Wellbeing and work-life merge in Australian and UK academics

ABSTRACT

Academic work environments are becoming progressively more digitalised and focused on performativity and commodification, increasing the potential to force an unwanted merge of the boundary between work and non-work domains. This study aimed to explore academic wellbeing and the role played by factors related to work-life merge. Data were collected from a cross sectional survey of 605 Australian and 313 UK academics, who were found to have a short version Warwick-Edinburgh Mental Wellbeing Score of 21.47 ± 4.11 and 21.35 ± 4 respectively, which is significantly below population norms. Australian men’s scores were significantly lower than Australian women (20.7 ± .31, p = .007). Job strain was evidenced by excessive work hours, high levels of intrusive work-related thoughts, reduced physical activity and a self-perception that work-life merge adversely affected psychological and physical health, and mostly only occurred to meet work demands. Action by government education and university leaders is urgently required to identify policy and management practices that are contributing to this ongoing health concern. The establishment of national and university based advisory groups and consideration of a data warehouse to curate a public dataset on the wellbeing of staff within universities could assist in ensuring the outcomes of any action are continually assessed.

KEYWORDS:

• Academics
• wellbeing
• work-life merge
• stress
• work-life balance

Introduction

Rapid change is currently occurring in the way universities operate as a result of the COVID-19 pandemic, exacerbating pre-existing issues of increased financial stress and rationalisation of resources, along with prompting a more comprehensive move to the digitalisation of teaching (Dolton 2020; Marshman and Larkins 2020). This change impacts a sector that was already experiencing high levels of occupational stress evident across a range of countries. These are countries where market principles, higher performance-based management styles and rapidly changing work environments are more prevalent, such as those in the United Kingdom (UK), Australia, Canada, Japan, The Netherlands, Finland, Hong Kong, Germany, China and Korea (Shin and Jung 2014). As universities plan to move through and beyond COVID-19 to a new normal there is now an urgency to the debate regarding how they can continue to meet societies’ needs for education, translational research and innovation; and how management and financial principles adopted by both funding bodies and universities can best support efforts that lead to those outcomes. An important aspect of these conversations has to be the impact current and future change to tertiary sector management has on academics and their wellbeing; particularly given wellbeing is key to positive performance and productivity outcomes (Haddon 2018).

Wellbeing of academics

Concerns regarding the wellbeing of academics extend back at least two decades in the United Kingdom (UK) and Australia to research undertaken by Kinman (1998), Kinman and Jones (2008) and Winefield et al. (2003). The most recent of these studies included a finding that half the sample of 844 UK academics were experiencing a level of psychological distress that required some form of intervention (caseness), as determined by the 12 item General Health Questionnaire. This percentage was found to be unchanged since the earlier study in 1998 and was considerably higher when compared to seven other professions, whose caseness varied between 26% and 44% (Kinman and Jones 2008). Similar results were reported in Australia by Winefield et al. (2003) and showed from a sample of 3,711 that the level of caseness was 43% compared to the population norm of 12%. A comprehensive report was also published on Workplace Stress in Australian University Staff (Winefield et al. 2008) that showed autonomy, job insecurity, procedural fairness, work-home conflict and academic workload explained 11% of the variance for psychological distress amongst academics (specified by the authors as psychological strain). Whilst research continues to highlight issues with job satisfaction, work-life conflict, reduced quality of working life and occupational stress in a varied range of smaller and more specific academic samples and countries (Bell, Rajendran, and Theiler 2012; Mudrak et al. 2018; Converso et al. 2019; Dorenkamp and Ruhule 2019; Fontinha, Easton, and Van Laar 2019; Singh et al. 2019), it is important to now follow up, more than 10 years later, on how the wellbeing of academics in Australia and the UK is faring. This is particularly relevant as pressures from the increased use of digital technologies and working from home mark the rise of the phenomenon where work and life merge (Haeger and Lingham 2014; Hinsliff 2013).

Work-life merge

There is an extensive and well-established tradition of research into the relationship between work and non-work domains in the lives of individuals and the impact the interactions and balance in this relationship has on personal, interpersonal and organisational outcomes (Sirgy and Lee 2018). A range of terms, such as spill-over, blending, work-life fusion and work-life merge have been used in the literature to describe when the boundary between these domains becomes blurred. As early as 1985, Greenhaus and Beutell described spill-over as occurring when working at home resulted in compromises to the focus and attention, personal investment, time allocation and physical presence demanded by the respective domains, causing work-life conflict and diminishing the overall experience, motivation and satisfaction of the individual concerned in both work and home life.

Since early first descriptions of this phenomenon there has been increasing recognition of the negative effects associated with the absence of a border between work and family domains; with Clark (2000) describing how the blending of domains could lead to either integration and a sense of wholeness, or alternatively, the awkward juggling of conflicting demands and ‘a sort of schizophrenia about sense and purpose’ (Clark 2000, 257). Clarke’s ideas now appear to have taken on an even greater relevance in the context of contemporary societal change, technological advances and the changing nature of work and employment that was occurring even prior to the more recent impacts of COVID-19. This is evidenced by the coining of a new term by Emily White, a then Facebook executive, to describe this phenomenon as work-life merge (Hinsliff 2013).

