The year 2020 will always be associated with the outbreak of the global coronavirus disease 2019 (COVID-19) pandemic. Globally, the repercussions of the pandemic infiltrated many areas within societies. Perhaps the most prominently affected were healthcare systems, which were forced to cater for patients beyond their usual capacity, and of course, the world of work. The pandemic forced institutions, including higher education institutions (HEIs), and society to rethink and restructure everyday operations and established systems. Work-from-home (WFH) became an increasingly appealing alternative to in-person or office-based operations, and therefore characteristic of the ‘new normal’ during this time (Wang et al., 2021, p. 17). The WFH arrangement was introduced in an attempt to minimise the spread of COVID-19 (Vyas & Butakhieo, 2020, p. 6). While this may seem like the easy option, employees and managers had to adapt to and cope with rapid and radical changes (Carnevale & Hatak, 2020, p. 183).

The COVID-19 pandemic impacted universities at both operational and institutional level (Paterson, 2021). Like all other institutions during the global COVID-19 pandemic, HEIs had to migrate to online and blended activities, which included teaching, learning and institutional planning (Obadire et al., 2020, p. 16778; Smith, 2020, p. 115). Hedding et al. (2020, p. 1) point out that academic staff at contact universities in South Africa had little experience or training in the pedagogy or delivery of online learning. Furthermore, historically disadvantaged universities may have fewer resources to support staff and students in flexible work and teaching arrangements (Hedding et al., 2020, p. 1). In addition, Zoom or similar communications may encourage multitasking and distractions, which may make participation difficult and may leave workers feeling frustrated, fatigued and complaining of ‘Zoom hangovers’ (Schulman, 2020, p. 384). Notwithstanding the possible adverse effects of WFH and online operations, it may build capacity in staff, as they are encouraged to try out new technology tools and systems (Marinoni et al., 2020, p. 26). As a result, work engagement may be positively or negatively impacted.

There is no single definition of work engagement; however, it can be broadly viewed as applying high levels of energy, intensity, interest and persistence in work effort (Colquitt et al., 2013, p. 165). Engagement is regarded as important, as it results in positive outcomes for both the individual and the organisation (Admasachew & Dawson, 2010, p. 71; Werner, 2021, p. 423). Engaged employees are believed to be more involved in, satisfied with and enthusiastic about their work (Amos et al., 2016, p. 124; Schaufeli & Bakker, 2004, p. 4), they are more able to deal with the demands of their job (Schaufeli & Bakker, 2004, p. 4), they strongly identify with their jobs and the organisation where they are employed, they demonstrate commitment to the organisation (Werner, 2021, p. 423) and they are more motivated and perform better (Wärnich et al., 2018, p. 260). Consequently, engaged employees benefit the organisation and result in more satisfied and loyal customers, increased profitability and enhanced quality of products and services (Wärnich et al., 2018, p. 260; Werner, 2021, p. 423). According to Rothman and Baumann (2014, p. 516), work–home and home–work interaction can influence levels of work engagement. Therefore, it is expected that working from home may influence the engagement levels of employees.

Given the background provided here, the purpose of the study was, firstly, to determine the perspectives on work engagement and WFH among employees of an HEI in South Africa during the global COVID-19 pandemic and, secondly, to ascertain the association between selected socio-demographic variables, work engagement and WFH.

Scholars, organisations and countries differ in their definitions and conceptualisations of work engagement (Bagraim, 2016, p. 125; Sun & Bunchapattanasakda, 2019, pp. 65–68). Two main approaches are used to understand work engagement. The first approach views it as the opposite of burnout. The second approach sees it as a ‘positive, fulfilling state of mind, unrelated to burnout’ (Bagraim, 2016, p. 125). In view of the latter approach, Kahn’s (1990) and Schaufeli et al.’s (2002) conceptualisations of engagement are discussed here.

