Who’s more vulnerable? A Generational Investigation of COVID-19 Perceptions effect on Organisational Citizenship Behaviours: Job Insecurity, Burnout and Job Satisfaction as Mediators


 This paper is an empirical investigation that examines a path model linking COVID-19 perceptions to organisational citizenship behaviour (OCBs) via three mediators: job insecurity, burnout, and job satisfaction. The research examines the path model invariance spanning Generations X, Y, and Z. Three countries in the Middle East and North Africa (MENA) were the focus of the study. The data was collected from a sample of employees in service companies. We used a Partial Least Square Structural Equation Modelling (PLS-SEM) to analyse the data. Our findings reveal that COVID-19 positively predicts job insecurity, which positively impacts burnout levels. Burnout negatively predicts job satisfaction. The findings established that job satisfaction positively predicts OCBs. The mediation analysis determined that job insecurity, burnout and job satisfaction convey the indirect effects of COVID-19 perceptions onto OCBs. However, our multi-group analysis revealed that the indirect effects of COVID-19 on OCBs were only valid amongst younger generations, i.e., Generation Y and Generation Z. Finally, our hypothesised model is not invariant across Generations X, Y and Z. Specifically, younger generations are substantially more vulnerable to the indirect effects of COVID-19 perceptions on their engagement in OCBs than Generation X whose job satisfaction blocks the effects of COVID-19 perceptions on OCBs.


Introduction
The causes and consequences of job insecurity have been the subject of substantial research in the last four decades since the publication of the seminal theory paper on job insecurity (Greenhalgh & Rosenblatt, 1984). The reasons for this rising interest in job insecurity research is due to the breadth of its impact: (1) the severe negative outcomes for workers anxious about their employment; (2) the explicit deleterious organisational effects. A series of meta-analyses have discovered different categories of adverse effects of job insecurity on the health of employees, their attitudes and the level of stress they experience(see multiple meta-analytic studies: Cheng & Chan, 2008;Jiang & Lavaysse, 2018;Keim et al., 2014;Sverke et al., 2002;Sverke et al., 2019). The consequences of job insecurity affect individual attitudes and behaviours, raising negative implications for the organisation. When faced with job insecurity, employees perform poorly; they engage in counterproductive workplace behaviours, making fewer voluntary contributions such as organisational citizenship behaviour (Reisel et al., 2010). Sverke et al.'s (2019) meta-analysis concluded that job-insecure employees tend to be less likely to contribute to achieving the organisational strategy. This fundamental inter-relationship between employees and employers means that negative effects have individual and organisational manifestations.
This research focuses on an aspect of job attitudes and organisational outcomes in the midst of the COVID-19 pandemic.
We seek to understand if COVID-19 in uences employee perceptions of job insecurity and how this in uences attitudes and behaviours. We ground this in both theory and evidence. We focus upon employees' contextual performance to explore the degree to which generational differences in uence discretionary organisational citizenship behaviours (OCBs). We rst validate our hypothesised model to learn how COVID-19 perceptions predict OCBs via job insecurity, given that employees have few safeguards against macro forces such as global GDP declines resulting from the pandemic. Second, we examine whether there is a signi cant in uence from generational differences among Generations X, Y, and Z. This approach has not been previously reported in the literature. Our study took place in the MENA geographical region, where the most recent COVID-19-related research has focused on the health sciences (Ibrahim et al., 2020). MENA region is particularly vulnerable to increased unemployment due to the COVID-19, which stems from economic and nancial hardships organisations are exposed to due to pandemic-induced contracting of business activities. The MENA region has long suffered one of the highest rates of youth unemployment globally (30% as of 2017)-its young people are ve times more at the risk of unemployment than their counterparts in other regions (OECD, 2020c). Therefore, the MENA offers an ideal context to study employees' attitudinal and behavioural responses to COVID-19-triggered uncertainty, notably amongst younger generations, i.e., Generation Y and Generation Z compared to their ancestors.