Work-life merge is the term adopted by the current study reported here in recognition that engagement with work and personal domains, particularly in relation to discrete demarcated time and resource allocation, has become blurred or ‘merged’ for many. This merge occurs within the context of an academic work environment that now values performativity and commodification (Kenny 2018), and is increasingly reliant on digital technologies to deliver education. However, with this merge of domains comes the potential for both negative and positive effects, which may either enhance or undermine an individual’s wellbeing.

Although extrinsic factors such as work and work-life merge have the potential to affect an academic’s wellbeing, it is also recognised there are intrinsic components that are important, such as how people view their work (Seigrist and Li 2016) and how they approach the maintenance of their physical health (Tkachuk and Martin 1999). Consequently, the current study, in addition to investigating wellbeing and the impact of work-life merge related variables, also sought to examine work orientation and physical exercise behaviours amongst the participants. Specifically, the following research questions were asked:

• What are the relationships between work characteristics, work orientation, work-life merge and wellbeing amongst Australian and UK academics?

• How do Australian and UK Academics describe their lived experience of work-life merge and its impact on their wellbeing?

This paper reports on findings related to the first question and provides an important baseline measure for the wellbeing of academics pre COVID-19, along with insight into the impact of the phenomenon of work-life merge. This will be important particularly to inform policy and practice in universities as COVID-19 prevention strategies result in workforce redundancies, pressure to further extend online learning and the implementation of working from home policies.

Methods

The study used a cross sectional survey design using an online questionnaire consisting of closed, Likert and open-ended questions to collect data from academics across Australia and the UK.

Sample and recruitment

Email addresses were sought from university websites and requests seeking assistance in recruiting participants were sent to 236 Deans and Pro Vice Chancellors or Heads of Departments in 39 universities across Australia and 240 emails to 42 universities in the UK. Data from Australia and the UK were collected during 2018 in April and October/November respectively, which marks the approximate middle of the 1st semester of each country’s academic year. The emails requested the recipients to distribute to their academics an invitation to participate through links to the study information letter and the questionnaire. Responses were anonymous and IP addresses were not downloaded. Given an approximate population size of 55,000 Australian academics (Norton, Cherastidtham, and Mackey 2018), and 200,000 UK academics (Universities UK 2018), a confidence level of 95% and 5% margin of error, sample numbers of 381 for Australia and 384 for the UK were sought (Bartlett, Kotrlik, and Higgins 2001; Gill, Johnson, and Clarke 2010).

Survey instrument

The online questionnaire consisted of open-ended questions exploring participants’ lived experiences of work-life merge and its impact on wellbeing (Research Question 2), but these findings will be reported elsewhere. Additionally, closed and Likert scale items collected the data reported in this paper on demographics, work characteristics, work orientation, work-life merge, and physical and psychological wellbeing, and were designed to answer Research Question 1. They are now described in more detail below.

Operational definitions

The questionnaire provided participants with operational definitions to guide their responses. These were:

• Work activities: any activities related to your paid work as an academic;

• Personal life activities: all other things outside of paid work, such as: domestic work, family responsibilities, leisure, community activities, self-improvement/education;

• Work-life balance: the division of one's time and focus between working and personal life activities;

• Work-life merge: the overlap that occurs when work and personal life activities become integrated (eg cooking at home whilst also attending work emails).

Demographics, work characteristics and work-life merge

Demographic data collected included age, gender, marital status, dependents, education, qualifications, and employment descriptors, as reported in Table 1. Additionally, participants provided descriptive data on their workload, reasons for, and impact of, merging work and life and the impact of communication technologies on work-life merge.

Table 1. Descriptive demographic data for Australian and UK samples.

