Diversity and inclusion in UK Higher Education: staff perspectives on institutional representations and their reality

ABSTRACT

This paper examines staff perspectives on institutional representations of a range of areas of diversity and inclusion in a key post-compulsory education sector: that of UK Higher Education (HE). The paper focuses on comparing representational statements made by institutions with the reality of their lived experiences as perceived by staff. The paper first reviews literature around key issues of promoting and progressing in these areas, and how these areas are represented by institutions and the sector as a whole. It presents and discusses data from a survey (n = 300) to show strong agreement with representational statements as mirroring those of institutions, yet statistically significant differences between representations and the reality of lived experiences, particularly in relation to the lived experiences of staff. These data are discussed in the light of progress made by institutions, and the challenges faced in translating representations into real lived experience. Suggestions for institutions and policy makers are made to help better align reality with these representations in the UK Higher Education sector and, theoretically, for post-compulsory education sectors elsewhere.

KEYWORDS:

• Diversity
• inclusion
• Neoliberalism
• rhetoric
• UK
• Higher education

Introduction

The paper explores the ‘lived experience’ (Schwandt and Burgon 2006) of staff working in the UK Higher Education (HE) university sector, regarding how closely institutional representations of policy aspirations and claims made in relation to working conditions and Equality, Diversity and Inclusion (EDI) align with the reality of lived experience. It does this for staff perceptions of both their own reality and of staff perceptions of the reality for students. The 2003 UK Higher Education (HE) government White Papers (Department for Education and Skills DfES 2003, 2003b), aimed to widen access ‘in the interests of social justice and economic competitiveness’ (Riddell, Tinklin, and Wilson 2005, 1). More recently, the 2016 HE and Research Bill set out to ‘deliver greater competition and choice that will promote social mobility; boost productivity in the economy and ensure students and taxpayers receive value for money from their investment in higher education’ (Gov.UK, 2016, para 5). Thus, promoting competition and productivity for the UK government is linked with a socially transforming ‘aim to create a future in which anyone with the will and potential to succeed, regardless of their background, has the opportunity to transform their lives through accessing an outstanding learning experience at a UK university’ (UK Gov., Manchester 2020, para 1). Here, competitiveness leads to value for money, helping promote and ensure improved conditions and better EDI. Individual institutions reflect such aspirations (e.g. the University of Manchester (Manchester 2020)). Moreover, aspirations of equality (Pickering 2021), diversity (Thornton et al. 2010) and inclusion (Meir 2018) are a key focus in post-compulsory education research and literature. However, a key issue is whether such aspirations translate into reality when set against a context of neoliberal ideological goals of competitiveness and productivity (Ball 2012; Mark and Peters 2005). This is relevant when ideological goals emphasise aspects such as casualisation of contracts (Cardozo 2017; Raaper 2016; Shore and Davidson 2014), individual resilience (Higgins and Larner 2017), responsibilization (Bonanno 2017) and individual human capital (Becker 1993, 2011), and where inequality in an economic sense is considered a positive state (Plant 2011) and justified through economic arguments (Piketty 2020).

Ostensibly, and in line with UK legislation, such as the Equality Act 2010 (UK Gov 2010) and the ‘Equality, Diversity and Inclusion Strategy 2020–2024’ (UK Gov 2020), universities are exemplars to promote EDI and to fight discrimination; and aim to do so (Hakkola and Ropers-Huilman 2018). Such EDI encompasses areas including, but not limited to: Lesbian Gay Bisexual Trans and Queer (LGBTQ+) (Stewart, Renn, and Brazelton 2015); disabled students (Riddell, Tinklin, and Wilson 2005); women (Ward and Wolf-Wendel 2004), and Black, Asian and Minority Ethnic (BAME) (Kim and Ng 2019). Indeed, EDI has benefits for organisations economically, as they outperform those with less diverse workforces (Kochan et al. 2003) and have enhanced capacities to innovate and value employees, thereby creating greater loyalty (Ovseiko et al. 2019; Richard, Kirby, and Chadwick 2013). Nevertheless, whilst progress has been made, many reports and studies advocate the need to do more (e.g. see UUK 2020, OfS, 2019; Cardozo 2017; Morrish 2019). Indeed, many question whether such policies translate into reality in areas such as Athena SWAN (Tzanakou and Pearce 2019), BAME (Bhopal and Pitkin 2020), LGBTQ+ (Rumens 2016), and Disability (Jammaers and Williams 2021). Often cited as preventing EDI initiatives is neoliberalism (Jammaers and Williams 2021; Mark and Peters 2005). If this is so, then regardless of best intentions and aspirations, as we argue below, the context and environment universities operate in, rather than creating conditions naturally leading to positive EDI outcomes, in fact prevent the attainment of such aspirations. Indeed, the impact of economic approaches on key aspects of diversity, such as widening participation, have been argued to mean aspirations can be only rhetorical in both FE (Thorley 2004) and also HE (Archer 2007) in the UK, and thus disconnected from reality and people’s lived experiences.