Our study focuses on the psychological aspects of employment in the service sector, where the pandemic crisis has severely battered employment conditions worldwide. Such precarity has been attributed to the services sector employees' limited ability to shift to the safety of remote or virtual work since their job nature requires them to interact with customers directly. It impacts customer service delivery that is particularly vulnerable when employees who engage with customers directly perceive their job security as being threatened , making the services industry an excellent context for our investigation.
Literature Review And Conceptual Model  This paper seeks to examine the association between COVID-19 perceptions and job insecurity, representing one of the rst research studies linking these two constructs and subsequent effects upon OCBS. We identi ed one prior study (Bazzoli et al., 2021) that examined the relationship between job insecurity and safety compliance in the U.S. during the pandemic. Our rationale for the association is grounded in job insecurity theory and evidence from the MENA region. From a theoretical standpoint, we know from an extensive body of research that employees experience increased job insecurity perceptions when they have greater uncertainty about their job status (Vander Elst et al., 2016). This can occur for structural reasons such as downsizing and strategic change, or it can occur from broad effects like macroeconomic shifts.
Uncertainty is very stressful to employees, yet there is often little in the way of responses that employees can use to address the threat of job loss. COVID-19 perceptions precipitate this sense of angst and worry because COVID-19 has traversed the world leaving unparalleled threats to society and economic systems. In particular, nations in the MENA region have the added challenge of systemic challenges to an adequate COVID-19 response due to vaccine production taking place elsewhere, the substantial decline in demand in sectors such as oil, heavy reliance on service economies, and corruption among politicians (OECD, 2020b). Prior research into macro in uences, such as war, has also shown to predict job insecurity (e.g., Mahmoud & Reisel, 2015). We reason that the scale of COVID-19 and its detrimental impacts upon GDP inside the MENA region will be positively related to job insecurity. For the purpose of this study, we de ne COVID-19 perceptions as the perceived probability of discomfort and/or worry about the potential negative impact on an individual.
The COVID-19 pandemic has dramatically disrupted the global business environment and workplace practices. It is broadly acknowledged that COVID-19 has contributed to the wholesale decline of businesses in 2020 (Hack-Polay, 2020;Nicola et al., 2020). Businesses that survived did so by reducing expenses, often targeting human capital reductions. These aggressive initiatives affect a variety of organisational domains and increase employee job insecurity. The political and socio-economic landscape that businesses operate in and strategies adopted by rms have been tremendously rede ned by the COVID-19 (Helm, 2020;OECD, 2020a).
Given that COVID-19 is an unfolding phenomenon, there is currently limited understanding of its total relationship with job insecurity. However, early indications and analyses suggest that workforce anxiety levels have risen globally (Hussain, 2020;OECD, 2020a). COVID-19 has caused structural economic and policy changes that have weakened job security. The current vaccination initiatives in developed nations have not yet widely affected the MENA region, suggesting that organisations will have to maintain their contingency measures in place for the foreseeable future (Hack-Polay, 2020;Helm, 2020). Hence employees' anxieties about job security are likely to persist. Some experts warn that workplace and employment security might never return as we knew it before COVID-19. The 'new normal' occasioned by the pandemic (Hussain, 2020) may lead organisations to create what Teece (2011) terms "dynamic capabilities" enabling organisations to "create, extend, or modify their resource base."

Job insecurity
Job insecurity is a perceptual construct that is uncomfortably familiar to employees worldwide as they come to perceive threats to their jobs. It is de ned as "perceived powerlessness to maintain desired continuity in a threatened job situation" (Greenhalgh & Rosenblatt, 1984, p. 438). In this research, we focus on aspects of job insecurity that are linked to the perceived threat of job loss rather than those pertaining to loss of job features or correlates such as feelings of hopelessness (Ashford et al., 1989;Shoss, 2017). According to Shoss (2017), job insecurity can be a consequence of many causes, inclusive of broad economic factors such as organisational restructuring, recession, etc. Further, job insecurity can be the result of job status (such as being employed on a short-term contract) or cognitive vulnerability (Shoss, 2017).
