Written Evidence Submitted by the Mathematical, Physical and Life Sciences Division of the University of Oxford

(DIV0063)

Summary

To improve diversity in STEM we need to take a more holistic and intersectional view of equality, diversity and inclusion (ED&I), using our skills and abilities as a scientific community to tackle disparities and ensure that the STEM environment is an inclusive one. Advance HE’s Athena Swan Charter has provided a useful framework for understanding the issues around gender diversity in higher education. Using that model of regularly analysing quantitative and qualitative data and consulting widely with relevant groups to create comprehensive action plans to advance change, would be beneficial across ED&I areas. We recommend that such a methodology should be applied to encompass intersectionality and a broadened view of ED&I that would enable individual STEM subjects to address the most pressing and problematic issues within those fields (e.g., gender imbalance in physics and engineering, or ethnicity imbalance in biology and earth sciences, discussed below).

To achieve this, data collection around protected characteristics and socio-economic background needs to be improved, standardised as much as possible, and have more granulation (particularly for ethnicity data), whilst adhering to data protection laws. While high quality data is invaluable to help us understand the nature and nuances of our challenges with diversity in STEM, progress cannot be stalled in an endless pursuit of more data. Resources and attention need to be focused on actions. It is important to have evidence-based actions, but we need to also be bold enough to try new, innovative, and creative interventions to add to our existing evidence-base. Monitoring and evaluation of interventions needs to become the norm in order to have continual improvement in ED&I.

The biggest gains will undoubtedly be made by changing the perception of STEM in our society, and by real cultural change within STEM disciplines so that everyone can feel that they belong. Therefore, more investment in STEM at early ages will reap rewards further down the line. Properly funding laboratory-based subjects in all schools is critical to overcome the socio-economic biases in higher education, alongside education regarding what careers are opened up by studying STEM subjects. Within STEM, people and particularly research leaders, need to be more conscious of how to create inclusive and welcoming spaces, and held accountable for poor behaviour that harms others. Those who enable collaborative, fair, and supportive environments, where people from all backgrounds are able to flourish, should be recognised and rewarded for their efforts.

Funding bodies can play a huge role in instigating cultural change by ensuring that funding is provided based on both scientific potential and the quality of the working environment for developing people from all backgrounds. The government and funding bodies have a responsibility to embed ED&I in all that they do, ensuring that people of all backgrounds are included in and can contribute to our scientific research.

Introduction

1. The Mathematical, Physical, and Life Sciences (MPLS) Division is one of four academic divisions of the University of Oxford. The University is widely recognised as one of the world's leading science universities, and is a research-intensive, international institution drawing in students and staff from around the world. The disciplines within MPLS regularly appear at the highest levels in world rankings and have been evaluated as conducting world-leading and internationally excellent research in UK research assessments. Our nine academic departments include: Biology (recent merger of Zoology and Plant Sciences), Chemistry, Computer Science, Earth Sciences, Engineering Science, Materials, Mathematical Institute, Physics, and Statistics. In addition, Begbroke Science Park, the Doctoral Training Centre, and the Oxford-Suzhou Centre for Advanced Research are also part of our Division. MPLS is home to approximately 7000 students (around 3500 of which are postgraduate students), 1100 researchers, 600 academics, and 1000 professional and support staff.

2. This submission was prepared by the MPLS Equality, Diversity and Inclusion (ED&I) team, with input from divisional colleagues, ED&I Steering Group members, ED&I Fellows, and colleagues from the central Equality and Diversity Unit and the Medical Sciences Division (MSD). It is informed by our University-wide Staff Experience Survey (SES 2021[1]) which reports our divisional results[2], focus groups, and Athena Swan work.

3. Our Division, and indeed the wider institution, has worked hard to embed ED&I at all levels, across disciplines, degrees, job roles, and with our policies, programmes, and practices. Each of our academic departments in MPLS and MSD have an Athena Swan Silver or Bronze award (current total of 18 Silver and 8 Bronze awards), evidencing their continued commitment and ongoing action to improve gender equality. Our divisional activity and resource (both in staffing and funding) has increased to include race and ethnicity, disability, LGBTQ+, and most recently, enhanced support for mental health and neurodiversity. We have actively sought to collect and understand our relevant ED&I data, consult widely with staff and students, create action plans to address identified challenges, implement actions, and regularly monitor and evaluate their impact. It is a collaborative endeavour, actively working with departments, central University teams, student groups, and ED&I colleagues in other divisions to create an inclusive culture within the University.

Topic 1: the nature or extent to which women, ethnic minorities, people with disabilities and those from disadvantaged socioeconomic backgrounds are underrepresented in STEM in academia and industry