Demographic	Australia (N = 605)	UK (N = 313)
Age (years)
25–34	12.9% (n = 77)	17% (n = 53)
35–44	29.1% (n = 174)	26.3% (n = 82)
45–54	33.1% (n = 198)	32.7% (n = 102)
55–64	22.2% (n = 133)	22.1% (n = 69)
65–74	2.7% (n = 16)	1.9% (n = 6)
Female	69.2% (n = 417)	71.5% (n = 233)
Male	30% (n = 181)	28.2% (n = 88)
Doesn't identify as male/female	0.8% (n = 5)	0.3% (n = 1)
Married or domestic partnership	80% (n = 481)	73.8% (n = 231)
Not in a partnership	20% (n = 120)	26.2% (n = 82)
Dependents	67.9% (n = 411)	62.6% (n = 196)
≤ 5 years	38.9% (n = 160)	37.2% (n = 73)
6–12 years	56.2% (n = 231)	45.9% (n = 90)
13–18 years	40.6% (n = 167)	40.8% (n = 80)
Children over 18	46.9% (n = 193)	38.7% (n = 76)
Highest education reached
Bachelor’s degree	2.3% (n = 14)	3.8% (n = 12)
Grad Cert/Diploma	2.3% (n = 14)	2.9% (n = 9)
Masters	16% (n = 97)	26% (n = 81)
PhD	79% (n = 478)	67.3% (n = 210)
Current Study
Higher degree (Research)	2.3% (n = 14)	12% (n = 36)
Post grad coursework	2.3% (n = 14)	7% (21)
Professional development	16.1% (n = 97)	9% (n = 27)
Employment type
Tenured	65.1% (n = 394)	75.7% (n = 237)
Fixed term contract	29.9% (n = 181)	23.3% (n = 73)
Sessional	5.1% (n = 31)	0
Full time Fraction	77.5% (n = 469)	83.9% (n = 240)
Mean years worked in universities	12.43 +/- 8.4	13.5 +/- 9
Position
Tutor (sessional/casual)	3.1%	1.9% (n = 6)
Lecturer A	6.1% (n = 19)	26.8% (n = 84)
Lecturer B	33.7% (n = 204)	25.3% (n = 79)
Lecturer C	26.4% (n = 160)	16.3% (n = 51)
Assoc Prof	13.6% (n = 82)	9.9% (n = 31)
Professor	12.1% (n = 73)	17.9% (n = 56)
Other	4.1% (n = 25)	1.9% (n = 6)
Employment area
Applied Sciences	49.6% (n = 305)	50% (n = 157)
Arts and Humanities	8.8% (n = 54)	5.4% (n = 17)
Natural Sciences	10.4% (n = 64)	11.5% (n = 36)
Social Sciences	31% (n = 191)	31.6% (n = 99)
Other	0.2% (n = 1)	0.5% (n = 2)

Work orientation

Data on how academics view their work were collected in questions that used three vignettes based on the work of Wrzesniewski et al. (1997). Each vignette represented either a job, career or calling orientation towards their work as an academic. Briefly:

• Academic A represented a ‘Job Orientation’ and worked primarily to earn enough money to support their life outside of their job.

• Academic B represented a ‘Career Orientation’ and basically enjoys their work, but their focus is on career advancement and they do not expect to be in their current position five years from now. Instead, they plan to move on to a better, higher-level job or position.

• Academic C represented a ‘Calling Orientation’ and felt that what they did for a living was a vital part of who they were and its one of the first things they tell people about themselves. They also tend to think about work out of work hours and also work at home because they enjoy it.

Following the presentation of each vignette respondents were asked to rate, on a 5-point Likert scale, how much they felt they had in common with each.

Wellbeing

Psychological (mental) wellbeing, was measured using the Short Warwick Edinburgh Mental Well-being Scale (SWEMWBS), a shorter but psychometrically robust version of the 14 item scale and relating more to function than to feeling, with most items representing aspects of psychological and eudemonic wellbeing, and fewer covering hedonic wellbeing or affect (Stewart-Brown et al. 2011). The short version has been validated across a range of populations and in different languages and has been recommended for use as a measure of psychological wellbeing in epidemiological, intervention and evaluation studies, as well as in clinical practice (Koushede et al. 2019). Physical health and the impact work-life merge had on physical and psychological wellbeing were measured as self-perceived, and through how often intrusive personal and work thoughts were experienced and the time participants spent undertaking moderate and aerobic activity in an average week.

Analysis

A total of 760 Australian and 383 UK respondents logged on and commenced the questionnaire. Following data cleaning, 605 questionnaires from the Australian cohort and 313 UK questionnaires were available for analysis. Descriptive data are reported as mean (M) and standard deviation (SD). Pearson’s correlations were used to investigate significant relationships between the variables describing workload, work orientation, work-life merge, demographics and wellbeing. A standard multiple regression analysis was then conducted, using those variables significantly correlated with the SWEMWBS score. Effect size was calculated as recommended by Tabachnick and Fidell (2007) and all p levels lower than .05 were considered significant. Internal consistency of the SWEMWBS was assessed using Cronbach’s alpha.

Results

Participants’ demographics

In both countries there was normality in the distribution of participants’ age, gender, marital status, dependents, current study and area of employment. Data were very similar across both countries (Table 1), except for slightly more tenured, full time participants in the UK sample, and about 10% more Australian participants with a PhD. There was a higher proportion of Lecturer A and Bs (Australian equivalent) in the UK sample. The largest area represented across both samples came from the Applied Sciences, which for the purposes of this study included health and life sciences/professions, engineering and technology, agriculture and computer science; and from the Social Sciences, which comprised economics, law, political science, psychology, sociology and education. Arts and Humanities (performing and visual arts, languages, literature, philosophy, theology, history and geography) and the Natural Sciences (biology, chemistry, mathematics, physics, earth and space) each only contributed around 5–10% in each sample.