Kahn (1990, 1992), who first developed the concept of personal engagement, highlights the importance of being present in one’s assigned role, as it allows for more effective role performance. Khan’s thoughts on engagement were heavily influenced by the classic sociological book The presentation of self in everyday life, written by Goffman (1961). In line with Goffman’s theory, Kahn suggests that ‘people act out momentary attachments and detachments in role performances’ (Kahn, 1990, p. 694). Kahn (1990) defines engagement as ‘the harnessing of organisation members’ selves to their work roles’ and disengagement as ‘the uncoupling of selves from work roles’ (p. 694). Therefore, Kahn’s definition focuses on ‘how people occupy and adjust to their work roles’ (Schaufeli, 2017, p. 10). According to Kahn (1990, p. 694), ‘engaged employees employ and express themselves physically, cognitively and emotionally during role performances’.

Relying on the works of Goffman (1961), Maslow (1970) and Alderfer (1972), Kahn (1990, cited by Shuck & Wollard, 2009, p. 135) suggests that the spheres of ‘meaningfulness, safety and availability’ are significant in ‘understanding why a person becomes engaged’. According to Kahn (1990), meaningfulness is the ‘feeling that one is receiving a return on investments of one’s self in role performance’ (p. 705). Meaningfulness is experienced when people feel useful and valuable and it is influenced by task characteristics (e.g. challenging, clearly demarcated, diverse, creative and somewhat autonomous), role characteristics (e.g. roles that carry status or influence) and work interactions (e.g. rewarding interactions promoting dignity and self-appreciation) (Kahn, 1990, pp. 704–708). Safety is experienced when ‘feeling able to show and employ one’s self without fear of negative consequences to self-image, status, or career’ (Kahn, 1990, p. 705). Safety is influenced by interpersonal relationships (e.g. supportive and trusting), group and intergroup dynamics, management style and process (e.g. supportive, resilient and clarifying) and role performances (consistent with organisational norms) (Kahn, 1990, pp. 708–713). Availability is regarded as the ‘sense of possessing the physical, emotional, and psychological resources necessary’ to complete one’s work (Kahn, 1990, p. 705). Availability is influenced by the depletion of physical energy and strength, the depletion of emotional energy (e.g. emotional labour is tiring), individual insecurity (e.g. about one’s work and status) and outside lives (e.g. work–family distractions).

Although Kahn presented a comprehensive theoretical model of psychological presence and although the model was extensively cited, it was critiqued because it lacked an operationalisation of the construct (Schaufeli et al., 2002, p. 74; Weidert, 2011, pp. 10–11).

Schaufeli et al. (2002, p. 74) introduced another approach to work engagement, which focuses on the positive side of workers’ well-being (i.e. human strengths and optimal functioning) instead of the opposite, which is regarded as burnout. Schaufeli et al. (2002) defined engagement as a ‘positive, fulfilling, work-related state of mind that is characterized by vigour, dedication, and absorption’ (p. 74). Vigour implies the application of high levels of energy while working, the willingness to put effort into one’s work, and mental resilience and persistence when encountering work-related challenges. Dedication implies that one is strongly involved in work because work is regarded as important and is a source of ‘enthusiasm, inspiration, pride and challenge’ (p. 74). Absorption refers to being focused on and happily occupied in one’s work, whereby one loses one’s sense of time and finds it difficult to detach (switch off) from one’s work. Based on these definitions, Schaufeli and Bakker (2004) developed the Utrecht Work Engagement Scale (UWES) that measures the three dimensions of work engagement, namely vigour, dedication and absorption. Over the years, the scale has been widely cited and used to measure the engagement levels of employees, and it was also employed in this study.

Kahn (1990) and Schaufeli et al. (2002) agree in their conceptualisations of engagement, as both comprise a ‘physical-energetic (vigour), an emotional (dedication), and a cognitive (absorption) component’ (Schaufeli, 2017, p. 10).