4. Diversity within and across categories – importance of intersectionality. The groups listed as part of this inquiry (including those who have lived experiences and identities across more than one of these categories) have shared and differing challenges, rooted in trying to survive and thrive in an environment that was not built with their needs in mind, and at various historical (and present) times, actively excluded them. This inquiry, covering such broad categories, needs to keep in mind the diversity of experience within these groups and intersectionality across categories. For instance, an able-bodied white woman from a lower socio-economic background will likely have a different set of challenges from a disabled woman of colour from an economically privileged background, while still sharing some experiences navigating a male-dominated sector. Each of us are unique and have complex identities comprising multiple characteristics.
• ‘Women’ as a category is incredibly broad, with differences across race and ethnicity, religion, age, disability, sexuality and identity, caring responsibilities, and other characteristics. Significant progress has been made to address gender inequality in STEM, with Athena Swan having a large impact on recording and analysing data and setting out actions that will address the evidential shortfalls. However, to move further forward we need to understand the nuances and intersectionality within large demographic groups. For example, addressing the broad gender inequality may still result in fewer women from UK-ethnic minorities.
• ‘Ethnic minorities’ is a large umbrella term that can homogenise vastly different experiences and histories, and obscure the lived experiences of people from particular ethnicities. For example, the term ‘Asian’ can include Indian, Bangladeshi, Pakistani, Chinese, and other East and Southeast Asian ethnicities. According to the APPG study on Diversity and Inclusion in STEM cited in introductory text for this inquiry, people of Indian ethnicity are over-represented in the STEM workforce (at 4%), while Pakistani/Bangladeshi people are underrepresented (at 2%), but they would often be placed in the same ‘Asian’ category.[3] Furthermore, STEM is an international endeavour, so terms such as ‘over-represented’ and ‘under-represented’ need to be clearly contextualised.
• With respect to disability, there is a huge variety of experiences depending on whether we are discussing specific learning disabilities, physical disabilities, long-term health conditions, mental health difficulties, or others. Our staff survey results showed that work-related stress was a key issue for all staff, but even more so for disabled staff – less than half of MPLS staff with a disability find the average level of work-related stress reasonable compared to 56% of staff without a disability. Furthermore, for staff with a disability and recent experience of mental ill-health, this drops to 36% [4].
• Religion and faith are not widely addressed in many ED&I initiatives, but people of faith can feel marginalised because of their religious beliefs. This can be a particularly divisive issue in STEM, where science and religion are still often regarded as incompatible.
• Socioeconomic status is not a protected characteristic under the Equality Act 2010, therefore it often gets overlooked in ED&I activity even though people from disadvantaged socioeconomic backgrounds can feel ‘othered’ and have lower attainment in schools. These considerations tend to be limited to student admissions, considerations of state versus private schools, bursary and scholarship decisions, hardship funding and other financial aid matters. With respect to staff, the broader issues of casualisation of labour (e.g., fixed term, insecure and zero-hours contracts), living wage payments, and marked decline in pensions and benefits are significant concerns for all staff, but can have a disproportionate impact on those from disadvantaged socioeconomic backgrounds. It is important to look at the intersection of socioeconomic background with other protected characteristics – for example, if ethnic diversity is improved, but access is limited to mostly ethnic minorities from privileged backgrounds, there is still more work to do.
5. Existing evidence. There is ample existing evidence of the nature and extent of underrepresentation among women in STEM, and increasing research on ethnic minorities and disabled people in STEM. The nature of the problems and barriers are well documented. The impetus needs to be on the actions carried out to address these longstanding disparities.
• The House of Commons Science and Technology Committee produced a comprehensive report on ‘Women in scientific careers’ from January 2014, based on a call for evidence in which the University of Oxford submitted a detailed response. While some time has passed and some change has occurred, many of the issues identified and recommendations shared remain relevant today.[5]
• More recently, the Royal Society commissioned several reports related to ‘Ethnicity in STEM’, looking at data for students and academic staff in higher education over an 11-year period, and reviewing the profile of postdocs in the UK eligible for Royal Society early career fellowship programmes.[6] The Royal Society also commissioned reports looking at quantitative disability STEM data for students and academics in higher education, and qualitative research examining barriers to progression for disabled scientists.[7][8]
• Universities UK report ‘Tackling Racial Harassment in Higher Education’ (November 2020)[9] and the ‘Black, Asian and Minority Ethnic Student Attainment at UK Universities: #Closingthegap’ report [10] are not STEM specific, but offer useful findings relating to the experience of people of colour in higher education.

6. COVID-19 implications. One area for further research can look at identifying the impact of the COVID-19 pandemic on the stated underrepresented groups. As more people become disabled from long COVID, and/or suffer from mental health difficulties during this difficult period, there will likely be an increase in disabled staff and students in STEM. There is already existing research studying the impact of the pandemic on academic women showing the reduction in publications from women researchers, risking their career status and opportunities for progression.[11] Ethnic minorities have been disproportionately impacted by the pandemic in terms of health outcomes and economic impacts. With a higher proportion of ethnic minorities on fixed term and insecure contracts, and from outside the U.K., there may be short- and long-term effects on their ability to secure employment in STEM, and stay in the U.K., if they wish to do so.

7. Nonbinary and transgender issues. We understand ‘women’ to include trans women. There are particular issues facing transgender people within STEM that warrant recognition and action, which have not been specified within this inquiry. Similarly, nonbinary, genderqueer, or intersex people who do not identify in gender binary terms can also be ‘othered’ within traditionally male and ‘masculine’ STEM spaces. Programmes aimed at ‘women’ in STEM may still operate and use language in a binary manner (even if they seek to include marginalised genders). There needs to be greater recognition of the differing challenges faced by people across the gender spectrum. A first step would be to make language and labels (if needed at all) more inclusive, and their correct usage commonplace.

8. Absence of LGBTQ+. The absence of LGBTQ+ communities within this inquiry on diversity in STEM is notable. We understand that data on LGBTQ+ groups are not regularly collected and there are challenges to gathering sufficient and accurate data. However, even with these limitations, our LGBTQ+ communities within STEM are becoming ever more visible and active, and deserve to be included. Staff and students in MPLS have been very active in this space for years, and have been involved in organising the annual LGBTQ+ STEMinar Conference[12], hosted online by the University of Oxford last year and part-sponsored by MPLS. For LGB+ staff in MPLS who responded to the staff survey, 75% felt able to be themselves at work, compared with 84% of heterosexual staff in MPLS, and 44% felt integrated into their department compared with 56% of heterosexual staff[13]. Creating and maintaining a welcoming and inclusive workplace for LGBTQ+ staff needs to be addressed.

9. Data concerns. To understand the nature and scope of the issues for these underrepresented groups in STEM, we need to collect sufficient quantitative and qualitative data, while ensuring confidentiality and data protection. While data categorised by sex (as a binary variable) has been collected and analysed extensively due to Athena Swan activity (although less so with gender identity), there is a less comprehensive approach to race and ethnicity data, disability, and – particularly – socioeconomic status. Additionally, in our institution and likely true in the wider STEM environment, there are high rates of nondisclosure (people responding ‘prefer not to say’ or not updating their records). As such, there are significant data gaps that hinder proper analysis.
• Data needs to be disaggregated as much as possible to dig deeper into large categories such as ‘ethnic minorities’ and ‘disability’.
• Data should be analysed with an intersectional approach in mind to consider how different aspects of a person’s identity could affect results.
• Qualitative data needs to be included to understand people’s experiences, not purely focusing on quantitative data. Groups with small numbers should not be summarily dismissed or ignored – different methods of gathering information from severely underrepresented groups should be considered.
• Socioeconomic background data is rarely collected outside the undergraduate stage, and even then, issues with accuracy, consistency and having contextual information pose challenges.
• It is more challenging to collect data regarding career progression, compared with recruitment or staff/students in post. Furthermore, we are not able to capture a STEM sector view to consider changes and trends within and between employers, academic institutions, public bodies, and others. Diversity lost in one type of employer may well be gained in another, but we do not currently have reliable data to track if this is the case.