Participants’ work orientation

74.3% (AU) and 70.5% (UK) of academics responded they had either nothing at all or only a slight amount in common with Academic A (job orientation), with the proportion reducing to 54% (AU) and 56.8% (UK) in relation to Academic B (career orientation). However, 73% (AU) and 74.3% (UK) felt they had either moderate, considerable or very much in common with Academic C (calling orientation).

Work characteristics and work-life merge

Descriptive statistics for work related variables potentially impacting work-life merge and wellbeing are reported in Table 2. Findings are summarised below.

Table 2. Descriptive results for work variables related to work-life merge and wellbeing (percentages, M ± SD).

Work and wellbeing variables	Australia (April 2018)	UK (Oct/Nov 2018)
Workload
% who worked hours in excess of contract in last 2 weeks	71.4% (n = 432)	71.1% (n = 223)
Mean excess hours worked/week in last 2 weeks	18 ± 15.3 h	15.7 ± 15 h
Male - Mean excess hours worked/week in last 2 weeks	*20.6 ± 15.9 h	17.5 ± 16.3 h
Female - Mean excess hours worked/week in last 2 weeks	*16.7 ± 14.8 h	15.2 ± 14.6 h
Non identifying - Mean excess hours worked/week in last 2 weeks	Range 15-60h (n = 5)	No data (n = 1)
% who worked in excess of contract in the last 6 months	88% (n = 543)	85.9% (n = 268)
Mean excess hours worked/week in last 6 months	12.2 ± 10.3 h	12.2 ± 12.1 h
Male - Mean excess hours worked/week in last 6 months	13.4 ± 10.9	13.1 ± 15.5
Female - Mean excess hours worked /week in last 6 months	10.1 ± 10	11.1 ± 10.6
Non identifying - Mean excess hours worked/week in last 6 months	13.4 ± 9.5	No data (n = 1)
Work Flexibility	Australia	UK
% who sometimes, rarely or never were allowed the flexibility to take time-in-lieu in quieter times	67% (n = 417)	56.5% (n = 174)
% who often or always felt they had flexibility in selecting where they worked outside of scheduled teaching and student contact times.	51.9% (n = 314)	63.6% (n = 199)
% who often or always felt they had flexibility in selecting when they could work outside of scheduled teaching and student contact times.	55.2% (n = 334)	58.5% (n = 183)
Work & Workplace Preferences	Australia	UK
% who often or always preferred to have separate periods of time allocated to work and life activities	70.2% (n = 425)	70.8% (n = 219)
% who wanted to conduct their work activities balanced between the university and their home	30.7% (n = 164)	57.7% (n = 180)
% who mostly or always wanted to conduct their work at the university	27.7% (n = 168)	37.1% (n = 117)
Work-life Merge	Australia	UK
% who agreed or strongly agreed they merged life and work activities only because they believed it was necessary to meet work demands	76.2% (n = 455)	68.5% (n = 213)
% who agreed or strongly agreed they merged life and work activities because they enjoyed living this way	30.5% (n = 184)	37.5% (n = 111)
% who agreed or strongly agreed they merged life and work activities because they felt more productive this way	37.2% (n = 225)	35.7% (n = 120)
% who sometimes, often or always felt their work and personal life merged in a negative way	89.3% (n = 533)	88% (n = 273)
Work-life merge and wellbeing	Australia	UK
% who often or always experienced intrusive personal thoughts during work	13.4% (n = 81)	18.8% (n = 49)
% who often or always experienced intrusive work thoughts outside of work	69.5% (n = 420)	63.6% (n = 199)
% who felt work-life merge sometimes, often or always adversely affected their psychological health	78.7% (n = 451)	76.1% (n = 220)
% who felt work-life merge sometimes, often or always adversely affected their physical health	72% (n = 417)	63.2% (n = 182)
% who felt their wellbeing sometimes, often or always suffered because work activities merged into family or other life activities	80.4% (n = 467)	74.7% (n = 218)
Work-life merge and communication technologies	Australia	UK
% of the day spent in contact with others via communication technologies of some sort	36.2% ± 23.3	33.4% ± 22
% who often or always checked their work communications outside of work even though they didn’t really need to	70.8% (n = 428)	71.9% (n = 225)
% who agreed or strongly agreed communication technologies facilitated positive work-life merge	43.8% (n = 265)	43.7% (n = 137)
% who agreed or strongly agreed they had considered employment options outside academia in the last 12 months	54.1 (n = 327)	42.5% (n = 133)

*t(582) = 2.93, p = .003.

Workload, work flexibility and workplace preferences

Over two thirds of Australian and UK academics claimed to have worked a mean of 16–18 h/week respectively in excess of their contract in the two weeks prior to completing the questionnaire. Almost 90% estimated they worked an excess 10–12 h/week over the last six months. However when academics exceeded their contracted hours during peak periods (e.g. marking, grant deadlines) over a half (UK) to two thirds (AU) of respondents reported managers either only sometimes, rarely or never allowed them the flexibility to take time in lieu during quieter periods, with more than twice as many in Australia compared to the UK rarely or never allowing it. Almost twice as many UK than Australian respondents wanted to conduct their work activities balanced between university and their home, with almost two thirds of UK respondents and half of Australians feeling they were provided flexibility either often or always in selecting where they worked outside of scheduled teaching and student contact times. Just over a half in both samples felt they were provided flexibility of when they worked.