This paper contributes to debates around the implementation of EDI in Higher Education (e.g. for students (Hakkola and Ropers-Huilman 2018) or for BAME (Kim and Ng 2019)) by presenting the perspectives of staff regarding how the discourses around EDI align with the reality of their lived experiences. To do this it attempts to answer the following questions: do university staff believe the representation of their individual institution’s policies towards staff and students align with generic aggregated statements from 10 institutions (8 UK institutions and 2 Australian ones); and, do staff feel such representations translate into lived experience for each individual area (e.g. LGBTQ+)? These questions are key and intertwined both with how representations of EDI translate into reality, and how they are enabled through policy. In addition, they are also key to how employees will feel towards the underpinning solidity of the representations claimed by institutions, and thereby affect their identity and alignment with the institution.

The paper attempts to answer the questions through statistical analysis of quantitative data generated from a survey aimed at staff working for UK HE institutions. These data are complemented by qualitative statements. Results show an overall uniform and heavily marked alignment regarding how generic statements of representation echo individual institutional approaches; but a comparatively uniform and marked disjunction (greater for staff than students) between how such representations translate into reality.

The paper starts by discussing lived experience, as an effective means to evaluate implementation of EDI policies and initiatives. This illustrates how lived experience provides an important lens for studying how realities of EDI are perceived in practice. Next, the method and approach to data collection and analysis is outlined and discussed. An analysis of the data, including qualitative comments, is then presented. The findings are subsequently discussed regarding the reality of HE EDI practice standing up to scrutiny through the lived experience of those closest to such practice. Conclusions then summarise the paper’s contributions, as well as highlight study limitations and areas for future research.

Lived experience

The notion of ‘lived experience’ denotes the reality of how individuals understand and interpret their worlds (Schwandt and Burgon 2006). Lived experience is critical for what Schwandt and Burgon (2006, 101) call ‘human science’ and consists of ‘studying how actors in a given social situation invest objects, events, and experiences with meaning and then reconstructing (interpreting those understandings)’, which is itself a ‘mainstay of interpretive social science.’ It thus plays a critical role by allowing an evaluation of how policies and initiatives are perceived to be effective in reality. Lived experience has been applied in a range of contexts, ranging from UK health communication professionals during the COVID-19 pandemic (Gregory, Nichols, and Underwood 2021); the role of emotional labour in line managing neurodiversity (Richards et al. 2019); teenage employees in suburban America (Besen 2006); sex workers in South Africa (Gould 2011), and in an education context, where undergraduate students in Ireland compared their experiences against institutional depictions of their time studying for an undergraduate degree (Émon and Timonen 2019).

The study of lived experience reveals how particular policies or assumptions can unfold in reality. For example, the experience of studying for an undergraduate degree often contrasts, negatively, with the marketing images of universities (Émon and Timonen 2019). Further, the approach has been used to build new theories of Human Resource Management (HRM) practices long-associated with prioritising rhetoric over reality (e.g. see Legge 1995), and increasingly being ‘irresponsible’ by failing to uphold already minimal employment law (Richards and Sang 2019). The lens of lived experience has exposed the reality of sex work, revealing acts of ‘trafficking’ and explaining why governments fail to deal with exploitation (Gould 2011). Specifically, the lived experience helps challenge assumptions about exploitation by shedding light on what happens and not what is supposed to happen. Interestingly, not all lived experience produces negative imagery; for example, research on teenage employment showed how, ‘these jobs, ironically, provide opportunities for workers to have fun and exercise their individuality, control, authority, and power’ (Besen 2006, 323). Thus, the lived experience lens represents a powerful and versatile tool in social science research, not only shedding light on a range of issues, but also providing insights critical for recommendation of changes to policy and practice. In this instance and given the importance of ‘authenticity’ for universities (Patey et al. 2022), the lived experience approach has the potential to provide the basis for improving the treatment of staff and students, leading to improved wellbeing and organisational success.

Thus, a key aim of the current research is, via the lens of the lived experience, to examine the experiences of staff working in a wide-range of UK HE institutions, all of which invest substantial resources into creating positive images of EDI. The study involves the use of an electronic survey allowing participants to compare rhetorical statements with their lived experience. The survey, largely quantitative by nature, allows measuring the gaps between rhetoric and reality, plus provides a limited, but important, opportunity for participants to elaborate on such experiences. It is the view taken in this paper, that only by exploring lived experiences, can the true state of EDI be exposed. However and more importantly, such an approach is likely to be critical in determining how UK HE can reform to go further on EDI initiatives.

Methods

The study of lived experience is associated with a range of methods, such as non-participant observation (Besen 2006), participant observation (Brannan, 2015), interviews (Richards and Sang 2019), focus groups (Riach and Loretto 2009), and surveys (Gregory, Nichols, and Underwood 2021). In the main, while lived experience is largely associated with qualitative methods, many studies pursue a mixed methods approach, opening up the possibility of applying experimental approaches leaning more towards quantitative over qualitative approaches. Thus, whilst remaining grounded in constructivist and interpretivist approaches, quantitative and statistical techniques can be used in analysis to provide what Miles, Huberman, and Saldana (2018) would call ‘magnitude’ in qualitative analysis. Such ‘magnitude’ seeks to substantiate the data gained in any qualitative approach and method. In this study, a key aim was to gain the broad and wide-ranging perspective of staff lived experiences using robust quantitative statistical analysis, but also to allow for qualitative comments (cf. Galasiński and Kozłowska 2010). The survey was ethically approved by The School of Computing ethics committee at Edinburgh Napier University on 28 April 2020, and generated anonymous quantitative data and also qualitative data requiring anonymisation before analysis and sharing of such data between the wider research team (Christians 2011). The approach to sampling involved purposive and snowball approaches (Noy 2008), and involved UK HE institutions and staff cold-emailed with requests for participation and distribution of the survey link. The survey ran for six months of the COVID-19 pandemic, from April to September 2020.