10. Broad definition of STEM. There is a need to look at different disciplines of STEM separately (for example between maths, physical sciences, and life sciences), as there are nuanced but important differences in how diverse each can be. For example, Oxford Biology and Chemistry have relative gender equity at the point of undergraduate admissions, but proportions of women decrease as seniority increases. While other departments have much lower proportions of women undergraduates due to limiting factors affecting the available pool of applicants such as lack of take-up in schools (for example, Physics, Engineering Science, and Computer Science), and this disparity is usually maintained or increased at later academic stages. With respect to race and ethnicity, based on July 2021 snapshot data, 16% of our staff in MPLS identify as ‘Black and Minority Ethnic’ (BME), with 18% unknown ethnicity but these proportions vary widely across our departments, with Engineering Science and Materials having significantly higher proportions of BME staff, and Computer Science, Earth Sciences, Statistics, and Zoology having lower proportions.

Topic 2: the reasons why these groups are underrepresented

11. Exclusive cultures. There are a number of different and wide-ranging societal and academic/STEM specific reasons for underrepresentation of women, ethnic minorities, disabled people, and people from disadvantaged socio-economic backgrounds. At the core, we live and operate in a culture that generally sees cisgender, white, able-bodied men from advantaged socioeconomic backgrounds as the “default” in STEM fields, with those who come from other backgrounds potentially feeling excluded from this culture. Some of the specific reasons articulated in the studies referenced in response to Topic 1 include:
• Historical exclusion from scientific fields;
• Stereotypes of who scientists are and what they do, that are limiting and can put off people who do not fit that mould;
• Hostile or toxic environments that push out marginalised people;
• Lack of visible role models, mentors, sponsors, and networks;
• Long-hours working cultures;
• Expectations around laboratory and/or fieldwork that can exclude particular groups;
• Lack of accessibility of both physical and virtual spaces;
• Inadequate access to resources (e.g. equipment, books, funding);
• Limited information on STEM careers throughout school.

12. Rethinking the ‘leaky pipeline’. It is not a ‘leaky pipeline’ implying a passive filter – people are forced out due to hostile or toxic cultures and an unsupportive system that limits choice.
• Retention is just as important as recruitment. Often the focus is on increasing numbers and numerical representation at the expense of understanding the experiences of underrepresented groups already present in an institution, and how to enable their progression and career success. We need to shift towards creating inclusive cultures where marginalised people truly feel that they belong, and know that they have psychological safety to contribute their thoughts and ideas, and that their abilities and achievements will be acknowledged and valued.
• Workplace safety. This is different from a feeling of discomfort or experiencing bullying and harassment. For example, for workers in many STEM industries who deal with specialised equipment, appropriate personal protective equipment (PPE) is necessary to complete a job safely. In many cases, the default for PPE is for able-bodied men, which means that it can be difficult for women and disabled people to access PPE that can allow them to complete their work safely. There are also issues for people of particular faiths (for example, people who wear hijabs, niqabs, and other coverings). A 2016 survey conducted by numerous organisations including the trade union prospect and the Trades Union Congress (TUC) showed that 57% of women found that their PPE sometimes or significantly hindered their work, and that just 29% of women said that their PPE was specifically designed for women – an issue that men do not face to the same extent.[14]
• Fieldwork safety. STEM research in areas or countries where being from particular underrepresented groups (e.g., LGBTQ+, women, ethnic minority) can expose STEM staff and students to unpleasant behaviour, discrimination, or threat to their safety. Limited access to (or interest in) fieldwork can impact careers and career decisions.

13. Explicit and implicit biases. There are a number of cognitive biases that lead us to hire and work with people who are similar to ourselves.
• Less access to networks. In a highly competitive sector, often who you know is more important than what you know. Those with access to social and professional networks, and specifically well-respected and powerful ‘sponsors’, have a greater chance of success. People from underrepresented backgrounds are less likely to have access to such networks and sponsors. For example, as a PhD/DPhil student, success can depend on the supervision received. Students from underrepresented backgrounds may find it more difficult to find a supervisor in their field that they can relate to and work well with, and who can understand the barriers they face.
• Impact of bias not limited to recruitment. Often the focus is on biases that affect the recruitment process, but we need to keep in mind the impact of biases related to inequitable work distribution and assignment of key career defining/advancing tasks that affect career progression. People from underrepresented groups often face a ‘prove it again’ bias where they have to continually work harder and prove themselves multiple times compared with white men.
• Maternal wall bias (also known as family responsibilities discrimination). This bias refers to the discrimination faced by working mothers or pregnant women because of pregnancies and maternity leave, or for those who have caring responsibilities. Colleagues can view working mothers or those with caring responsibilities as less competent or less committed to their jobs, resulting in barriers to career progression and a potentially hostile working environment.
• Research group dynamics, as well as manager/supervisor relationships. Staff and students can experience bullying and harassing behaviour from their peers and their managers/supervisors that can affect their health, safety, and wellbeing. Senior academics may think that junior members need to ‘earn the right’ to a career in STEM, e.g., by working long and unsociable hours, as that is what they did. Yet this perpetuates the status quo, and goes a long way to ensuring that we only retain the same kind of people who have the capacity to work in this manner. We should encourage and reward those scientific leaders who recognise and support diverse talent.

14. Societal issue. Lack of diversity in STEM is a wider societal issue, of which higher education is one element of a bigger picture. There is ample research on STEM issues in education and the importance of ‘science capital’. Stereotypes of STEM scientists contribute to young people not pursuing STEM careers, and this starts very early. The wide range of outreach programmes could be more coordinated, address young people at earlier stages, and engage with teachers, parents and the wider community.
• According to the 2014 WISE report, ‘pupils from age 10 start to self-identify as “not STEM”’, and ‘careers from STEM are not popular aspirations for students age 10-14’[15].
• From Year 3 (age 7–8) onwards girls’ views regarding future careers tend to conform to traditional notions of ‘girls’ jobs’ and ‘boys’ jobs’.[16] These notions are reinforced by parents’ views. Professor John Perkins’ Review of Engineering Skills: The Department for Business, Innovation and Skills, November 2013.[17]
• Teachers and the wider community need to be supported to recognise and challenge these stereotypes, and reduce/eliminate the impact of these biases on young people. STEM should be taught in an engaging, relevant, and creative way.