In addition to the aforementioned, the Job Demands-Resources (JD-R) model explains the factors that may increase or decrease levels of work engagement. The JD-R model was initially developed by Demerouti et al. (2001) to understand the antecedents of burnout (Schaufeli & Taris, 2014, p. 43). The model identified 8 job demands and 13 job resources that may contribute to burnout. According to Demerouti et al. (2001, as cited in Schaufeli & Taris, 2014), job demands are seen as ‘those physical, social, or organizational aspects of the job that require sustained physical or mental effort’ (p. 45) and when high, may result in physical and psychological costs, for example, continual tiredness or weakness and irritability. Job resources are ‘those physical, social, or organisational aspects of the job’ that may contribute to realising work goals, decreasing job demands and consequently, the physiological and psychological costs resulting from the demands and inspiring personal growth and development (Demerouti et al., 2001, as cited in Schaufeli & Taris, 2014, p. 45). Job resources include ‘feedback, job control and social support’ (Schaufeli & Taris, 2014, p. 43). In 2004, Schaufeli and Bakker introduced a revised version of the JD-R model, which not only provides explanations for burnout but also underlines the inherently motivational qualities gained from job resources that contribute to work engagement. Therefore, job resources foster work engagement by providing ‘a positive work-related state of mind’ through the satisfaction of basic needs and the realising of work goals, resulting in organisational commitment and performance (Schaufeli & Taris, 2014, p. 47).

To enhance the engagement levels of employees, Crim and Seijts (2006, as cited in Amos et al., 2016, pp. 124–126) suggest that managers incorporate the 10 Cs in working environments: Connect (manage to engage employees), Career (provide work that is meaningful, challenging and stimulating), Clarity (clear communication regarding the vision, strategy and goals of the organisation or relevant division), Convey (clarify expectations of employees and provide feedback), Congratulate (provide recognition and praise immediately and often), Contribute (create understanding of the contribution of employees’ jobs to the success of the organisation or relevant division), Control (create opportunities for employees to exert control over their work, to participate in decision making and to take ownership of problems and the solutions thereof), Collaborate (create opportunities to collaborate on individual, group and organisational goals), Credibility (strive to maintain the reputation of the organisation and demonstrate high ethical standards), and Confidence (encourage high ethical and performance standards to create confidence in the organisation).

It is evident that work engagement is a multifaceted construct; it is expected that an abrupt change in a steady work environment brought about by a crisis such as the COVID-19 pandemic may have drastic effects on the stability of an established working relationship and, consequently, the engagement levels of employees.

During the COVID-19 pandemic, WFH was introduced as an alternative work arrangement to minimise the risk of COVID-19 infection (Vyas & Butakhieo, 2020, p. 6). The International Labour Organization (ILO, 2020) refers to WFH as ‘a working arrangement in which a worker fulfils the essential responsibilities of his or her job while remaining at home, using information and communications technology (ICT)’ (p. 5). According to Edgell and Granter (2020), homeworking involves ‘undertaking paid work at home, a workplace that is considered to be and still is an atypical spatial location in modern societies’ (p. 157). The term ‘work-from-home’ is often used interchangeably with the following terms: remote work, industrial homeworking, teleworking or telecommuting and virtual work (ILO, 2020; Watson, 2017, pp. 218–219). Although the definitions of the concepts differ slightly, they all refer to flexible working practices that enable employees to work remotely, for example, at home or any other place outside of the traditional corporate office building, using technology to communicate and execute work tasks (ILO, 2020, p. 5; Vyas & Butakhieo, 2020, p. 6).

Neo-Fordist and post-Fordist thoughts on work (industrial and service) agreed that there was a trend away from the standard Fordist model (i.e. permanent full-time employment) of work towards non-standard forms of employment (i.e. impermanent, individually negotiated and spatially variable) (Edgell & Granter, 2020, p. 193). This trend was encouraged because of the demise of Fordism as a production system (Edgell & Granter, 2020, p. 194), globalisation, technological change, being responsive and competitive in the marketplace (Hutchinson, 2017, p. 195; Watson, 2017, p. 215), managerial strategies to enhance functional (task flexibility) and numerical (i.e. adjusting the labour supply) flexibility and the increased use of ICTs (Edgell & Granter, 2020, p. 148), among other factors.

