. But I am not sure that these inequalities quite explain the low submission of women in the RAE and now, perhaps, the REF. Is there something that favours men in the way assessment exercises are set up? Or are women simply less good at research?

There will be official investigations. As part of the guidance on submissions for the REF, institutions had to carry out equality impact assessments on their staff selection processes and identify any differential impact on specific groups of staff. These had to be submitted to the Higher Education Funding Council for England at the end of February and should eventually be published. I very much hope that these documents will be useful, but there are concerns about them. After the RAE in 2008, the Equality Challenge Unit examined 32 institutions and their equality impact assessments. Only 22 provided evidence of having undertaken an equality impact assessment; moreover, “save for a few notable exceptions, the quality of the EIAs provided was poor”. The Equality Challenge Unit also reported that interviews with unit of assessment leaders and pro vice-chancellors revealed “a general lack of understanding and engagement” with the equality impact assessment process.

In 2009, Hefce published a report, Selection of Staff for Inclusion in RAE 2008, which showed that there were more men than women eligible to be submitted to the 2008 RAE, and that they were also more likely to be selected. The chance of a permanent female academic being selected for RAE 2008 was 48 per cent, against about 46 per cent in RAE 2001. For men, the selection rate was 67 per cent, and the gender gap persisted even after other factors were taken into account. The report suggested that this could “be linked to selection bias resulting from age and gender”, but that “it could equally be a result of deeply rooted inequalities in the research careers of men and women”. The Equality Challenge Unit was more specific in its own report, citing “gender occupational segregation, both horizontal in terms of disciplines and vertical in terms of academic grades and research experience; work-life balance issues; a tendency for women to have greater teaching, pastoral care and administrative working loads compared with their male colleagues; and lower application rates for research funding” as possible factors affecting the low submission of women to the RAE. The report also made several specific recommendations, which were taken on board by Hefce in the lead-up to the REF.

While waiting for official reports on this round of assessment, I decided to ask my Facebook friends about women and the REF. Within hours I had a stream of useful comment. Maternity came up straight away: Hefce recognises the impact of parental leaves, but not the enduring impact of parenthood (and we can assume that this failure affects women more acutely than men). Some friends insisted that it is more difficult for mothers of young children to produce four substantial research outputs. I doubted this, but apparently one of the issues here is social politeness. Textbook contributions, book reviews and articles promised for conference proceedings do not necessarily make good REF outputs, but still need to be delivered, unless we are prepared to say “no”.

Statues (20 March 2014)

Excellent research involves both originality and the ability to harness previous work, but gender may affect the relative emphasis placed on each

In the sciences, multi-authored papers can be a problem: when researchers work in the same institution, they cannot all claim the output. A second or third author of a paper will not usually be able to submit it. (Again this is likely to affect women more acutely than men because there are proportionally more junior than senior women, and hence more second and third authors, and because women tend to encounter greater family obstacles if they want to move institutions or collaborate closely with long-distance colleagues.) Style came up, too. REF reviewers read many publications in a short time and must surely appreciate a clear statement of what a specific output contributes “in terms of originality, significance and rigour” (as the guidelines state). Opening paragraphs in which the author claims to profit from the insights of others may not be so effective. Of course, excellent research involves both originality and the ability to harness previous work, but gender may affect the relative emphasis placed on each – at least in terms of rhetoric.

For me, however, the real headache is how research impact is evaluated and how it affects women in the REF. Hefce’s intention was for the number of academics submitted to determine the number of impact case studies required. In fact, what happened was often the reverse: academics working in the same unit and with outputs of similar standards could not all be included because there were not enough impact case studies to support a full submission. In such scenarios (which are common), criteria other than excellent research outputs must apply. One way to discriminate when predicted research ratings are similar is to choose researchers who feature prominently in supporting documents, such as the “environment template”, where indicators of esteem, conferences organised, public lectures given and other extras are mentioned. Academics who work part time will fare badly on that criterion, and they tend to be women.