Table 1 (below) shows statements for the themes surveyed. These statements were generated by gathering from ten university websites worldwide (eight from the UK and two from Australia), their representations of policy towards staff and students in key areas of EDI, and then constructing aggregated statements. The approach to constructing these statements involved a dialogic process of four of the authors discussing the 10 institutional statements for each category and creating one statement that was considered to include all elements. The statement was then refined stylistically so it was as concise as possible. Thus, whilst this was a subjective process, as well as recognising its limitations, it involved focused discussion and verification by four of the authors. Moreover, it arrived at a statement that was then the focus of the subsequent data collection in terms of whether it accurately represented institutional policy. The aim was to measure the extent to which believed statements chimed with their own institutions’ statements, as well as their lived experience. Although data gathering was UK-focused, institutions with similarly bound neo-liberal contexts were included, given their statements appeared at least broadly the same and produced for very similar purposes. Questions asked after each statement were:

1. To what extent does this statement mirror the representations made in your own institution?

2. To what extent is this representation accurate of the reality in your institution?

Table 1. Statements examined in the survey. Each statement represents the policies, understandings and approaches to the specific group considered.

Theme	Statement
Staff overall	“We must support our remarkable staff to continuously upskill and take advantage of new pedagogies and technologies to ensure their classes are of the highest quality … We are committed to supporting staff career growth —building on the capability, leadership, skills and motivation of our staff … while creating a rewarding and engaging place for our people to work.”
Students	“Our students are at the core of everything we do, we provide excellent teaching, engagement and curricula embedded with the latest research, and this is externally recognised. We develop our students socially, culturally, and prepare them for their future careers.”
LGBTQ+	“This institution values all staff and students equally regardless of sexual orientation and is committed to supporting staff and students who identify as LGBTQ+ and against any discrimination or barriers to opportunities for LGBTQ+ staff and students”.
Gender equality and women	“This institution is committed to gender equality for all and at all levels of staff for recruitment and promotion. It follows the ATHENA Swan charter and approaches in line with key government Acts and is recognized by government and industry for its efforts.”
BAME	“We are committed to equal opportunities for all students and staff regardless of race, colour and ethnicity, for recruitment, promotion, pay equity, and are against discrimination. We have dedicated working groups to promote equality for BAME students and staff and all backgrounds.”
Disability	“The institution is committed to ensuring equality in recruitment and opportunities for all students and staff with disabilities, to improving working environments, accessibility, and any required adjustments and specialist support, and is against any unlawful discrimination.”
Carers and those with dependants	“The institution fully appreciates staff may have caring and dependants responsibilities and many policies relate to flexible working and unpaid leave. Every effort will be made to accommodate requests but they are not guaranteed or automatic and must be balanced with work obligations.”
Mental health	“In line with governments Acts and those such as ACAS, we recognise positive mental health to improve performance and poor mental health to cause stress and reduce effectiveness. We encourage openness and confidentiality and provide services to help.”

Question 2 was subdivided into ‘reality in your institution for staff’ and ‘reality in your institutions for students’ (i.e. two questions) for the following themes: ‘LGBTQ+’, ‘Gender equality and women’, ‘BAME’, ‘Disability’, ‘Carers and those with dependants’, and ‘Mental health’. Furthermore, for the category ‘Gender equality and women’, the students’ question was reworded to ‘To what extent is this representation accurate of the reality (excluding aspects related to promotion and ATHENA Swan) in your institution for students?’ The survey was primarily designed to capture quantitative data, used for the statistical analysis below, but also yielded qualitative data, through options to provide additional comments. The aim of such an approach was to allow a statistical analysis of the quantitative data, and accompany this with qualitative data, providing depth to statistical accounts of lived experience.

Statistical analysis methods

The quantitative data were subject to two statistical analysis methods, via the Statistical Package for Social Sciences (SPSS) v.27 software.

The first method, one-way Analysis of Variance (one-way ANOVA), allowed the identification of whether significant differences in responses occurred between different participant groups. One-way ANOVA is based on the assumption that each sample is an independent random sample, and the distribution of the response variable follows a normal distribution (i.e. variance is homogeneous) (Field 2009). The first hypothesis was met, as comparisons were made between independent participant groups, whilst the homogeneity of variances hypothesis was tested using Levene’s test, with a confidence interval set to 95%. Homogeneity of variance was met for all cases.

The second method used, the dependent t-test, compares means between two related groups on the same continuous, dependent variables (Field 2009). This test allowed comparing the ‘representation’ and ‘reality’ answers given by participants. Four assumptions should be met for the t-test results to be valid (Field 2009): 1) The dependent variable should be measured on a continuous scale, using interval or ratio levels (this applies to the semantic scales used here); 2) The independent variable should consist of two categorical ‘related groups’ or ‘matched pairs’, i.e. the same participants should be present in both groups (which is the case here); 3) There should be no significant outliers in the differences between the two related groups, i.e. there should be no single data points that do not follow the usual pattern; in this study, all the ‘Not sure’ answers were removed from the t-test analysis to avoid such issues; 4) The distribution of the differences in the dependent variable between the two related groups should be approximately normally distributed. The dependent t-test requires only approximately normal data because it is quite ‘robust’ to violations of normality, i.e. the assumption can be a little violated and still provide valid results. Normality was tested in SPSS by comparing data histograms with normality curves, which showed all the differences in the dependent variables were approximately normally distributed.