Work-life merge characteristics and wellbeing

Almost three quarters of both samples often or always preferred to have separate periods of time allocated to work and life activities. Despite this, similar proportions believed it was necessary to merge life and work activities to meet work demands, with only around a third in both countries saying they merged work and life activities because they enjoyed living this way or because it felt more productive. Similarly, almost 90% felt that work and life activities overlapped either sometimes, often or always in a negative way. Almost three quarters of Australian and UK academics reported intrusive work-related thoughts either often or all the time outside of work hours.

Most respondents in both countries, 80.3% (AU) and 74.6% (UK), felt their wellbeing suffered either sometimes, often or all the time because work activities merged (or overlapped) into their family or other life activities. More specifically, over three quarters of all participants reported the merge of work and personal life activities adversely affected their psychological health either sometimes, often or always, with around two thirds believing it affected their physical health. More Australian (56.9%) than UK (46.1%) participants either agreed or strongly agreed they had considered employment options outside academia in the last 12 months.

Almost half of respondents in both countries either agreed or strongly agreed communication technologies facilitated positive work-life merge with approximately one third of a typical day spent in contact with others via communication technologies of some sort. However, almost three quarters also said they often or always checked their work communications outside of work even though they didn’t really need to.

Academic wellbeing

Self-perceived physical health was rated by respondents on a 5-point Likert scale with just over half (56.7% AU; 60.2% UK) rating their physical health as good or very good. This is despite only 20.4% (AU) and 18% (UK) of participants spending a minimum of 150 min on moderate aerobic activity/week; and 19.5% (AU) and 18.5% (UK) reporting they spent more than 90 min on vigorous exercise. Most (77.4% AU and 88.2% UK) said they had not undertaken any University offered lifestyle activities or programmes with 10.8% (AU) and 9.6% (UK) reporting their university didn’t offer any.

The SWEMWBS scores for the UK and Australia were not significantly different (p = .69) at 21.35 ± 4.0 and 21.47 ± 4.11 respectively. Men (21.55 ± 4.12) and women (21.45 ± 4.13) scored similarly (p = .8) in the UK sample, however men in Australia (20.7 ± .31) had significantly lower scores (p = .007) than women (21.45 ± 2.0). Internal consistency (Cronbach’s alpha) for the scale for the Australian dataset was .86, and for the UK it was .87. Intrusive personal thoughts during work were experienced by 14% (AU) and 18.8% (UK) respondents either often or all the time during work hours.

Relationships between demographic and work characteristics, work-life merge and wellbeing

Pearson’s correlations identified the same 20 workload, work orientation, work-life merge, demographic and wellbeing variables in the Australian and UK data that were significantly associated with the SWEMWBS in addition to one extra UK variable: marital status (Table 3). No multicollinearity was found. A standard multiple regression analysis was performed to estimate the proportion of variance in psychological wellbeing scores that could be accounted for by these variables (Table 3). Checking for assumptions and normality, the Durbin Watson was 1.95 (AU) and 2.21 (UK), collinearity tolerance ranged from .57 to .89 and extreme cases accounted for 5.2% (n = 28) for Australia; and for the UK from .34 to .83 where extreme cases accounted for 4.2% (n = 10). Standardised residual histograms and P–P Plots indicated normality. Shrinkage indicates if the Australian model was derived from the population it would account for approximately 5% less variance in the outcome and 4% less for the UK model.

Table 3. Standard regression statistics for UK and Australian cohorts.

Variable	UK Un-standardised coefficients B	UK Std. Error B	UK standardised coefficients β	AU Un-standardised coefficients B	AU Std. Error B	AU standardised Coefficients β
Constant	27.541	2.589	–	25.153	1.809	–
Marital Status	−0.975	0.471	−.101*	N/A	N/A	N/A
Excess hours worked in last 6 months	0.006	0.016	.022*	−0.205	0.460	−.016
Excess hours worked in last 2 weeks	−1.236	0.617	−.101	−0.019	0.009	−.074*
Job orientation	−0.705	0.244	−.164*	−0.518	0.144	−.149**
Career orientation	−0.072	0.169	−.020	−0.226	0.118	−.066
Calling orientation	0.055	0.219	.014	0.422	0.140	.123***
Merged work-life activities because its enjoyable	0.388	0.281	.103	−0.216	0.168	−.061
Merged work-life activities because its productive	0.107	0.235	.028	0.133	0.154	.037
Merged work life activities to meet work demands	0.335	0.253	.078	0.297	0.154	.076
Checking work communications outside of work	−0.298	0.226	−.068	−0.166	0.142	−.041
Work-life activities overlap in a positive way	0.300	0.342	.056	0.144	0.240	.029
Work-life activities overlap in a negative way	−0.527	0.376	−.092	−0.153	0.241	−.030
Intrusive personal thoughts during work hours	−0.421	0.227	−.094	−0.570	0.153	−.131***
Intrusive work-related thoughts outside of work	−0.274	0.274	−.065	−0.585	0.194	−.127**
Wellbeing suffers due to work-life merge	−0.919	0.331	−.214**	−1.162	0.225	−.278***
Wellbeing enhanced due to work-life merge	0.151	0.283	.033	0.327	0.203	.078
Psychological health adversely affected by work-life merge	−0.518	0.324	−.113	−0.021	0.218	−.005
Self-perceived physical health status	0.868	0.256	.186	0.909	0.160	.228***
Physical health adversely effected by work-life merge	0.151	0.266	.034**	0.221	0.188	.054
Moderate aerobic activity	0.326	0.116	.148**	−0.002	0.081	−.001
Vigorous aerobic activity	−0.221	0.123	−.097	−0.073	0.084	−.035