There are also more subtle worries: units of assessment are small, tight-knit communities that need to remain functional before, during and after the REF. It may be tempting for heads of department, directors of research and others involved in the selection process to submit colleagues who would otherwise become resentful or uncooperative. For the long-term good of the unit, it may be less damaging to exclude those who have outputs of a similar standard but seem to worry less about their status. These need not be women, of course, but it is possible that gender plays a role.

The classicist Mary Beard, in a recent lecture on the public voice of women, contrasted the profundity heard in male voices with the frequent accusation that women sound strident or whiny. In such circumstances, it is not surprising that some women try to sound like men. This has obvious drawbacks. If colleagues tell us that we did not make the cut for the REF, we may conclude that our play-acting has been discovered – that since we do not actually sound like men, we are not really authoritative after all.

My Facebook feed provided plenty of anecdotal evidence in support of these concerns. Many women worried about how colleagues perceive them. One described her habit of slinking into her department at weekends, hoping that nobody would notice her and conclude that she does not have a “proper”, fulfilling private life – only to be told months later (in a formal meeting on her promotion prospects) that she did not seem ambitious and was never seen working out of hours. Many pointed out relevant research – on men quoting each other more often than women, for example. One suggested that we adopt the slogan of the Memphis sanitation workers: I AM A MAN. It was not, on the whole, a particularly lachrymose or downbeat string of comments, more like Women on the Verge of a Nervous Breakdown. One friend revealed that she is pregnant; another suggested that a satisfactory sex life immediately makes you seem unsuitable for REF submission. The few men who contributed did not always cover themselves in glory. Just like the characters in an Almodóvar film, some sounded confident, but not especially bright. One told the woman-who-tries-to-disappear-at-weekends that if she would just work smart she would eventually get promoted (I’m sure he did not realise he was addressing one of the most eminent scholars of ancient science). Another breezily declared that academic hierarchies favour men and that women are less good at research. On that reading, all is well; but I am not so sure.

I rashly promised my Facebook friends that I would do something in answer to our concerns about women and the REF – even though I was not too sure what I should do. After some reflection, I went to see our (excellent) dean of diversity and equality and suggested to her a simple numerical analysis. If it is true that, in borderline cases, women are more vulnerable to exclusion from the REF than men, it should be possible to reconstruct this from internal ratings of research outputs. It stands to reason that men will have higher ratings overall. But I wonder whether, among those not submitted, women have higher ratings than men. If this is so, the analysis would suggest that women just missed being submitted, whereas non-submitted men were clear cases for exclusion. The dean immediately saw the point of the analysis and said she would take it forward.

I then thought about the importance of using networks and did something bold. I have been writing book reviews for Times Higher Education for years and considered that I could offer to write an article. I tentatively suggested this to my contact person, the books editor. It turns out that some nine in 10 academics who approach her with unsolicited offers to write for the magazine, or petition for their own books to be reviewed, are men. We both celebrated the fact that I was helping to balance out the figures. At that point, though, I started to think hard about what I wanted to say, if given the opportunity.

I remembered Zeus and decided that the third thunderbolt is indeed the one that matters. Research excellence is more important than money or power. I could now go on a training course to strengthen confidence and networks. I could try to rise through the administrative ranks: a few people at Durham have suggested that I should do just that. Those plans would, however, deprive me of my already limited time for research – and it is research, together with teaching, that attracted me to the profession in the first place. It seems to me that there are risks, if we get distracted. And, in any case, the distractions of power and money are available only to women who already have a good track record. What is of central importance for all academics, and at every stage of their career, is research. That is why we need to understand what happened in REF 2014 and, if women are indeed still lagging behind, do better next time.

Times Higher Education free 30-day trial

Register to continue

Why register?

  • Registration is free and only takes a moment
  • Once registered, you can read 3 articles a month
  • Sign up for our newsletter
Please Login or Register to read this article.


Featured jobs