15. Parents are strong influencers in the lives of their children. Additionally, if parents are providing financial support for their children to attend university, there may be added pressure for students to pursue careers deemed acceptable to their parents.
• The 2011 Public Attitudes to Science Survey highlights that ‘the data suggest parents are somewhat more likely to see science as an activity for boys rather than girls, showing the need to engage parents as well as children.’ [18]
• A 2013 study of the career aspirations of young people[19], looked at ‘science capital’ which refers to ‘science-related qualifications, understanding, knowledge (about science and “how it works”), interest and social contacts’ and found that ‘families exert a considerable influence on students’ aspirations’. The study also found that ‘some Black parents recognised that they personally lacked knowledge and awareness of where science can lead, and felt that this might be a disadvantage for themselves and their children’.[20]
• The WISE ‘Not for people like me?’ report[21] highlighted:
1. There is untapped potential in the family as an important encourager or influencer for young people, particularly in the Asian population.
2. It is important to ensure that parents, particularly in lower income groups are aware of the full range of careers available.
3. Mothers in particular, need to know their daughters could be happy in a career from physics/in engineering, and that the working environment would be supportive.
• Based on these findings, the WISE report recommends to ‘Communicate with parents who are significant influencers. Use information about the demand for STEM skills and qualifications, particularly the commercial value of mathematics and science qualifications.’[22] Not everyone understands the careers that are possible with an education in STEM.

16. Resistance and backlash. Given increased ED&I and positive action efforts to right historical wrongs and address existing inequity, there are some perceptions that these are unnecessary ‘handouts’ and that underrepresented groups are actually given more opportunities than cisgender, white, able-bodied men from advantaged socioeconomic backgrounds. This minority view needs to be strongly and regularly challenged, coupled with greater recognition for those who do lead by example by ensuring everyone has the opportunity and space to flourish, whilst also challenging such views.

Topic 3: the implications of these groups being underrepresented in STEM roles in academia and industry

17. Underrepresented groups in STEM should not have to justify their right to exist in this space. Inclusion of people from all backgrounds in the STEM space should be viewed as a matter of course. Recruitment and advancement should follow transparently from talent and accomplishment, without bringing in irrelevant criteria. Fixing systems within academia that disadvantage underrepresented groups will have huge impact across the board.

18. Diversity benefits everyone, and is everyone’s responsibility. It is not the responsibility of underrepresented groups to solve their own oppression. There is a significant resource demand for those who are in STEM from underrepresented backgrounds to take on unpaid labour to address ED&I issues. This detracts from time they have to focus on work, and reduces their competitiveness compared to others on the job market. It is also mentally and emotionally draining for them to deal with issues relating to their own identities rather than focusing on their science, and is perhaps a deterrent in itself from staying and progressing to higher-level positions.

19. Research, innovation, teaching, STEM, and society as a whole will suffer if STEM lacks diversity. There will be less potential for high-quality research and innovation, and more limited solutions for local, national and global scientific issues. Scientific problems of today require a global perspective and an interdisciplinary approach. The global scale of our scientific questions require solutions made up of diverse teams that can look at complex issues from a wide range of perspectives and in new ways, keeping in mind inclusive design practices. Advances in science have significant implications for other sectors and disciplines. For example, developments with artificial intelligence (AI) may have ethical implications as well as legal, political, and economic ones, and we have to work collaboratively with experts across fields. There is potential to replicate our existing biases into new technology. Well known examples include driverless cars having more trouble detecting darker skin tones, AI recruitment systems that continue to favour men based on existing data patterns, and a recent November 2021 study from the University of Oxford finding that AI systems developed to detect skin cancer run the risk of being less accurate or effective when used for people with darker skin.[23]
• Similarly, in medical research the current Eurocentric approach to medical advancement is no longer fit for purpose in a global society. COVID-19 disproportionately effects BAME communities, which are not proportionally represented in clinical trials. Drug studies show that trial participants are overwhelmingly white despite trials taking place in more diverse countries. BAME communities on the whole are less likely to participate in medical research despite evidence to show they are willing. People with disabilities are largely absent from research unrelated to their particular disability, but are statistically more likely to experience poor health outcomes and more likely to have chronic health conditions. It has been shown again and again that diverse research groups produce higher quality medical research, with groups that are gender diverse and ethnically diverse having better innovative capacity.[24][25]

20. Diversifying and decolonising STEM subjects. Not only do we need to diversify the STEM population, we should broaden the ways in which we understand scientific knowledge production to help us come up with a wider range of solutions and approaches to our global scientific problems. From understanding the history (and troubling continuation) of scientific racism in biology – with racial categorisation, eugenics and human zoos from the 19th century – to studying the ways that our deep-rooted biases are manifested in digital technologies, artificial intelligence, and health inequalities and medical education; the issues are complex and wide-ranging. There has been work in the area of decolonising STEM curriculum to acknowledge the legacies of colonialism and how a Eurocentric frame has influenced what is being taught, how it is being taught, and critically what is missing. Beyond the curriculum, the colonial legacies influence our very ideas about what constitutes ‘science’ and who ‘scientists’ are. Historical efforts of British scientists (many from Oxford) to extract and exploit resources, specimens and knowledge from colonised countries and people, intertwine science with imperialism and notions of white supremacy, which devalue indigenous knowledge. ‘Science’ and ‘modernity’, and ideas of scientific ‘advancement’ or ‘progress’ have been given a ‘Western’ frame, to the exclusion of other forms of knowledge. History has shown us many examples of instances where contributions from women and people who are racially and ethnically minoritised have been ignored or appropriated. We should expand our understanding of these issues and add to the curricula throughout the UK education system, in order to highlight the key roles played by people from underrepresented groups, the plethora of methods that can be used to advance scientific knowledge across the world, and the importance of diverse and inclusive teams and practices.

