The policeman at the centre of ‘Gategate’, who may or may not have been called a pleb by the Conservative Chief Whip, has been called in the press a ‘jobsworth’, a term certainly not complimentary even if not in the same league as the presumed ‘toff’s’ alleged insult of ‘pleb’ (Isn’t our class system wonderful!). I spend a lot of my time talking to and working with our university’s HR (Human Resources) Division, whom I am sure in the eyes of many in the community are seen as no better than ‘jobsworth’ people themselves. HR is too often seen as that annoying bunch of people who insist on you filling in endless forms and who waste your precious time apparently just for their own whim and pleasure. The reality is rather different, and I do get frustrated listening to the moans of colleagues who should know better.
Undoubtedly our UK governance structures, loosely interpreted, require academic institutions to fill in a depressingly large amount of paperwork. Take visas for visiting students and scholars, for instance. These were always a nightmare and have become increasingly so under current rules. However, I need say no more than London Met to make my point that correct form-filling is necessary for the protection of all. That is not to say I agree with government policy on this, of course I don’t (my University Council – to which I belong – has issued a statement about these issues); removing students from the migrant quota is the obvious logical thing to do and one with which BIS almost certainly would agree. The battle is with the Home Office and as long as they remain obdurate, doing anything other than complying with the insane paperwork is daft. That is not HR’s fault, so don’t blame the messenger.
What about all that paperwork associated with appointments? Surely that could be dispensed with? But that is exactly where there is some hope – if anyone reads the paperwork and monitors the statistics – of tracking whether equality law really is being properly followed. A recent, shocking, study published in PNAS looked at how identical CVs, submitted under either male or female names, led to very different outcomes, with the ‘male’ applicant being significantly more likely to be offered the job than the ‘woman’. The study demonstrated that there is an ongoing need to monitor what goes on when jobs are on offer. This US-based study revisited the problem of how much individuals – male and female alike – are unconsciously affected by the gender of applicants in ways that colour their decision-making because they implicitly believe women are less ‘competent’. The following is taken from the study’s abstract:
In a randomized double-blind study (n=127), science faculty from research-intensive universities rated the application materials of a student – who was randomly assigned either a male or female name for a laboratory manager position. Faculty participants rated the male applicant as significantly more competent and hireable than the (identical) female applicant. These participants also selected a higher starting salary and offered more career mentoring to the male applicant. The gender of the faculty participants did not affect responses, such that female and male faculty were equally likely to exhibit bias against the female student.
The putative position in question was therefore at the most junior end of things, for which a graduate student was being recruited, but there is no reason to expect that things would have been any less skewed for higher-ranking positions. Indeed I would argue that things are liable to be only worse for hiring in the higher echelons. The authors concluded that
Rather, some women may persist in academic science despite the damaging effects of unintended gender bias on the part of faculty.
I suspect many of my readers may be nodding in agreement at this point. The challenge for an HR department is to work out what to do to overcome such gender bias. My PVC with responsibility in this area has already asked just that question having read the PNAS paper. Some years ago, courses were run in my university called ‘Responding to Merit’ in which some of the underlying issues about unconscious bias were raised and discussed. I attended one of these (they were of course voluntary), but I fear that most of the attendees were people like myself who were already fairly aware of the pitfalls that gender bias can produce. Monitoring statistics for every job: number of applicants by gender, number shortlisted/interviewed etc plus the final outcome is at least a step in the direction of providing hard evidence with which to work. It is insufficient and of course the conditions are far less controlled than the study itself, making the results so much harder to interpret.
However, if academics in universities just shake their head every time a directive turns up from the HR department demanding a set of forms be filled in for every vacancy, clearly little progress is going to be made. Nothing short of a culture change, a realisation at every level in the organisation that problems persist, is likely to improve the situation. This is not a question of conscious discrimination, which makes it all the harder to root out. The study authors conclude with the statement that
These results suggest that interventions addressing faculty gender bias might advance the goal of increasing the participation of women in science.