For both the one-way ANOVA and the dependent t-test, results were considered statistically significant for p < 0.05 and p < 0.01, where p is the significance value. Furthermore, effect sizes were computed by using eta squared (η²) for one-way ANOVA and Cohen’s d for the t-test, when results were found statistically significant. Such effect sizes can be categorised as small (0.2), medium (0.5) or large (0.8), using Cohen’s scale (Cohen 1988). It is important to note small effect sizes indicate differences observed are trivial, even if statistically significant. Therefore, it is critical to report effect sizes (η² or d) in addition to significance values (p) for one -way ANOVA results.

Participants

A total of 300 participants responded to the survey. Variation in participants was ensured through representation of different university types, age groups, and reasonable variation in gender and ethnic groups. For the purpose of transparency, such details are analysed and detailed below.

Participants were almost evenly split between Post-1992 universities (48%) and 1994 (6.7%)/Russell Group (26.1%)/Plate Glass (6.6%)/and Other (12.4%) universities (totalling 52%), indicating good split between contemporary and traditional incarnations of UK universities.¹

University characterisation did not affect participants’ rating for most answers (20 out of 22), with the exception of statistically significant differences observed for the rating of ‘disability reality for students’ (F(4,239) = 5.04, p = 0.001, η² = 0.078) and for the rating of ‘mental health reality for students’ (F(4,250) = 6.13, p = 0.000, η² = 0.089). The highest rating of ‘disability reality for students’ was given by Post-1992 participants (M = 3.72, SD = 0.96), whilst the lowest rating was given by Russell Group participants (M = 3.00, SD = 1.08) (see Section 5.1 for the ratings’ meaning). Similarly, the highest rating for ‘mental health reality for students’ was given by Post 1992 participants (M = 3.75, SD = 0.93), with the lowest rating provided by Russell Group participants (M = 3.09, SD = 1.08) and 1994 (M = 3.07, SD = 0.96). However, the effect sizes η² were very small (<<0.2), so the practical implications of such findings are limited.

The vast majority of participants fell into the 30–59 age bracket (81%), with the remainder split between 20–29 (8.3%) and 60 and above (9.0%) age brackets, and 1.7% selected the ‘prefer not to say’ option. No statistically significant differences were found in all ratings given by different age groups.

The split between male and female participants was as follows: ‘female’ at 55.9%; ‘male’ at 37.8%; ‘prefer to self-describe as … .’ at 4.3% and ‘prefer not to say’ at 2.0%. Statistical analysis of differences between male and female participant ratings was performed (other participants being excluded due to low numbers which would have limited the statistical relevance of findings). Results showed no or little differences for most answers (19 out of 22), and cases where differences were statistically significant have little practical implications due to the very small effect sizes η² (<<0.2), i.e. such differences can be disregarded.

Most participants were ‘White British/English/Welsh/Scottish/Irish/Traveller community’ (64.3%), with the remaining 35.7% as ‘Other’ (17.3%); Asian/Asian British (5.7%); Black/African/Caribbean/Black British (5.0%); Chinese (3.3%); Prefer not to say (3.0%); and Arab (1.3%). The split allowed for meaningful statistical comparisons between groups and all other ethnicities. A statistical analysis of differences including all ethnicity groups was not carried out. This was due to low numbers in all but one group, and if done, would have limited the statistical relevance of the findings. One-way ANOVA between ‘White British’ participants and all other participants showed no differences for most answers (17 out of 22), and cases where differences were statistically significant have little practical implications due to the very small effect sizes η² (<<0.2), i.e. such differences can also be disregarded.

Overall, the sample could have been more representative of issues surrounding EDI covered in the research, yet the similarity in responses across a varied sample suggests the data gathered is sufficient to generate important exploratory findings on lived experience of EDI in the UK HE sector.

Results

Quantitative analysis of results

Table 2 summarises the ratings obtained across all questions, whilst Table 3 shows the percentage agreements with the statements. To interpret the data correctly, it should be noted the following values were attributed to the word ratings:

Table 2. Mean ratings, M, of statements and standard deviation, SD, given in parentheses.

Theme statement	Representation	Reality for staff	Reality for student
Staff overall	3.83 (SD=1.03)	2.87 (SD=0.88)	N/A
Students	4.24 (SD=0.90)	N/A	3.47 (SD=0.92)
LGBTQ+	4.32 (SD=0.77)	3.61 (SD=1.01)	3.67 (SD=0.92)
Gender equality and women	4.20 (SD=0.94)	3.23 (SD=1.01)	3.52 (SD=0.95)
BAME	4.01 (SD=1.01)	3.05 (SD=1.12)	3.22 (SD=1.10)
Disability	4.23 (SD=0.84)	3.24 (SD=1.13)	3.47 (SD=1.06)
Carers and those with dependants	4.05 (SD=0.87)	3.36 (SD=0.97)	3.25 (SD=1.04)
Mental health	3.96 (SD=0.94)	3.03 (SD=1.15)	3.45 (SD=1.03)

Table 3. Percentage agreement with statements and average percentage agreement across all statements (bottom row).