Flexible working, such as WFH, has positive and negative outcomes for employees and employers (Hutchinson, 2017, p. 196). Proponents of flexible working practices argue that employees tend to prefer flexibility, as it gives them control over how, when and where they work, gives them a sense of empowerment and contributes to a better work–life balance (Hutchinson, 2017, p. 196). Therefore, employees will be more satisfied, motivated and engaged, which will result in higher levels of productivity and performance (Hutchinson, 2017, p. 196).

However, authors such as Felstead and Jewson (as cited in Edgell & Granter, 2020) argue that homeworkers are engaged in a distinctive type of struggle, regularly, regarding ‘the management of the self’ (p. 158). This includes that homeworkers have to organise the space in which they work and their working time, establish social boundaries and supervise themselves to maintain the quantity and quality of their work and enhance their occupational credibility (Edgell & Granter, 2020, p. 159). This applies to low-discretion (i.e. workers who receive low pay and experience poor working conditions) and high-discretion (i.e. workers who receive better pay with superior working conditions and who have greater access to dedicated workspaces and technologies) homeworkers (Edgell & Granter, 2020, p. 159). Felstead and Jewson (as cited in Watson, 2017, p. 219) found that homeworkers often find it challenging to establish a work routine; to manage quality, health and safety standards; and to maintain their relations with both people living in their household and people in the organisation to which they relate (Watson, 2017, p. 219). They also tend to work intensively for long hours (Edgell & Granter, 2020, p. 159). Although for employers, homeworking provides the advantage of numerical flexibility, it is not without challenges, as maintaining control over workers at a distance places a premium on trust and a manager’s skills to delegate and communicate. The technological potential of ICTs seems to be constrained by ideas of how work should be organised, through principles espoused by Taylorism and the ideal bureaucracy (Edgell & Granter, 2020, p. 159). Furthermore, it is more difficult to control and motivate workers at a distance than doing so in a face-to-face situation (Edgell & Granter, 2020, p. 160).

As a result of the strict lockdown regulations introduced by governments on a global scale in the early days of the pandemic, HEIs, as most other institutions, had to quickly migrate to online and blended activities (Obadire et al., 2020, p. 16778; Smith, 2020, p. 115). Employees were forced to work from home regardless of their personal preferences, abilities, resources, and the nature and extent of their jobs (Wang et al., 2021, p. 17). Prior to the pandemic, individuals had not had much experience of working in a remote environment, nor were institutions prepared to support practices of this nature (Wang et al., 2021, p. 17). Primary activities, including teaching, research and community outreaches, were affected, and the shift to virtual operations left employees who were responsible for critical university operations ill-equipped (Agasisti & Soncin, 2021; Piotrowski & King, 2020). Support staff who were operating face-to-face in the physical university space were potentially considered irrelevant, leaving their job security in question (Piotrowski & King, 2020).

Across sectors and organisations, challenges reported regarding the WFH structure include an increase in the experience of work–life conflict, frequent interruptions from family members, a lack of experience with the organisational climate, interruptions in concentration (Chanana, 2020, p. 1), increased work hours and workload (employees reported that they were overworked) (Pathak & Majumdar, 2020), ‘Zoom hangovers’ (Schulman, 2020, p. 384), increased parental demands because of the closure of schools and childcare services, loneliness, a lack of purpose and belonging, and health and well-being issues (Carnevale & Hatak, 2020, p. 183). In highly heteronormative societies, it can be assumed that women’s productivity in particular may have suffered, based on the added strain of childcare and home-schooling, which fell predominantly on their shoulders (Hedding et al., 2020, p. 1). Given the WFH structure, the pandemic caused a so-called double-bind on specific mothers, which may have had definite implications for work engagement during this period (Heggeness, 2020, p. 1053). Similarly, people’s marital status and familial responsibility may have exacerbated issues of gender inequality, which forced women to adopt even more domestic labour in pandemic-related circumstances (Fisher et al., 2020, p. 250). The opposite was also found to be true. In a traditional male–female (heteronormative) relationship, the fact that more men are working from home could also hold certain positive outcomes. The flexible work arrangements present them with a unique opportunity to share in childcare and domestic work (Fisher et al., 2020, p. 250).