They do not specify what such interventions might look like. Lining up all the professors in a hall and talking at them is unlikely to be successful. Nor is an online training programme, I fear. But constant reminders – verbal reminders probably work best, from heads of department or equivalent – to be on guard could help. So could mentoring and training for newly recruited faculty, to make sure that before they ever get involved in hiring they’ve had any unconscious prejudices and preconceptions drawn to their attention. Note, these remarks are addressed to women just as much as men. Somehow, our societal expectations, ingrained from our earliest years, cause us to believe that young women are less ‘competent’ overall, when it comes to a scientific job, than the identical individual masquerading under a male moniker. This is prejudice, plain and simple.
So, next time HR sends a form in your direction, think not just about the probable tedium of having to fill it in, but about the underlying reason for its existence. It may be to satisfy some unsatisfactory government policy, it may be to try to identify where problems lurk in local decision-making, but the chances are there is some underlying rationale. Maybe in a perfect world these forms would be unnecessary (I’m sure readers will be keen to nominate instances where they really do believe there is no point in this or that HR directive), but I for one think HR serves a hugely important purpose and I’m delighted to be doing my bit, as the University’s Gender Equality Champion, to champion the work they do against those charges of ‘jobsworthiness’ to which they are so often subjected.
If you want to read a fuller discussion of the implications of the PNAS paper take a look at this post and fellow OT blogger Jenny Rohn over at Occam’s Corner. I am sure there are others out there and I have not attempted to discuss the findings per se here. Suffice it to say I find it shocking and none of us should ignore the implications. I have also begun to see analyses suggesting how the study protocols could have been improved, though only so far through informal routes so I can’t post links. Nevertheless, all analysts I have come across seem confident that the evidence supports the overall conclusion of the existence of unconscious bias in the study.
We actually have such a course at UCL, called ‘Avoiding Unconscious Bias’ provided by Pearn Kandola for us. It was introduced for the first time last year, and trialled in Engineering (Anthony Finkelstein has been an important supporter of and innovator in this area). I then attended it as a HoD for Arts and Humanities. I think it was pretty much compulsory- at least strongly encouraged- for all Heads and people such as faculty tutors. I found this to be very impressive indeed, although I am, of course, already interested in such issues. I think it is vital that such courses should be taken at least by HoDs, and anyone else likely to be chairing an appointment panel, as it has made be think carefully about what i do, and how decisions are made in such circumstances.
this situation in the first world , makes me worry about my daughter future ,
great effort, thanks .
It is interesting and a little disturbing to me that the thoughtful analysis of recent research on embedded bias in Athene Donald’s blog is being set beside Jenny Rohn’s Guardian piece. The differences in these two accounts may be subtle but they are also significant; and whoever generated the title of Rohn’s post embraced its spirit with, “Women in science, you have nothing to fear but your own subconscious.” Dame Athene notes the role of HR in assisting with bias-training as one small but necessary step in good practice, while Jenny Rohn suggests that a reflective pause plus training is likely to result in marked improvement. The latter argument makes me uneasy, and I’ll try to explain why.
Explanations of women’s stalled advance in certain areas are often stacked up like layers on a wedding cake. The constituents of the heavy bottom layer have changed over the past three decades. Previously, it consisted of market forces and social trends: since women earned less the rational family allocated women the unpaid roles of caring labour while the men devoted their time to paid employment. This rational response to market forces then influenced women’s decisions about education and career investment, and the social norm of the housewife was born. In recent years this large, most substantial layer has been mixed and messed by neurosexism (a term coined by Cordelia Fine), stolidly stating that women are hard-wired to eschew science things in favour of more touchy-feeling things.
The second layer has remained pretty much the same over the past decades. This layer looks at male prejudice: Men, the argument goes, want to sustain their privilege and power, and so keep women out, either with biases hiring procedures or, when that does not work, harassment or more subtle forms of denigration.
The third layer is covered in the thick sticky icing of women’s psychology. In the 1980’s it was argued that women scupper their chances of success by being overcome with the desire to be looked after (The Cinderella Complex). Today the argument is only slightly different: women share the bias against women’s ability and competence, and rule themselves (and other women) out via their endorsement of male bias.