21. Disconnect with the wider public. Less diversity within STEM academia and industry will create a disconnect with the rest of the population, and risk delegitimising scientific knowledge. Engaging the public with science is more effective when people see themselves reflected in those doing the work/research rather than merely feeling like subjects to be studied or completely detached from the work. Particularly for fields with relevance to climate change, public health, and conservation, this is vital. There is already a growing minority eschewing science and ‘experts’, and this needs to be vigorously fought.

22. Perpetuating the status quo. Without diverse STEM role models, mentors and sponsors, and a more welcoming and supportive environment in STEM, future generations will be impacted and our existing make-up privileging narrow groups will be perpetuated. Our institutions need to work together, otherwise the same stereotypes and restrictive narratives of who scientists are (i.e., older white men), and what they do (i.e., work in labs with white lab coats and chemicals, or isolated, alone on a laptop) will continue.

Topic 4: what has been done to address underrepresentation of particular groups in STEM roles

23. ED&I action plans. Our activity encompasses a mix of policies, practices and programmes at the central University and divisional levels, and in our departments. This submission focuses on activity within our Division and in our academic departments. We aim to enable underrepresented groups to succeed in higher education, while guarding against deficit model thinking that focuses on ‘fixing’ people rather than systems. We have divisional and departmental action plans, including broad ED&I plans covering many topics, and some departmental gender-specific and race-specific plans. These plans are as detailed and measurable as possible, with clarity over timeframes and responsibility.

24. Embed ED&I in our governance and committee structures to ensure progress and accountability. In order to ensure wider consultation, gather input, and effectively monitor and evaluate activity, we embed ED&I principles and activity throughout our divisional structures in MPLS and MSD. ED&I is a standing item in our senior divisional committees, and we have a divisional ED&I Steering Group chaired by our Associate Head for People. Each of our academic departments have an ED&I committee with academic and administrative leads. Our divisions are represented on central University ED&I related committees and working groups. We work towards making these committees as diverse as possible so that underrepresented groups are included and an integral part of divisional and departmental decision-making.

25. Building networks and capacity. Staff and students who are in the minority in their departments and/or research group often look to networks, societies, and particular spaces to connect with others similar to them.
• ED&I Fellows scheme.[26] To guard against isolation and build a strong network of engaged members across the Division keen to advance ED&I, we developed an ED&I Fellow scheme of representatives from each of our departments to champion ED&I and help progress our divisional action plan.
• BIPOC STEM Network.[27] We have provided administrative and funding support to a new BIPOC STEM Network formed by DPhils students in MPLS and MSD to create a supportive and inclusive space for researchers, academics, and postgraduates that identify as BIPOC (Black, Indigenous and People of Colour) or BAME (Black, Asian and Minority Ethnic).
• Mentorship programmes. MPLS and MSD colleagues have joined the University’s Pivot mentoring scheme for BME staff members. There are many departmental mentoring schemes for students and staff across MPLS and MSD, and MSD operates an annual divisional mentoring circles scheme. The BIPOC STEM Network is currently developing a mentoring programme.
• Departmental level societies for underrepresented groups, including Women in Computer Science Society, Women in Engineering, Women in Materials, Women in Physics, the Mirzakhani Society (Oxford Mathematics' society for female undergraduates named after Maryam Mirzakhani, the first woman to win a Fields Medal), Oxford Women in Chemistry, and the Biology LGBTQ+ Society.
• University ED&I networks. The division supports university staff networks such as the BME Staff Network, Disabled Staff Network, LGBT+ Network, Allies and Role Models.
• Faith-based spaces. Provision of a multi-faith prayer room in the science area to ensure that are staff and students with different religious beliefs are able to practice their faith in the workplace.
26. Funding. We provide small grants for local departmental initiatives that have potential for wider divisional impact through our MPLS ED&I Fund. The Fund provides resources for new, innovative projects, and helps to sustain existing ones that have shown positive impact. This complements the University’s larger Diversity Fund scheme that enables cross-divisional, University-wide innovative ED&I activity. The following projects are examples of successful applications:

• Mfano-Africa Mentorship Programme[28]: African applicants undertake an 8-week internship with Oxford-based mentors to help build a career in maths.
• ED&I Reverse Tutorials: graduate students deliver reverse tutorials on ED&I topics to current tutorial staff to encourage them to be more actively inclusive.
• Addressing bias in graduate applications: a joint submission from Chemistry and Pharmacology to develop redaction software, which evaluates admissions processes.
• Tackling colonial attitudes towards fieldwork: Biology are developing a training course, which enables staff and students to address colonial attitudes in fieldwork.

27. Athena Swan commitment. Athena Swan has been a transformational programme for addressing underrepresentation of women. Each of our academic departments hold awards. There is no similar scheme for other groups that has had the same impact. This award has recently been expanded to encompass people of all genders and includes consideration of intersectionality in departmental culture.

28. Visible role models. Seeing others with similar backgrounds and experiences to yourself throughout STEM areas and across levels of seniority is vital to inspiring, attracting, and retaining talent. Our departments use their visual spaces (both online and physical spaces) to highlight diverse STEM role models. Here are some representative projects.
• Beyond Boundaries science inspired art competition and exhibit. Oxfordshire state school students in Years 5-8 create art based on profiles of Black, Asian, Minority Ethnic (BAME) researchers in MPLS who shared their personal/career stories and their current research. An awards ceremony and public exhibition is held in the Oxford University Museum of Natural History; winners and runners up receive prize money, and all artwork is showcased online.[29]
• Women in science videos. MSD produced a series of online videos from women in medical science interviewed in 2014/15, discussing many topics, including their careers, the culture of science, publishing, obtaining fellowship funding, having a mentor and Athena Swan. MPLS contributed an update to this project in 2017 to showcase women scientists in the mathematical, physical, and life sciences. The website and videos were available online until 2020.
• 100 Women of Medical Sciences. As part of the women’s Centenary project, MSD asked 100 diverse women across the division to take part in a project to showcase diversity of roles women now hold, each reflecting on their personal journeys, their place in medical sciences and their vision for the next 100 years.[30]
• BIPOC STEM Network videos. This network is working in collaboration with MPLS to produce a series of videos showcasing the research of BAME and BIPOC staff and students within MPLS and MSD. The first two in the series are available to view online, with further videos being planned for 2022.[31]
• Oxford University’s Diversifying Portraiture project. This University-wide initiative involved (1) cataloguing existing portraits from around the University that highlight the range of pioneering figures whose achievements over the centuries have challenged the stereotypes of their time, and (2) selecting current and former staff and students to sit for portraits that became part of a wider exhibit and now permanently displayed in the Examinations School. Current and former MPLS staff and students were selected for portraits.
29. Training and learning opportunities. ED&I related workshops, training, and online provision are delivered by MPLS ED&I team members, colleagues in MPLS and the wider University, and by external providers to offer a range of online and face-to-face ED&I learning opportunities. MPLS delivers sessions on topics such as implicit bias, responsible bystander, allyship (both targeted and intersectional), anti-racism, disability, mental health, neurodiversity and LGBTQ+ role models and allies. We aim to provide a variety of learning opportunities for staff and students within the Division as standalone ED&I sessions, and embedded within our researcher development and public engagement with research provision.