The literature review provided a brief overview of work engagement, WFH and workplace practices in higher education within the COVID-19 context and informed the instruments used in this study.

The study used a quantitative-based cross-sectional design. A cross-sectional design is a structured observation that collects data to make inferences from a sample of a population at a single point in time (Bryman, 2012, p. 44).

The Corporate Relations and Marketing division of the university under investigation distributed the web-based survey to the target population via all the relevant e-platforms of the university. In total, 399 responses were received. Table 1 presents the biographical and demographical information of the respondents.

From Table 1 it is evident that more women (72.9%) than men (25.8%) participated in the research. The most responses were received from Campus A (75.7%), followed by Campus B (15.3%) and Campus C (8.3%). The majority of the respondents were between 30 and 59 (82.5%) years old and were married (67.4%). Eighteen per cent of the respondents had a college or university degree, 32.8% a postgraduate degree and 27.6% a PhD degree. In total, 70.7% of the respondents had worked for the university for 6 years and longer. The majority of the respondents were employed in support positions and 37.8% in academic positions. A large majority (90%) of the respondents were employed in full-time contracts. During the research period, most of the respondents (40.4%) had been working from home since the beginning of the lockdown, but came to the office when required and 24.6% who had been working from home had not returned to campus.

Confirmatory factor analyses showed that the three-factor structure of the UWES fit the sampled data adequately (see Figure 1). The measurement of vigour was supported by six items, of dedication by five items and of absorption by six items. All factor loadings were statistically significant at the 0.05 level. The standardised regression coefficients were interpreted as factor loadings. The factor loadings for vigour ranged from 0.520 to 0.865, for dedication from 0.520 to 0.939 and for absorption from 0.494 to 0.879. According to Field (2009, p. 644), in general, a factor loading with an absolute value of more than 0.3 is regarded meaningful and the significance of a factor loading will depend on the sample size. Field (2009, p. 644) suggests that a factor loading for a sample of 100 should be greater than 0.512, for 200 it should be greater than 0.364, for 300 it should be greater than 0.298, for 600 it should be greater than 0.21 and for 1000 it should be greater than 0.162. The Cronbach’s alpha coefficients for vigour, dedication and absorption were 0.8 (see Figure 1), indicating excellent reliability and internal consistency according to Field (2009, p. 675).

The goodness-of-model-fit indices (chi-square statistic divided by degrees of freedom CMIN or DF, the comparative fit index [CFI] and the root mean square error of approximation [RMSEA]) for the work engagement measurement model are presented in Table 2. From the results it is evident that the CMIN or DF (5.757) yielded a value that indicates a close fit between the measurement model and the sampled data. The value obtained for the CFI (0.885) indicates an acceptable model-data fit. However, the value obtained for RMSEA (0.109; 0.101 [low]; 0.117 [high]) indicated not a good fit between the measurement model and the sampled data.

A seven-point Likert scale was used to measure employees’ levels of engagement, ranging from 0 (never) to 6 (every day). The mean scores (see Figure 1) for Vigour (M = 4.63), Dedication (M = 4.78) and Absorption (M = 4.55) were above 4, leaning towards the positive side of the scale, although there is much room for improvement.

An exploratory factor analysis was conducted on the 30-item WFH scale measuring employees’ perceptions regarding working from home. The Kaiser–Meyer–Olkin (KMO) test yielded a value of 0.879, thus the sample size was adequate for factor analysis. According to Field (2005, p. 640), a KMO value between 0.7 and 0.8 is regarded as excellent. The p-value of Bartlett’s test of sphericity was 0.000, showing sufficient correlation between the items (see Field, 2005). Five factors (WFH challenges, Organisational support, Conducive WFH environment, Work–life balance and Social alienation) were extracted through Kaiser’s criteria (Field, 2005) that explained 57.21 of the total variance (see Table 3).

