In any organisation, the effects of institutional culture on staff and performance are very evident. But while it is easy for senior managers to assert their cultural aspirations in high-level documents, implementing them on the ground is hard ¨C especially in such complex organisations as universities.

The University of Oxford, for instance, expresses strong commitment to equality in its policy statements, yet women and minorities remain under-represented in tenured and leadership positions. Moreover, in published in , my colleagues and I find that women¡¯s experiences of Oxford¡¯s culture are less positive than those of men on six out of 12 ?for those in medical science departments and 10 out of 12 dimensions for those in social science departments.

In both disciplines, women find work less energising and personally satisfying than men do. They also feel less confident in their ability to progress their careers and overcome barriers, and are less convinced that the university treats women and BME staff equitably. A significantly lower proportion of women than men have a permanent contract of employment (conferring job security), a formal leadership role (conferring higher status) or a college affiliation (conferring in-kind privileges and often significant monetary benefits). Not surprisingly, therefore, more women than men intend to leave Oxford?owing to dissatisfaction.

Our study finds that women suffer more than men do on account of Oxford¡¯s culture of overwork, non-transparent tenure and promotion criteria, devaluation of teaching and administration, and the rewards accorded to stereotypically masculine patterns of behaviour, such as individualism and self-promotion. Women also feel more acutely the exclusion from collegial and departmental life of researchers and academic-related staff on fixed-term contracts, as well as divisions between academic faculty and administrative, professional and other support staff.

É«ºÐÖ±²¥

Importantly, both women and men highlight the salience of intersectional connections between gender, socio-economic class, race/ethnicity and sexual orientation/gender identity. Class is particularly prominent in the responses, with many expressing the view that Oxford is dominated by the upper and middle classes. According to one disenchanted respondent, college life is ¡°the logical next step for posh boys from Eton who feel at home with all the pomp¡±.

Yet the disciplinary differences between the responses are very prominent. When we conducted our study, all 16 medical science departments at Oxford had joined the Athena SWAN gender equality scheme and implemented action plans to attain a silver award, on which eligibility for funding from the National Institute for Health Research depends. By contrast, no social science departments or colleges had done so.

É«ºÐÖ±²¥

Many medical departments had recruited equality and diversity professionals. They had?made??to increase support for women¡¯s careers, better appreciate caring responsibilities in career progression, improve conditions for flexible working, and challenge discrimination and bias.

The responses to our survey reveal that this is paying off. Although long-standing power and pay imbalances remain, a more supportive and inclusive culture in medical sciences is emerging. More faculty and staff have positive mentoring experiences regardless of gender, for instance. And a significantly lower proportion of faculty and staff intend to leave the university?because of dissatisfaction.

Although we cannot claim causality, we argue that the widespread implementation of Athena SWAN, driven by the NIHR funding incentives, has contributed to creating a more supportive and inclusive culture in medical sciences.

The scheme is not without . It will not, by itself, give rise to the feminist revolution that we need in academia, both for social progress and as . Its reliance on peer review of the action plans generated by participants inevitably constrains the most radical thinking, and the ongoing Athena SWAN also highlights concerns about both the award process and administrative burden on institutions.

É«ºÐÖ±²¥

Efficacy is further limited by the devolved nature of academia, whereby the keys to real change are held by individual principal investigators, research teams and departments. Those who are not fully committed to can sit out the uptake and implementation of the action plans and game the system merely to qualify for NIHR funding.

Yet, with its implementation in , , and, most recently, and , Athena SWAN has become an influential global movement. It is a generational opportunity to modernise the Dickensian system of academic capitalism with a functional equivalent of the welfare state. For instance, it has prompted many departments ¨C and some external funders ¨C to offer bridge funding so that female postdocs have protected maternity pay even if their fixed-term contracts end during maternity leave, and so that they can have more time to generate papers and grant proposals if they return shortly before their contracts end.

Such schemes need to be significantly scaled up, but they are an important step in the right direction. So rather than sighing about its limitations, we should all rally round Athena SWAN and work hard to continuously improve its design and implementation. In doing so, we need to develop ¡°a science of Athena SWAN¡± to establish what works best. We also need to raise the profile of Athena SWAN professionals and, ideally, follow Oxford¡¯s example in critically examining its own efforts to improve institutional culture ¨C and openly sharing the results.

Evolution may be inferior to revolution, but it is far better than stagnation or regression.

É«ºÐÖ±²¥

is senior research fellow in health policy and management in the Radcliffe department of medicine, University of Oxford.