Theme statement	Representation		Reality for staff		Reality for student
	(a)	(b)	(a)	(b)	(a)	(b)
Staff overall	84%	64%	70%	21%	N/A	N/A
Students	96%	78%	N/A	N/A	84%	46%
LGBTQ+	93%	82%	73%	51%	70%	49%
Gender equality and women	92%	79%	72%	40%	70%	47%
BAME	89%	71%	59%	33%	60%	35%
Disability	96%	80%	65%	39%	68%	44%
Carers and those with dependants	82%	65%	67%	36%	51%	29%
Mental health	89%	71%	63%	36%	72%	43%
Average percentage agreement	90%	74%	67%	37%	68%	42%

Note: For ‘representation’ the percentages were calculated grouping: (a) ‘Somewhat similar’ with ‘similar’ and ‘Extremely similar’; (b) ‘similar’ with ‘Extremely similar’. For ‘reality’ the percentages were calculated grouping: (a) ‘Somewhat accurate’ with ‘accurate’ and ‘Extremely accurate’; (b) ‘accurate’ with ‘Extremely accurate’.

Extremely different/Extremely inaccurate: 1

Different/Inaccurate: 2

Somewhat similar/Somewhat accurate: 3

Similar/Accurate: 4

Extremely similar/Extremely accurate: 5

Therefore, the higher the rating value, the higher the similarity (for ‘perception’) or accuracy (for ‘reality’, i.e. lived experience) of the statement. It should also be noted all ‘Not sure’ answers were excluded from the mean rating analysis (Table 2).

Table 2 shows the ‘lived experience’ to variously contrast with ‘representation’ ratings, i.e. the statements were always better descriptors of representation than reality. Furthermore, in all but one case, reality ratings ‘for staff’ were always lower than reality ratings ‘for students’. This provided important evidence to suggest such statements appear generally more applicable to students than staff. Regarding ‘reality’ ratings, the lowest mean ratings were obtained for ‘Reality for staff’ for the ‘Staff overall’ statement and the ‘Mental health’ statement (values of 2.87 and 3.03 respectively). Comparatively, the highest mean ‘reality’ ratings were obtained for ‘Reality for students’ for the ‘LGBTQ+’ statement and the ‘Gender equality and women’ statements (values of 3.67 and 3.52 respectively).

Columns (a) in Table 3 show percentage agreements with statements when grouping ‘Somewhat similar/accurate’ with ‘Similar/Accurate’ and ‘Extremely similar/accurate’, again highlighting a drop in agreement for ‘reality’ compared to ‘representation’. The average percentage agreements show a drop of 23% for reality ratings ‘for staff’ and 22% for reality ratings ‘for students’. It is, however, worth noting average ‘reality’ agreements are still high, as about two-thirds of answers fell in categories ‘Somewhat accurate’, ‘Accurate’ or ‘Extremely accurate’. That said, the ‘reality’ percentages drop significantly when the ‘Somewhat accurate’ answers are removed (columns (b) in Table 3), agreement dropping to around 40% on average. Particularly noticeable is the very low ‘reality for staff’ percentage of 21% obtained for the ‘Staff overall’ statement, i.e. only 21% of participants found statements to be either ‘Accurate’ or ‘Extremely accurate’ for staff. The complete survey data is given in Figure 1 and statistical analysis results are given in Tables 4 and 5.

Figure 1. Survey answers (percentages). Legend: 1 = Extremely different/Extremely inaccurate; 2 = different/inaccurate; 3 = Somewhat similar/Somewhat accurate; 4 = similar/accurate; 5 = Extremely similar/Extremely accurate; 6 = not sure.

Table 4. T-test results comparing ‘representation’ vs. ‘reality’ for staff and for students.

Theme statement	Representation vs reality for staff	Representation vs reality for students
Staff overall	t(287) = 15.87, p = 0.000, d = 0.935	N/A
Students	N/A	t(282) = 14.31, p = 0.000, d = 0.850
LGBTQ+	t(244) = 11.70, p = 0.000, d = 0.747	t(224) = 10.29, p = 0.000, d = 0.686
Gender equality and women	t(280) = 15.47, p = 0.000, d = 0.923	t(245) = 10.04, p = 0.000, d = 0.640
BAME	t(255) = 13.80, p = 0.000, d = 0.862	t(233) = 11.46, p = 0.000, d = 0.749
Disability	t(256) = 14.44, p = 0.000, d = 0.901	t(243) = 11.65, p = 0.000, d = 0.746
Carers and those with dependants	t(231) = 10.75, p = 0.000, d = 0.706	t(187) = 9.26, p = 0.000, d = 0.676
Mental health	t(269) = 13.06, p = 0.000, d = 0.795	t(247) = 7.72, p = 0.000, d = 0.490

Table 5. T-test results comparing ‘reality’ for staff vs. ‘reality’ for students.