*p < .05, **p < .01, ***p < .001.

Australian Results. In combination, seven factors accounted for a significant 46.2% of the variability in the Australian cohort’s mental wellbeing, R²  =  .459, adjusted R ² = .44, F(19,503)  =  22.8, p < .001; and a significant 51.92% of the variability in UK mental wellbeing, R²  =  .562 adjusted R = .52, F(22,216) = 13.38, p < .001. In Australia inversely related factors were: the number of hours of academic work in excess of contracted hours in the last 2 weeks (t(512) = −2.07, p = .039), characteristics in common with a ‘job’ work orientation (t(512) = −3.60, p < .001), experiencing intrusive personal thoughts during work hours (t(512) = −3.73, p < .001), work related thoughts outside of work (t(512) = −3.01, p = .003), and perception wellbeing suffered because work activities merged into family or other life activities (t(512)=−5.17, p < .001); whilst physical health (t(512) = 5.69, p < .001) and a ‘calling’ work orientation (t(512) = 3.02, p = .003) were positively associated. A large effect size was observed (Cohen’s f²=.85).

UK Results. The UK factors inversely related to the SWEMWBS were working in excess of contract over the last six months (t(219) = −2.00, p = .046), characteristics in common with a ‘job’ work orientation (t(219) = −2.89, p = .004), the amount wellbeing was perceived to suffer because work activities merge into family or other life activities (t(219) = −2.77, p = .006), and being single (t(219) = −2.07, p = .04), whilst self-perceived physical health (t(219) = 3.40, p = .001), and the amount of moderate aerobic activity undertaken (t(219) = 2.80, p = .006) were positively related. A large effect size exceeding 1 was observed (Cohen’s f² = 1.08).

Discussion

Psychological wellbeing

The SWEMWBS scores for psychological wellbeing across both samples give significant cause for concern. Although a different instrument was used in this current study, compared to those previously conducted in UK (Kinman 1998; Kinman and Jones 2008) and Australian (Winefield et al. 2003) academics, findings are comparable in so much as the scores in the current study were also very similar across both countries, and in that they remain well below population norms. Whilst the Kinman and Winefield led research revealed approximately half the academic population studied had a level of psychological distress that required intervention, the current study shows, 10 and 20 years on, a SWEMWBS score that is well below the English population norms of 23.6 and falls in the bottom 61–80% of responses (Stewart-Brown et al. 2011; Fat et al. 2017). Although population norms using the SWEMWBS have not been undertaken for Australia, a study of self-identified Australian Gay men aged over 40 years (N = 415) scored 24.8 (Lyons, Pitts, and Grierson 2013), and in Denmark (N = 3508) and Iceland (N = 6344) the population norms were higher again at 26.4 and 25.4 respectively (Koushede et al. 2019). There were no gender differences identified in the UK academic population in this current study however the significantly lower scores of Australian male academics compared to their female counterparts was particularly worrying with their mean score classified as poor and in the bottom 20% of responses for English population norms.

Men and women have been reported to have different types and levels of psychosocial exposures at work, although the lack of consensus as to which types of exposures impact work and life stress are limited by the complexities of varying workplaces and the inability to distinguish between sex and gender in work and health research (Padkapayeva et al. 2018). However, we do know that common psychological and behavioural conditions are experienced by more females (22.3% AU, 20.7% UK) than males (17.9% AU, 13.2% UK), making the finding of a significantly lower SWEMWBS among Australian male academics one that needs to be investigated and addressed urgently.

Job demands, work-life merge and wellbeing

It is acknowledged in the literature that both intrinsic and extrinsic factors influence psychological wellbeing, and this is also a finding in the current study’s Australian and UK regression models. The strongest negative extrinsic predictive factor for both cohorts was that of the perceived impact of work-life merge, with up to three quarters believing it adversely affected their psychological and physical health. Job demands on an academic’s time are likely to play an important role in driving work-life merge, which was evident in excess hours not only worked at the peak time of mid semester but also over the longer term. The burden of excess hours worked is also on the rise not only in the proportion of academics it affects but also in the numbers of hours worked when compared to previous studies (Kinman and Jones 2008; Winefield et al. 2003).