The Cronbach’s alpha coefficients for all five factors were above 0.7 (WFH challenges: α = 0.748; Organisational support: α = 0.762; Conducive WFH environment: α = 0.885; Work–life balance: α = 0.745; Social alienation: α = 0.763), showing excellent reliability and internal consistency. Eight items loaded on WFH challenges (ranging from -0.370 to -0.528), five items on Organisational support (ranging from 0.591 to 0.848), eight items on Conducive WFH environment (ranging from 0.449 to 0.823), four items on Work–life balance (ranging from -0.655 to -0.785) and five items on Social alienation (ranging from -0.374 to -0.801). The factors obtained the following mean scores: WFH challenges: M = 2.344, Organisational support: M = 4.427, Conducive WFH environment: M = 4.384, Work–life balance: M = 3.231 and Social alienation: M = 2.740. The WFH scale used a five-point Likert scale ranging from 1 (strongly disagree) to 5 (strongly agree).

T-tests were conducted to determine the association of gender, nature of employment and employment contract with work engagement and WFH; the results of the t-tests are presented in Table 4.

The results of the t-tests revealed the male (M = 4.43) and female (M = 4.69) respondents for Vigour (p = 0.019); the effect size showed a small effect (d = 0.27). Regarding nature of employment, statistically significant differences between the mean scores of academic and support staff were evident for Vigour (p = 0.016; Academic: M = 4.47; Support: M = 4.71), WFH challenges (p = 0.000; Academic: M = 2.55; Support: M = 2.22), Organisational support (p = 0.007; Academic: M = 4.31; Support: M = 4.50), Conducive WFH environment (p = 0.016; Academic: M = 4.27; Support: M = 4.45), Work–life balance (p = 0.002; Academic: M = 3.42; Support: M = 3.10) and Social alienation (p = 0.018; Academic: M = 2.88; Support: M = 2.65). The effect sizes indicated small to medium effects (d ranging from 0.244 to 0.468). Furthermore, the t-tests revealed statistically significant differences between the mean scores of respondents who were employed full-time (M = 4.59) and those who were employed on fixed-term or temporary contracts (M = 4.92) for Vigour (p = 0.044); the effect size indicated a medium effect (d = 0.349).

Analysis of variance tests were conducted to determine the association of location, marital status and WFH scenario with work engagement and WFH. The results revealed no statistically significant differences between the mean scores of the different campuses of the university and the marital status categories; the p-values for all the work engagement and WFH factors were above 0.5.

With regard to the WFH scenario, the results indicated statistically significant differences between the mean scores of the different categories for Vigour (p = 0.008), Absorption (p = 0.045), Organisational support (p = 0.000), Conducive WFH environment (p = 0.000), Work–life balance (p = 0.002) and Social alienation (p = 0.000). The results of the post hoc tests are elaborated on in the discussion section of the article.

Spearman’s rank correlation coefficient was used to measure the linear association between ordinal biographical variables and work engagement and WFH. The results are reflected in Table 5.

The results showed small to moderate positive correlations between age and Vigour (p = 0.012, r = 0.126), Dedication (p = 0.000, r = 0.182) and Absorption (p = 0.000, r = 0.220). Furthermore, small to moderate positive correlations were found between Work–life balance and age (p = 0.002, r = 0.155), highest qualification (p = 0.031, r = 0.108) and duration of employment (p = 0.000, r = 0.181). The results revealed a small to moderate positive correlation between duration working from home and Conducive WFH environment (p = 0.000, r = 0.260) and a small to moderate negative correlation with Social alienation (p = 0.010, r = -0.130).

Spearman’s rank correlation coefficient was used to measure the linear association between work engagement and WFH; the results are displayed in Table 6.