Theme statement	Reality staff vs students
LGBTQ+	t(220)  = -0.68, p = 0.496, d = -0.046
Gender equality and women	t(243) = -6.38, p = 0.000, d = -0.408
BAME	t(223) = -3.18, p = 0.000, d = -0.209
Disability	t(232) = -4.80, p = 0.000, d = -0.314
Carers and those with dependants	t(188)  =  2.06, p = 0.041, d = -0.150
Mental health	t(247) = -7.03, p = 0.000, d = -0.446

Table 4 indicates that differences between representation and the reality of the lived experience are statistically significant for all the statements (p < 0.01), with a large (d > 0.8) or medium-large (0.5 < d < 0.8) effect size, further highlighting the robustness of this finding. Table 5 shows that differences between reality ratings for staff and students are statistically significant (p < 0.05) in all but one case, with a small-medium effect size (0.2 < d < 0.5).

Qualitative findings

(A) ‘Staff overall’ answers

The survey generated a modest but useful collective of 103 comments. Of these, only three were positive, commonly stating personal satisfaction with opportunities available. The rest variously depicted the apparent gulf between rhetoric and reality. For example, ‘there is a lot of tokenistic statements and policies but very little enactment.’ A key explanation given for the gulf was insufficient infrastructure or support, for example, ‘the aspiration is there but the infrastructure to support it is not present.’

(B) ‘Students’ answers

There were 92 comments regarding ‘students’. Analysis suggested reality for students experience was higher than that for staff, exemplified by the following quote: ‘[It] is better [for students] than for staff’. The comments revealed themes of difference between representation and reality. The comment that follows exemplifies the superficiality of EDI representations as only accurate: ‘ … on the surface’. Similarly, the gulf was widely attributed to insufficient support, with one participant noting: ‘If student experience really was the core of what we do, we would be staffed and resourced sufficiently. This is not the case in my experience.’ Indeed, reality and lived experience was felt more closely aligned with representation for students because staff typically closed the rhetorical gap with their own efforts. As one participant noted: ‘staff are committed to making student experiences as good as they can be, but are asked to do more than they can realistically handle and have a life outside of the institution.’ This helps explain why there is a greater gulf between statements and staff experience, as staff typically have to fend for themselves regarding EDI.

(C) ‘LGBTQ+’ answers

There were 78 comments concerning EDI and LGBTQ+, with many of such comments positive. One positive quote suggests: ‘LGBTQ+ colleagues and students appear to be well included.’ Only occasionally was a gulf between lived experience and the survey statement highlighted in the qualitative comments. One comment noted more infrastructure was needed to improved EDI in relation to LGBTQ+ activities: ‘These beliefs are included in our policy and, I believe, are shared by the majority of our leadership and policy-making team; however, I think there could be more infrastructure … to support this for staff.’ A further quote drew out the difference between representation and reality suggesting: ‘there is an attempt to consider and cater for LGBTQ+ and BAME issues. But in reality, there is still much discrimination and hatred – from staff and students.’ One consideration with LGBTQ+ highlighted was the reality of the lived experience for trans staff was far from its representation, for example, ‘there is good LGB awareness and support; however, I don’t feel this is the case for trans colleagues.’ Importantly, the quotes captured the reality and contradictory nature of UK HE institutions expansion into countries with often questionable observations of human rights conventions, especially related to LGBTQ+.

(D) ‘Gender equality and women’ answers

Comments in relation to ‘Gender Equality and Women’ totalled 94. A key theme of such data was a focus on Athena SWAN. A sense of pessimism permeated such comments, especially in terms of the alleged promotion of gender equality via this UK HE-wide initiative. The following quote captures such cynicism: ‘there is a tendency to see Athena SWAN as the be-all-and-end-all of gender equality work, whereas this is not the case at all.’ Such cynicism was apparent in terms of a scheme designed to generate gender-based equality, but which has little to no impact on critical aspects of senior women’s’ careers: ‘we do well on Athena Swan but … the gender pay gap is still huge in our institution and there is no sign of this being redressed at present. Women are under-represented across all senior management levels, in strategic management and in leadership.’ As noted previously, comments indicated how EDI initiatives had more impact on students over staff, for example: ‘we are better at supporting female students than … at supporting female staff members or non-binary staff or students.’

(E) ‘BAME’ answers

The survey captured 97 comments regarding EDI and BAME, and not for the first time, suggested the statements more closely represent the views of students over staff. For BAME, staff quotes widely revealed perceptions of poor career opportunities and pay gaps, for example, ‘there is excellent support for BAME students, from inclusive pedagogical approaches and curriculum design to mentorship programmes, but not so for staff. No BAME representation at senior executive levels.’ It was widely reported how initiatives to address EDI-related issues for BAME staff were more ‘heat than light’, for example: ‘there are lots of working groups, a BAME staff network and an application in progress to the Race Equality Charter. However people’s lived experiences within the organisation doesn’t always reflect this mission statement.’ Some comments were positive, however: ‘again, not well placed to comment. However, we have more BAME staff than anywhere else I have ever worked, so I would hope that it is accurate.’ Further, a number of comments highlighted progress on EDI and BAME, but more was needed: ‘we are beginning to explore racism and decolonisation at the university but we are at the beginning of this journey.’

(F) ‘Disability’ answers

Comments about ‘disability’ totalled 98, and reflected similar themes as above. On the one hand lived experience closely matched representation, with one participant believing the statement of disability to be: ‘very accurate, it’s something the University does very well.’ On the other, negative comments prevailed revealing important gaps between representation and lived experience of disability: ‘again better for students, students will find it relatively easy to manage impairments, but with staff, HR insist they can only make temporary reasonable adjustments. We are not sure if they don’t understand equality law or deliberately defy it wherever they can.’ Noted in the finding was how: ‘students have a disability office. Meanwhile, the staff disability officer post was discontinued a few years back.’ In general, the main theme was once again that more work needs to be done: ‘they are working on this, but far from achieving it’, with problems in this area attributed to: ‘hav[ing] a legacy of high quality disability support, but this has been underfunded and resourced in recent years.’