The significantly lower wellbeing scores of men in the Australian cohort, and the significantly increased number of excess hours they worked compared to women in both countries, would seem to add weight to the negative impact of excess hours on work-life merge and wellbeing. Particularly as over two thirds to three quarters of academics in both countries indicated they mostly only chose to merge life and work activities because they believed it was necessary to meet those demands, even though most preferred to have separate periods of time allocated to work and life activities. This illustrates an increase on Kinman and Jones (2008) UK findings where only 39% expressed a desire for completely separate work and home lives.

The increased preference in the current study for a temporal separation (when the work is done) may be due to the negative impact of work-life merge perceived by the participants and be further elucidated in the qualitative data collected. However, the increased role of technology and how it facilitates work-life merge was acknowledged by participants in both cohorts, with around a third of each day spent using communication technologies and almost three quarters of participants checking work communications outside of work although they didn't really need to. The latter behaviour may point to the extension of the automaticity of what have become habitual technology-based communication behaviours in the workplace to ones that spill over to outside of work and then begin to form part of an obsessive compunction or addictive behaviour associated with workaholism, which in part can also be driven by work conditions such as increased work demands (Kim 2019).

Despite the majority of academics expressing the preference for a temporal border, only around a third mostly or always desired the physical separation of only conducting their work at the university. However, even if the implementation of more stringent temporal borders were able to be implemented by participants, spill-over can always occur from an emotional perspective (Clark 2000; Staines 1980). Emotional spill-over was evident in the Australian regression model where both personal and work-related thoughts intruded on their alternate domains, and to the extent where they were both negative predictors for wellbeing. Although, intrusive work-related thoughts were experienced by many more academics in both countries than intrusive personal thoughts.

Intrusive work-related thoughts are often seen as a symptom of high job strain, which can occur where there are high job demands and low levels of employee control (Burns, Butterworth, and Anstey 2016). The excess hours worked by academics and the high prevalence of intrusive work related thoughts in the current study are concerning, not only because they were a negative predictor for psychological wellbeing, but because job strain has been implicated in a range of economically high burden diseases such as common mental health disorders, diabetes, obesity, cardiovascular disease and poor health behaviours such as physical inactivity (Burns, Butterworth, and Anstey 2016).

The known adverse effects of job strain not only impact the individual academic’s own life but impact the health economy, workplace productivity, presenteeism and turnover (McTernan, Dollard, and LaMontagne 2013; Wang et al. 2014), all of which can only have deleterious effects on the functioning of the university and the experience of students. In relation to turnover, job strain from a high level of job demand is also associated with lower job satisfaction and higher leaving intentions, and the current study has seen similarly high proportions as were reported previously by Kinman and Jones (2008), with approximately 50% of academics considering other employment options. A comment from the qualitative data has been chosen to illustrate the reality of the impact of increased workload and job demand:

I have been trying to not check emails on weekends, and indeed we are encouraged not to. But at the same time, it's almost impossible to stay on top of the work if you don't somehow catch up, after hours. It's exhausting. I've used lots of different strategies and have performed well and been well, but I think it's finally catching up with me and I keep needing sick leave and am finding it very hard to maintain the wherewithal to keep going … .. There is no way to justify whether a workload is unreasonable or not. You just quietly go insane from the pressure. Or from trying to argue that the pressure is too high. So it's better not to bother trying to articulate it, and just try to get on with the work and adjust where you can. … .. And I still love it, I still can see myself doing good things, helping the world make better decisions with better information for more people. But I'm at a loss because I can't keep up like this.

Work orientation and wellbeing

The majority of academics in both cohorts held such a calling orientation as is expressed in the above comment (where the focus is on the enjoyment of fulfilling, socially useful work); a factor that was a positive predictor for mental wellbeing in the Australian sample. Whereas a job orientation (where the focus is on financial reward and necessity) was held by approximately a quarter of participants and was a negative predictor for mental wellbeing in both countries. These findings support the original work done by Wrzesniewski et al. (1997) on the relation of job orientation to wellbeing dimensions, where ‘calling’ respondents rated significantly higher on life and job satisfaction than both ‘job’ and ‘career’ respondents, and that they also missed significantly fewer workdays. However, universities’ work demands, combined with the neoliberal revisioning of what it is to be an academic, have the potential to erode the meaning that ‘calling’ academics find in their work, further threatening their wellbeing, and indeed their desire to remain working in the sector.