When a cool look at the numbers of women in science forces home the hypothesis that there is bias at work, the cake is cut, and in a free for all the guest analysts picks and choose, when what is required is a pattern or puzzle-spotter who transforms the layer model into one of interconnected structures. Jenny Rohn notes other factors – most importantly, the shape of a science career – but seems to brush these aside as she focuses on unconscious bias. The notion that women express bias (as well as men) is so startling that it can obscure how deeply interconnected bias is with context. But above all, it lets us rest easy with the notion that we can fix matters without changing much else in the world.
This approach suggests a huge underestimation of the persistent power of bias. One problem in catching it out is that bias of this kind is not static. It becomes more salient and aggressive in some circumstances and purrs harmlessly like a sleeping kitten in others. Unfortunately for women in science, any gender deviant (that is, a man or woman who does not comply with gender norms) is likely to be punished by others. (See, for example, Rudman and Fairchild. 2004: Reactions to counterstereotypic behavior: the roll of backlash in culture stereotype maintenance”) Therefore, resisting bias is not just a matter of becoming aware of it; it is a matter of social and personal courage, and willingness to take the consequences.
Bias is also, as Cordelia Fine warns in A Mind of Its Own, highly secretive, good at disguising itself, easily harnessed to the “hotbeds of motivation and goals” that follow us around. While serving on an appointments committee, think about who you might want to impress or undermine. Among the candidates, who do we think will make us feel good about ourselves, and who might ignore us? Our agenda may not readily make itself known to us, and we fallible and self-serving humans need more than the most astute HR trainer to coax us towards fairness. Public conversations such as these generated by Athene Donald and (for all my reservations) Jenny Rohn offer a promising start.
Many thanks to @utafrith for tweeting links to both this and the Guardian blog.
I agree that HR–and other types of checks and balances into hiring practices–should serve a more important role in academic hiring. A problem as complicated and multivariate as bias/prejudice, with many different origins and a wide range of outcomes cannot be addressed by a single fix. It will take many interwoven approaches to improve this real problem, including new hiring structures from the top and increased awareness at the grassroots.
I wish I could forward your blog to the head of the department. I don’t want to sound like a sour grape (sigh!!), so just working extra hard to compensate for the bias.
Great post – sent it on the UEB & HR
At the University of Manchester there’s a pretty good online course for newcomers on equality and diversity regarding all kind of biases (gender and otherwise). I consider myself pretty unbiased, since I really don’t have any problems with other people’s gender/ethnicity/religion & so on (provided they return me the favour, that is). Yet it made me better appreciate how people with disabilities might be more sensitive to certain comments/behaviours even if they should be considered innocent by everyone else.
Going back to HR: filling in the online forms & viewing videos was, sometimes, tedious. But I DID learn something.
I’ve come across a couple of other relevant articles. This relates directly to the same PNAS article as I describe here. Whereas this discusses other and earlier studies, looking at a wider range of diversity types and considering how faculty respond (or don’t) to initial email contacts from putative students, but is every bit as shocking.
rs
Why not send it to your head of department? As another commenter says, he’s sent it to his University Executive Board as well as HR and a tweeter told me she’d sent it on to her university HR. I certainly sent the PNAS article on to my PVC, as I indicated above. Scientists like evidence before they act. Well, the PNAS article is evidence!
And now this commentary on the PNAS article in Science has been drawn to my attention, as well as a piece in the New York Times. It is excellent to see this story getting so much attention. I am sure the only way to overcome unconscious bias is to make it conscious, so that people spot their own internal assumptions.
What struck me is that the bias was reported to be relatively uniform across age, gender, tenure status, and field. That suggests that the effect is somehow a shared internal response to common external factors, or common experience (using the term loosely). It seems unlikely that “old male physicists” and “young female biologists” would somehow have such similar internal response to different external factors.
Perhaps that common factor is that sexism is so pervasive that even young women in fields with relatively high proportions of female researchers internalize this bias as well.
But I wonder whether it is possible that the common factor is the emphasis on programs for women in STEM, diversity training, (official or unofficial) targets, positive discrimination, etc.
Perhaps people internally compensate (over-compensate?) for their sense of common institutional practice, especially where the point of the exercise is judging a CV in the abstract (nobody collects external evaluations for a lab manager position in real life).