(G) ‘Carers and those with dependants’ answers

Comments related to ‘carers and dependents’ led to the generation of 96 comments, with similar themes reoccurring in this area of EDI. While many comments were positive, for example: ‘the University’s policy in relation to staff with disability, or staff with caring responsibilities is very supportive’, negative comments prevailed, such as: ‘we bend over backwards for students (especially those with families) and staff do not get quite the same level of support.’ Here also the theme of a disconnection between representation and lived experience was widely noted, for example: ‘the rhetoric is always positive and supportive – but the reality “on the ground” is very different with ever increasing workloads and pressure. The senior management and university “admin” seem divorced from the realities of academic delivery.’ Emergent from such comments were further indicators of the gap between lived experience and disability-related statements: ‘in principle, the University supports this statement but in reality certain managers are very inflexible with providing flexible working hours to staff with caring responsibilities.’ In effect, the gap between statement and lived experiences was widely attributed to efforts and skills of line managers.

(H) ‘Mental health’ answers

The issue of ‘Mental health’ received most comments with a total of 113. The pattern of a small number of positive comments substantially outweighing appeared regarding such matters, with a closer gap noted between students over staff. For example: ‘as with disability, I think substantial progress has been made in this area’, it is ‘better for students. Staff are simply given a number for a counselling service. Nothing proactive at all on this matter’, and, ‘as far as my employer is concerned “mental health support” = tea and colouring in. Mostly we’re encouraged to develop “resilience” and told it’s our responsibility to look after our own mental health, i.e. they’re not interested.’ Participants widely reported: ‘A lot of lip service and too many “Thera-Pets” rather than serious provision for enhancing mental health and managing mental ill health.’ As such, it became similarly apparent how gaps were attributed to a lack of infrastructure, and insufficient resources. Importantly, the gap regarding mental health was attributed to a combination of rising workloads and rhetorical attempts to combat such harmful practices: ‘Nice words – but increasing workload for staff and lack of resources to support students.’ A further quote reveals how staff believed senior staff are aware of how work creates and reinforces poor mental health:

‘The institution is acutely conscious of the problem and working hard to find solutions. For staff, many of the factors are related to national problems around precarity, unrealistic workload expectations, pensions. Some steps are being taken but they are not yet sufficient. Support for students is widespread but not yet effectively joined up and the burden often lands on academics alone. Work is being done in curriculum design to improve matters but resources, funding, and additional staff are in short supply, creating more mental health pressures on the existing staff.’ Such lack of activity provoked cynicism as demonstrated as follows: ‘The issues that cause stress relate to workload and until that is addressed effectively then you will not address issues of stress. Any other initiative for mental health is just sticking plaster.’

Discussion

The results show all statements were perceived better descriptors of university ‘representation’ than the lived experience of university ‘reality’. Differences between ‘representation’ and ‘reality’ were found to be statistically significant (p = 0.000) for all 22 questions analysed. Furthermore, all effect sizes of t-test comparisons were medium or large, highlighting the robustness of this finding. The average percentage agreements with the statements given were found to be 90% for ‘representation’, 67% for ‘reality for staff’ and 68% for ‘reality for students’. Although differences between ‘representation’ and ‘reality’ were found to be statistically significant, it is worth noting ‘reality’ percentage agreements were still high (in the order of two-thirds of the answers fell in the categories ‘Somewhat accurate’, ‘Accurate’ or ‘Extremely accurate’). However, it is important to note ‘reality’ percentage agreements dropped to around 40% when the ‘Somewhat accurate’ answers were removed. The lowest mean ‘reality’ ratings were obtained for ‘Staff overall’ and ‘Mental health’ statements for staff (Table 2). In particular, only 21% of staff found the ‘Staff overall’ statement to be either ‘Accurate’ or ‘Extremely accurate’ with their lived experiences. Comparatively, the highest mean ‘reality’ ratings were obtained for the ‘LGBTQ+’ statement and the ‘Gender equality and women’ statement for students (Table 2). This suggests differences in perceptions were more marked for staff representation vs. reality (e.g. support and reward) and mental health representation vs. reality (e.g. limited recognition of mental health issues), rather than equality and diversity representation vs. reality.

Comparisons between the accuracy of the statements for staff vs. students, indicate statements were found to be significantly less accurate for staff in four instances (‘gender equality and women’, ‘BAME’, ‘disability’ and ‘mental health’), equally accurate in three instances (‘staff overall’, ‘students’ and ‘LGBTQ+’) and significantly more accurate for staff only in one instance (‘carers and their dependants’). However, effect sizes of statistically significant differences observed were only small-medium, highlighting the limited importance of such findings. Participants largely gave comparable answers, regardless of university type, age, gender, and ethnicity. The few exceptions showing statistically significant differences were unimportant, due to their very small effect sizes.