30. ED&I in enterprise. Through our RisingWISE programme, created in partnership with the University of Cambridge, we support and enable enterprising women and nonbinary people in Oxford and Cambridge, and connect them with industry mentors, facilitators, and participants to foster greater understanding of the different sectors and encourage collaboration. This is aimed at early career researchers and is now running for its third year.[32] A similar SeedWISE programme has been developed for PhD/DPhil students.[33]

31. University’s Race Equality Task Force. A University-wide task force (with MPLS representation) launched in December 2020, with the aim to develop a strategy and business plan, identifying the specific steps and practices needed to accelerate change and promote racial equality across the collegiate University. A draft of the strategy document was open for internal consultation in November and December 2021. The Task Force is currently reviewing the results from the consultation, and will finalise the strategy, action plan and business plan for Council approval by the end of the 2021/22 academic year.

32. Support for carers. The University provides excellent childcare provision, and access to a ‘Returning Carers Fund’ to support academics and researchers aiming to re-establish their research careers after a period of leave. Several MPLS departments are looking to supplement this with enhanced job cover before and after a period of parental leave, so that momentum is not lost and support continues during the return-to-work phase. Additionally, MPLS, in collaboration with Childcare Services and MSD, created informative workshops for those returning from career breaks due to parental leave, and networking opportunities for other carers.

33. Recruitment. Our departments, like others in the University, have spent significant time and effort to improve recruitment processes to attract and appoint underrepresented candidates, with particular focus on women, given Athena Swan efforts. These efforts have included ensuring diverse selection panels, revising job descriptions, creating search committees separate from selection panels to encourage underrepresented candidates to apply, providing recruitment panel training, and improving the interview experience.
• All-male shortlists. For Associate Professor posts (our main academic grade), on the rare occasion that no women are shortlisted, the departmental selection committee must submit a report to the Division to explain why this is, and to seek approval to proceed further with the recruitment process. This is consistent with a similar central University requirement for Statutory Professor posts (the most senior academic level).
• ED&I statements. For a few recent academic posts, we have piloted an application process where candidates provide an ED&I statement, similar to what they would share with respect to their research and teaching, and are asked about it at interview.

34. Outreach and supporting students, and public engagement with research. All of our departments run outreach activities, and many of these target state schools and underrepresented groups. Our researchers, academics, and professional services staff regularly engage schools and students in a variety of ways in-person and online. A few of these programmes include: an ‘It All Adds Up’ conference organised by the Mathematical Institute and Statistics for girls and nonbinary students from Years 9-12[34], a Conference for Undergraduate Women in Physics UK and Ireland (CuWiP)[35] bringing together women and nonbinary undergraduate students which has now expanded to a national annual conference with different host universities, and an Oxford Women in Computer Science ’Challenge Club’ which is a year-long online workshop for Year 12 girls studying at UK state schools[36]. Multiple MSD departments also offer innovative work experience programmes with strategies to attract and support students from socioeconomically deprived areas, and state schools. Other pioneering programmes are listed here.
• Parents for STEM Futures. Given the research identified in paragraph 15, MPLS created a parent engagement programme designed to actively engage parents of primary school aged children in STEM subjects, particularly parents of girls and BAME families. The project ran from 2016-2018, and included (1) a print and digital ‘Just Add Imagination’ resource that is still available online, sharing creative STEM activities, and career stories and advice from (current and former) students and staff, and (2) a parent ambassadors scheme training parents to engage fellow parents at local primary schools. [37]
• Bridging Programme. MPLS offers a residential bridging programme to students about to begin their undergraduate studies in Physics, Chemistry and Earth Sciences[38]. This gives a taste of undergraduate life, departmental expectations, workload management, tutorials and lab experience. Furthermore, it provides a support network of contemporaries to the participants as they make the huge transition to university, which may be particularly beneficial to those coming from disadvantaged socio-economic backgrounds. Departments in MPLS also support Opportunity Oxford[39], another academic bridging programme that helps to prepare offer-holders from underrepresented backgrounds for successful student careers in a wide range of subjects.
• University’s UNIQ programme, and Foundation Oxford. MPLS participates in several programmes aimed at supporting those from disadvantaged backgrounds to make a successful transition to university life. The UNIQ course aims to help students from diverse and underrepresented backgrounds to make successful applications to be undergraduates at Oxford[40]. UNIQ+ research internships introduce undergraduate students from underrepresented backgrounds to graduate study[41]. Foundation Oxford is a year-long fully-funded foundation course beginning in 2023 for UK state-school students who have high academic potential, but have experienced severe personal disadvantage or disrupted education[42]. Alongside academic teaching, the course will develop students’ confidence and self-belief.
• Science Together. This is a community-led engagement with research programme that connects community organisations with the resource and expertise of Oxford’s universities. Each group has tailored support and is resourced to explore and produce research, engagement or other transformative activities that are grounded in community needs, priorities, perspectives and opportunities. The central ethos of this programme is to value all forms of expertise, and through doing so enable collaborative, societally-driven research and engagement.[43]

35. Scholarships. There are several scholarships created for underrepresented students in STEM within MPLS. For example, DeepMind have supported graduate student scholarships aimed at UK women, BME and those from disadvantaged areas pursuing a master’s in Computer Science.[44] Booking.com previously supported master’s scholarships for women in Statistics, Maths, and Computer Science.[45] Most recently, we collaborated with Formula 1, along with five other universities, to offer undergraduate scholarships in Engineering Science for those from underrepresented groups, including ethnic minorities, all genders, and from lower socio-economic backgrounds.[46] There are additional schemes that have been developed and others in the works.