How individuals view the meaningfulness of their work is explained by Wrzesniewski, Dutton, and Debebe (2003) to be significantly affected by others in the organisation and how they value or devalue the worth of their jobs and roles. Currently, it is clear from the strikes by academics taking place immediately prior to the COVID-19 environment in the UK (Wheale and Al-Khalaf 2020) that many academics there do not currently feel the work they undertake is valued, with the University and College Union citing excessive workload, loss of autonomy, a 15% gender pay gap, increased casualisation and changes to pensions as some of the reasons for the strike. The impact of such a working environment on wellbeing has been acknowledged by the Higher Education Policy Institute (HEPI) in the UK (HEPI 2019), who have called for a national measure that provides a public dataset on the wellbeing of staff within universities. HEPI (2019, 7) believe such a dataset would allow for more detailed comparisons within the sector and to the general population, and subsequently ‘promote and understand higher education beyond that of a financial transaction’. Whilst such comprehensive data sets would certainly assist in the detailed understanding of different groups within and between universities, and their associated needs, data does already exist in this study, and numerous others that have gone before it, to indicate that action is required now if wellbeing issues are to be addressed.

Individual behaviours and wellbeing

Key features of success for improving health and wellbeing in the workplace are well documented and interventions are commonly ‘focused on addressing individual behaviour change through programmes to encourage healthy eating, physical exercise, smoking cessation and stress management’ (UCL Institute of Health Equity 2014, 4), or by providing support through employee assistance programmes for those suffering consequences associated with work pressures. However, when work conditions are eroded, as is now being experienced by academics in universities, it is difficult for such interventions to be taken advantage of, even when available. We see this exampled in the findings in this study related to physical health and exercise.

Physical health is well known as a positive predictor for psychological wellbeing (Tkachuk and Martin 1999), and this was reflected in the regression models for both countries. Despite this, physical health was perceived to be either good or very good by only just over half the participants with around two thirds believing work-life merge had an adverse effect. Physical activity is an important component of physical health (Australian Department of Health 2014), however only approximately 20% of the Australian and UK academics in this study reported participating in the recommended amount of either moderate or vigorous physical activity, which is well below that of 55% for adults in the Australian population (Australian Bureau of Statistics 2019). In addition, most participants (77.4% AU and 88.2% UK) reported they had not undertaken any workplace physical activity despite only around 10% reporting their university did not offer any. Given the excess number of hours worked that were reported by participants, combined with the long hours of sitting required for university work, and the effects of job strain on physical inactivity, the participation rate is perhaps not surprising.

Limitations

Bias is always an important consideration in cross sectional questionnaires with the potential, in this instance, for those with wellbeing and work-life merge issues being those most attracted to participate. The UK sample also fell just short of the numbers required to reach power, however reassuringly the Australian sample was large enough to be representative and the descriptive and demographic data across both countries were very similar, with the SWEMWBS, a well validated tool, yielding the same results. Some differences in the regression models across countries existed, with intrusive thoughts and a calling orientation failing to reach significance in the UK, and being single not predictive in the Australian cohort. Despite this, there were strong predictors in common in relation to work-life merge, physical health and activity, excess hours worked and having a job orientation, with a large effect size and application to the broader academic population for both models. This application however only exists in relation to the variables explored and there may be a range of other potential confounders likely to have been omitted from the questionnaire. Also, although the gender response to participation was very similar across samples, there was an over representation of women participants when compared to that existing in the Australia and UK academic populations, with 12-15% more women responding, respectively. Finally, although an operational definition was set for work-life merge for the purpose of the questionnaire, how academics in the UK and Australia experience this phenomenon has not been fully described. However, there is extensive qualitative data available from the questionnaire to address this, which will be published separately.

Conclusion and recommendations

This study shows the psychological wellbeing of academics in Australia and the UK is well below population norms and has been a neglected long-term problem that previous literature indicates has been evident for up to two decades. Issues highlighted in this study included excessive work hours, high levels of intrusive work-related thoughts, inactivity and the perception work-life merge adversely affected both psychological and self-perceived physical wellbeing. The psychological wellbeing of Australian men was of considerable concern.

Action is now required by government education and university leaders to identify and remediate policy and management practices that are contributing to this ongoing problem. This study indicates it is not enough to just encourage healthy lifestyle behaviours through wellbeing programmes when work pressures that lead to excessive work hours and work-life merge are likely to make participation difficult. We recommend both government education and university leaders convene national advisory groups to ensure policy, management and practice decisions are made with consideration to workplace health within universities. A meta-analysis of existing studies on the wellbeing of academics is required to inform an action plan that can be adopted by the entire sector, whilst also considering literature highlighting a range of factors seen as contributing factors, such as neoliberal management practices and the commodification of academic work. As a rapid change in tertiary education policies and practice is continuing to occur as a result of COVID-19, university employee wellbeing and the current factors influencing it, including work-life merge, need follow up through international collaborative research. As called for by the Higher Education Policy Institute in the UK (2019), a data warehouse is required to curate a public dataset on the wellbeing of staff within universities so the outcomes of any action can be continually assessed.

Ethics Approval

Approval to conduct the study was granted by the Murdoch University Human Research Ethics Committee (Approval number 2017/183).

Disclosure statement

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

Additional information

Funding

This work was supported by School of Health Professions Dean’s Grant Murdoch University.