The qualitative comments confirm, but more importantly, compensate for the lack of depth in the quantitative findings. While there were occasional positive comments on all areas of EDI, most conveyed the theme of the reality of the lived experience more closely perceived to match representation for students, and of the difference in representation and reality being most marked for ‘staff overall’ and for ‘mental health’. On the one hand, the fact there was perceived to be less of a gulf between representation and reality for the areas of EDI such as LGBTQ+, Women and Gender, BAME, Disability, and for staff with Caring responsibilities, supports a view of how government legislation and action, such as the Equality Act of 2010 and the Equality Diversity and Inclusion Strategy 2020–2024 (UK Gov 2020), plus HE sector initiatives such as Athena SWAN, have had some positive impact on EDI in the UK HE sector. Yet there were numerous tensions and issues, such as the importance of line managers in both closing and widening the gaps between statements and lived experience, but perhaps most notably in relation to flexible working for those with dependents. Further key issues arose out of lack of senior level staff who are women or BAME, and the challenges faced by trans individuals. Also, there was a theme of progress being made but more needing to be done, and a theme of there being a clear difference between the representation of many areas and the reality of working conditions, particularly for ‘staff overall’ and ‘mental health’. Notably, mental health received the most optional additional comments, arguably illustrating the strength of feeling amongst staff regarding this area.

What is perhaps most notable however, is the theme of insufficient infrastructure and a lack of resources and institutional will underlying why the reality of the lived experience does not meet its representation. This was a theme that occurred within almost all areas. Clearly, this raises questions regarding resource allocation and resource generation which are beyond the scope of this paper, and possible avenues for further research opened by these findings are discussed in the Conclusions. Nevertheless, the results and the paper add to the large body of research focusing on these areas individually (e.g. Bhopal and Pitkin 2020; Jammaers and Williams 2021). They show that when surveyed and studied as a whole there are common themes arising that perhaps, if addressed, could better help institutions align the perceived differences between the representation of working in HE and the lived experience of its reality.

Conclusions

In this paper, we gathered and analysed data from a survey of UK HE staff, with the aim of exploring gaps between lived experiences and institutional portrayal of commitment to key areas of EDI. Although the staff surveyed were UK based, many UK institutions have campuses internationally, and HE has followed similar trajectories internationally as the UK is following, for example, Australia, USA, and South Africa. Results show a strong perception of alignment between the aggregated statements and how staff perceived their own institutions’ representations of EDI, but a strong, important and worrisome disconnection between these statements and the lived experience, particularly for the categories of ‘staff overall’ and ‘mental health’. On one hand this illustrates that the categories of EDI in relation to LGBTQ+, women and gender equality, BAME, disability and staff with caring responsibilities are perceived to be more closely aligned in their reality to their representation. Such results suggest that institutional policy in these specific areas is achieving more success than when policy towards EDI is viewed overall. Nevertheless, there were also common perceptions that more needs to be done, and, notably, without further resources, better infrastructure and high-level and sectoral-wide leadership, it would be difficult for institutions to achieve their vaunted aspirations.

The findings help extend the work of the wide body of research into these areas individually (e.g. Riddell, Tinklin, and Wilson 2005; Stewart, Renn, and Brazelton 2015; Ward and Wolf-Wendel 2004) and complement research illustrating the benefits of ensuring organisational EDI (Ovseiko et al. 2019; Richard, Kirby, and Chadwick 2013). The results provide an overall representation of the difficulties and challenges involved with translating the aspired claims of EDI in a number of areas, highlighting how the real lived experience can suffer from a lack of resources, poor infrastructure and leadership. Notably, the work presented adds to the existing research drawing on lived experience, by pinpointing a wide range of areas and by doing so using both quantitative and qualitative data. The work opens up areas of debate regarding how resources could be found and allocated, and the necessary infrastructure to ensure they are. It also begs the question of whether the situation has become worse over time or better, and in turn the question of the impact of recent policy initiatives, particularly the marketisation of education vaunted and promoted as creating social mobility (Gov.UK, 2016).

Ultimately, an analysis and discussion of the findings show in granular detail how there is an important, sizeable and under-researched gap between how UK universities portray themselves, and the reality of the lived experience of staff working in them, and this research, as such, fills part of that gap, but in doing so raises some important questions. It begs the question of whether or not what has been captured constitutes sectoral-wide ‘gaslighting’ or are these aspirations yet to be fulfilled? How sustainable is it for a whole sector, one reliant on domestic and international labour and ‘customers’, to exaggerate, oversell or even misrepresent working life/student ‘learning conditions’? What is the potential for long-lasting damage to employer–staff relations, especially those based around equalities, if such aspirations remain unfulfilled? What do the architects (and governments) of such visions have to say about the sizeable gap between rhetoric and reality, i.e. how do they intend to close the gaps? What role do university management play in creating and perpetuating these discrepancies between rhetoric and reality? More generally, given this research and others show that FE and HE have such a reality gap, do other sectors have similar differences? A much larger scale follow-up study would be very useful, perhaps on an international basis (HE), and domestically, drawing on a wider range of sectors. Finally, the research indicates a demand for more qualitative research, principally to explore the ‘lived experience’ of aspirational values/sector-wide gaslighting.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1. In UK HE Russell group universities consist of 24 ‘elite’ academic institutions, the 1994 group are other high ranking institutions, and plate glass institutions follow. In comparison, post-1992s are former polytechnics that became new universities after 1992 and tend to be more vocation focused.