36. Harassment advisors. The University does not tolerate any form of harassment and has established a Harassment Advisor Network. There are at least two harassment advisors (of different genders) in each MPLS department and within the divisional office who can provide support and signposting to students and staff who have experienced bullying and harassment. In MSD, a Harassment Training Network was formed. Anti-bullying and bystander training across MPLS and MSD is provided on a regular basis. 

37. Mental Health First Aiders. MPLS has invested in three cohorts of trained mental health first aiders (MHFA), across a range of professional roles from researchers, academics and professional services staff. These are supplemented by department-trained MHFAs. MHFAs provide confidential and non-judgemental listening support and signposting to any student or staff member who may need it, acting as a first port of call in supporting those with mental ill-health and a critical stepping-stone to clinical help. This is in addition to a large suite of University-wide activity addressing mental health and wellbeing for students and staff.

38. Awards, recognition, and celebration. The challenging, emotionally draining, but ultimately rewarding work that it takes to advance ED&I and create meaningful culture change should be recognised and celebrated. As such, we developed MPLS ED&I awards to honour passionate individuals and teams making a difference within our Division. These awards inspired the creation of the VC’s Diversity Awards. We also run a number of ED&I related events marking key national and international ED&I dates such as Black History Month, LGBT+ History Month, International Women’s Day, Mental Health Awareness week, Pride, and others. While we acknowledge that one day, week, or month is not enough, these activities help to foster an inclusive culture that recognises the contributions of underrepresented groups.

Topic 5: what could and should be done by the UK Government, UK Research and Innovation, other funding bodies, industry and academia to address the issues identified

39. No single or simple solution. Just as the underrepresented groups specified in this inquiry are incredibly broad, the possible solutions are equally varied, complex, and multi-faceted. Interventions will need to include policy and procedural changes, and new or improved programmes that are adequately resourced.

• Culture change needed. Even with all of the policy and programme changes implemented, we need real culture change on the ground at local and national levels to embed ED&I in all that we do and put inclusive behaviours into practice on a daily basis. We need to work to change research cultures that are hostile, harassing, competitive and individualistic environments to ones where everyone from any background can thrive and flourish; and those who harm others are held accountable.
• Equivalence of effort across ED&I streams. Currently gender equality has been given the most attention, focus, and resource (particularly at the departmental level), driven by Athena Swan and the clear gender disparity across STEM. The Race Equality Charter, only awarded at the institutional level, has not had a similar level of engagement and commitment across the sector. While addressing gender equality issues should absolutely continue, we need to expand this effort to other areas.

40. Review differences in STEM subjects. As mentioned previously (see paragraph 10), different STEM subjects pose different challenges and may need tailored solutions. For example, with respect to race and ethnicity, Earth Sciences has found that their subject tends to be less racially diverse, and some barriers identified include geology A-levels being more prevalent in wealthier schools, some communities having less access to ‘outdoor pursuits’, lack of role models, and family perceptions that an Earth Sciences degree is more expensive to undertake and has less financial security or transferable skills upon graduation.[47] Additionally, evolution is often highlighted as evidence against God and religion (e.g., The God Delusion), and this can lead to people from a wide variety of religions feeling that a biology department (and other STEM departments) would not provide an environment in which they could work.
• Good practice sharing. We can also learn from departments/subjects/sectors that are doing well in their numerical representation and inclusive cultures, and consider what good practice can be replicated elsewhere. Improved sharing of good practice across the sector would be very useful.

41. Improve data, but focus on action. Government, Research Councils and our institutions need more, better quality, and consistent data for all the referenced underrepresented groups to understand the issues we are facing in deeper and nuanced ways. This should be done from undergraduate through to senior academic levels. At the same time, we need to focus on action rather than merely trying to understand ‘the problem’. As a scientific community, we should take a scientific approach to tackling ED&I in STEM, using experimentation to explore ways to tackle injustices and opening new lines of enquiry, analytical thinking and review to monitor their effects and make improvements, and strong communication skills to share our learning with the wider STEM community. Moreover, more evidence is not needed for known problems. The House of Commons Science and Technology Committee ‘Women in scientific careers’ report from January 2014 referenced above, stated, ‘Our inquiry has not uncovered any new issues on the topic of gender diversity in STEM subjects. This indicates that the problems and solutions have long been identified, yet not enough is being done to actively improve the situation’.
• However, careful, nuanced and intersectional analysis of quantitative and qualitative data is needed. For example, the finding in the APPG report that ethnic minority women are overrepresented in STEM as compared with white women does not mean there is no issue to address. Ethnic minority women (while possibly ‘overrepresented’ compared to their proportion in non-STEM sectors) are still the smallest group compared with white men, white women, and ethnic minority men. Data should be reviewed regularly to monitor change in response to any actions implemented and to enable timely interventions.
42. Embedding ED&I in Research Council processes. All grant funders should monitor and report on ED&I data for awards, and use the Athena Swan model of reviewing data, completing wide consultation, developing an action plan, and monitoring impact on their activities. All applicants for funding should be required to articulate how they have approached inclusive design in their research and team diversity, and how ED&I will be embedded in their work going forward to ensure inclusive teams, and this should be reviewed in a similar way to the research at the end of the grant. Some funders have required such statements, e.g., Wellcome Trust requesting information on “How will you contribute to a positive and inclusive research culture?”, along with impact statements, and there should be greater consistency across funders.
• Transparency and accessibility. Application processes for funding and fellowships should be as transparent as possible while maintaining confidentiality where appropriate. Selection panels for funding, recruitment, studentships, internships, and others, should be diverse and well-trained in ED&I. Application processes and recruitment systems should be as accessible as possible to applicants.
• Accountability. Academics and senior leaders with known and evidenced patterns of bullying and harassment should not be granted funding and be able to move to new institutions with a clean slate. The sector needs to address the issues of power and hierarchy that places postdocs, students, administrative staff, and more junior academics in vulnerable positions vis a vis their supervisors/managers. Conversely, those academics who have a proven track record of good practice should gain advantage in subsequent funding applications.
• Funding of individuals. Research Councils should adopt the best practices in allocating funding, in particular for individual fellowships. There is typically a default presumption of a need to move institutions in fellowship applications (i.e., applicants are required to justify why they are not moving). Treating this as the default can be off-putting for people with personal reasons for remaining in the same institute. Both STFC and BBSRC have moved away from such requirements, which we believe is moving in the right direction. Furthermore, moving to a system where funding is weighted more on scientific potential, rather than past accomplishments would mean that those people that have not had as many opportunities in their career, would not be severely disadvantaged. Along similar lines, publication record should be assessed on quality rather than quantity. ERC grants applications request listing the best 5 or 10 publications with reasoning, and this would be a good example to follow.