I don’t know. It’s a hard and important problem and it’s not clear what the solution is. But I’m not convinced I agree with the conclusion “do more of what doesn’t seem to have worked in the past”, especially if the hypothesis above is correct. (But I’m pretty sure it shouldn’t involve more bureaucracy and paperwork.)
As one of the presenters of the ‘Responding to Merit’ workshops it is great to see the PNAS paper having such impact. I also agree that most of the participants, and there were only a few tens of them in total over four years, were people who were already aware of the issues. This raises the question: what would work? Something like Professor Virginia Valian’s Tutorials for Change (http://www.hunter.cuny.edu/gendertutorial/.) have the advantage that people can do them in their own time rather than take half a day out of their already crammed schedules. They differ from online training courses in that they were written by an academic for academics. The STRIDE project at the University of Michigan (sitemaker.umich.edu/advance/recruitment__stride_) has had some success. (In the two years pre-STRIDE there were 10 women in 73 appointments (14%). In the three years following the introduction of STRIDE there were 46 women in 134 appointments (34%).) Key features that have been identified (Ref.) as contributing to STRIDE’s success are:
– STRIDE is run by academics for academics. Representatives from different disciplines bring cultural knowledge, examples and experience. Having committee members who are respected in their own fields gives credibility.
– Compensation for time spent on committee, that is, it is part of the job not an optional extra.
– Staff support, for example, someone to locate and circulate relevant articles.
– Debating issues rather than avoiding conflict.
While it might not be possible to replicate the STRIDE programme, it makes sense to look at the factors that contributed to its success and see if they can be incorporated in other programmes. Virginia Valian’s book, Why so few? the advancement of women, has a number of suggestions both for individual women and for organisations (Chapter 14), including practical suggestions for minimising the effects of unconscious bias in hiring decisions. The National Academies’ Beyond Bias and Barriers report contained many recommendations but they were mainly at a strategic level.
We should be careful not to concentrate solely on recruitment. Evaluation bias can occur in many contexts,for example: student evaluations of lecturers, which students go to which conferences, who gets encouraged to apply for prestigious fellowships, who sees themselves as a credible candidate for a prestigious fellowship, which committees people are invited to join. As Terri Apter has noted unconscious bias is not a simple variable independent of context. Context affects whether and how unconscious bias is expressed but also what the underlying unconscious biases are.
Practical steps that could be taken include:
– Incorporating material on unconscious bias in existing training for recruiters, managers (including principal investigators) and lecturers.
– Preparing a one page briefing on unconscious bias for members of selection panels backed up with web material and/or a workshop or facilitated discussion. (Though one has to bear in mind that assembling a selection panel in the same place at the same time is hard enough to achieve for the purposes of doing the selection. Adding another meeting just may not be feasible.)
– Having a group of academics from science and engineering who are prepared to engage with the issues raised by unconscious bias and to discuss them with other academics (along the lines of the University of Michigan’s STRIDE committee).
I agree that reminders from Heads of Department and mentors, if people have them, are likely to be the most effective ways of maintaining a culture in which people feel they should take unconscious bias into account but I think there needs to be a push to get the information out in the first place. It also needs to be part of a 5-10 year strategy, not a 1-2 year initiative.
I think one of the mistakes I made was to assume that when people were presented with facts, for example, the results of Steinpreis et al in The Impact of Gender on the Review of the Curricula Vitae of Job Applicants and Tenure Candidates: A National Empirical Study (pdf)(Sex Roles, 1999), they would change their minds. In fact, people often have an emotional reaction to these studies. This is one of the reasons that simply telling people to do something is ineffective. It also means that people need to be able to talk about what they have learnt. There is a problem here, of course. Discussions take time, but time is what people are not willing to spend.
I think it would help if HR were able to see beyond compliance and think about marketing themselves. Why not put on your forms ‘We collect these data to comply with the xxx Act yyyy’ or ‘We collect these data in accordance with the University’s Policy on xxx’? People may feel that filling in forms is onerous. They should try comparing the effort involved to the time, energy and financial costs of clearing up the mess after someone has got it wrong. Do it this way because we say you must is not an argument likely to appeal to academics.