43. Funding. Funding decisions made regarding what STEM research should progress has a significant impact on ED&I, and therefore, such decisions need to be made with ED&I in mind.
• Team based and collaborative projects. These have the potential to be inherently more diverse and inclusive, overcome single-point failures in supervision and line management and foster a collaborative spirit to research. Therefore, introducing mechanisms to encourage applications from teams, rather than individuals, while providing commensurate resources to ensure such teams are not disadvantaged, would be welcome.
• Monitor and address trends in funding decisions. Government and funders should monitor the ED&I impact of their funding decisions. If increased funding is going to ‘hot topics’ or areas where there are significant known disparities, additional funding should be provided to also help increase diversity in those areas. For example, current trends show more funding in areas of AI and Machine Learning, which are known to have fewer women and ethnic minorities from particular groups, so in addition to providing funding for the science, resource should be granted to help address the lack of diversity in that field, from primary school level upwards. Otherwise, funders are further incentivising and perpetuating existing inequalities.
• More ED&I related STEM research. As noted in previous sections, there are many examples of where our science and technology have had ED&I implications and have perpetuated and exacerbated existing inequities (e.g., racism in AI algorithms). There should be more funding available for researchers doing ED&I related work in STEM, and cross-disciplinary projects with social sciences and humanities subjects.
• Interdisciplinary funding. The government and Research Councils can provide interdisciplinary funding in such a way that focuses on collaboration rather than competition. Funding that encourages STEM disciplines to reach across to other STEM areas, but also to completely different fields in social sciences and humanities, will enhance our research and science overall.
• International collaborations. Funding and support should be made available to enable international collaborations particularly with countries from the Global South and others where ethnic minorities in the U.K. make up the majority population.
44. Immigration considerations. If we are truly to be a global university producing world-leading research and teaching, the government’s immigration laws and policies need to help enable this so that we can bring talent from around the world to help tackle global problems. Having a ‘hostile environment’ immigration policy or pursuing an isolationist approach (such as with Brexit) will deter underrepresented groups from coming to the U.K. and sharing their skills and abilities. For those who consider coming to the U.K., needing to engage in burdensome bureaucratic processes and paying high fees for visas, NHS health surcharges, indefinite leave to remain, and citizenship, may make the move untenable. The enormous cost of bringing a family to settle in the UK to take up a job in academia (with limited salaries and job insecurity from fixed-term contracts) can discourage highly qualified candidates from accepting job offers. This impacts our ability to attract the best minds to the UK and as such will result in the UK becoming less productive in STEM, which in turns leads to more of the best people not wanting to move to (or stay in) the UK.

45. Support work across sectors. Diversity in STEM is a societal issue that needs to be addressed across the STEM sector and beyond. Collaborations within and across sectors should be supported and funded.
• Between academia and industry. Collaborations between academia and industry to tackle ED&I in STEM together should be resourced and supported. There can be a lot learned from working across STEM in academia and industry, for our institutions and for individual career development. (See paragraph 30.)
• Joining up efforts across education sector. Diversity in STEM should be considered holistically across the education sector, from primary schools, secondary schools, further and higher education. For instance, ED&I related training should be offered at every stage in an age-appropriate way for students and staff. Researchers and academics should be supported and funded to learn how to implement inclusive research design practices in their own work. Quality learning opportunities provided on a regular (rather than ad hoc or one-off) basis, with follow-up are necessary.

46. Specific recommendations. There are a whole host of other recommendations for STEM related challenges, some of which are identified below.
• Expanding positive action. Given that the law on positive action in recruitment is restrictive, causing institutions reluctant to risk legal challenge to be more cautious, positive action is currently limited to scholarships and studentships. If this could be broadened to include early career fellowships, that would help with the attrition point between postdoctoral researchers and academics.
• COVID-19 impacts. There needs to be clearer understanding of and actions put in place to mitigate the disproportionate effects of the COVID-19 pandemic on under-represented groups (see paragraph 6.) UK Institutions and STEM employers could make a formal commitment to protect diversity in light of the COVID-19 impact. One such example is the Statement of No Detriment being drafted by the Royal Society of Chemistry and Oxford’s Department of Chemistry. Similarly, Australian Universities and employers have signed a pledge to “formally monitor and report on gender equity impacts (including intersectional factors)” of the pandemic and to preserve progress[48]. 
• Childcare and other caring responsibilities. Funders have made some progress with recognising the impact of maternity leave, shared parental leave, and other childcare responsibilities on research careers. There has been less awareness with other caring responsibilities (caring for elderly and disabled parents, and disabled dependents, etc.) and childcare needs for older children, all of which were exacerbated by the COVID-19 pandemic. How carers (more broadly defined) can be supported in STEM should be explored.
• Impact of fieldwork. The need for fieldwork across some STEM disciplines (Earth Sciences and Biology in particular) can pose problems for underrepresented groups who may face real dangers and risks to their health and safety while conducting fieldwork and/or abroad in other countries (see paragraph 12). This area can benefit from sector-wide review. Stronger guidelines, a sense of being led and protected by their employer, and demonstration of good practice could help to reduce this fear of a threat to safety.
• ED&I should be embedded in the Research Concordat. Due to the hierarchical nature of academia, postdocs and researchers are in more vulnerable employment positions, often at the mercy of their manager or supervisor. Research culture needs to promote inclusive practices that protect its most vulnerable members. Additionally, more stability in funding and payment for those at early career stages is essential, and will help make the sector more appealing to people from all backgrounds.

(January 2022)