Job insecurity represents a considerable threat to individuals, and this is highly stressful. Stress theory is one of the most useful paradigms for understanding the effects of job insecurity, given that employees use personal resources such as energy to counteract the stressful experience of job insecurity (Folkman & Lazarus, 1984). Yet, employees facing the threat of job loss experience stress intensely because they are ill-equipped to counteract the threat to their job. The ensuing response cycle is distracting and draining, and the ultimate behavioural result is worsened attitudes and subsequent performance de cits. Additionally, job insecurity has been proposed as a breach of the psychological contract between employees and their employers (Rousseau, 1995). This implies that employees make efforts in exchange for pay, recognition and other key work-related outcomes. Nevertheless, job insecurity triggers a perception of violation of these expectations and leads to negative attitudes and behaviours in the workplace.
Prior research into broad macro factors such as wartime crisis has shown that war effects disrupt everyday life and economic stability, causing anxieties about job losses (Mahmoud & Reisel, 2015). The vast uncertainties stemming from the COVID-19 crisis have increased the prospects of job loss and, thus, should predict increased job insecurity. Therefore, we propose: H1: COVID-19 perceptions among service employees in the MENA region positively predict job insecurity Burnout Consistent with the predictions of stress reactions, we hold that employees will be unable to deploy resources to counteract the COVID-19 impact on job insecurity simply because there are few options to address such a broad macro stressor. Job insecurity will, therefore, be related to negative effects such as employee burnout. Employee burnout is de ned as a psychological syndrome comprised of emotional exhaustion, depersonalisation and reduced personal accomplishment (Maslach & Jackson, 1981). Burnout has been conceptualised recently as a psychological state (Bellou & Chatzinikou, 2015). Burnout erodes an employee's sense of achievement once employees become disconnected from work. Employees become burnt out in the absence of psychological need satisfaction and rewards (Abós et al., 2019;Taormina & Law, 2000). It may be simplistic to see burnout as mere physical exhaustion owing to employees being over-worked. Burnout develops when physical exhaustion interacts with other psychological factors as suggested above and the lack of preparedness to acknowledge the true demand of the work (Leiter et al., 2013;Taormina & Law, 2000). This fundamental aspect of burnout is well established in the employee burnout scholarship (Bellou & Chatzinikou, 2015;Datta et al., 2007;Maslach & Jackson, 1981). It is widely documented that burnout has signi cant negative consequences for both the worker (stress, lack of interest in career development and self-con dence, and depression) and the organisation owing to decreased productivity, labour turnover, and loss of reputation (Elder et al., 2014;Ullrich et al., 2012).
COVID-19 has already provided evidence of a link between the change caused by the pandemic and rising levels of burnout (Giurge & Bohns, 2020). This derives largely from increased remote working and isolation (Bowden, 2020;Giurge & Bohns, 2020), the burden of work and understa ng (Wu et al., 2020). A signi cant aspect of the existing literature closely related to our study is Bellou and Chatzinikou's (2015), whose conclusion links burnout and organisational change.
As the COVID-19 pandemic continues, burnout reactions expected, thus providing one of the motivations of the present research. Previous research shows that job insecurity can be one of the primary sources of burnout (e.g., Bosman et al., 2005;Maslach et al., 2001). The literature also posits job insecurity as a proxy through which external factors can affect other attitudinal and behavioural variables (Mahmoud & Reisel, 2015). Further, the most recent studies investigating COVID-19 psychological effects (e.g., Shuwiekh et al., 2020) have not extensively covered the MENA region. Thus, we expect that COVID-19 perceptions effects will be transmitted via job insecurity into burnout. Accordingly, we hypothesise that: H2: Job insecurity among service employees in the MENA region positively predicts workplace burnout.

Job Satisfaction
The adverse effects of job insecurity on job satisfaction are perhaps one of the most frequently reported ndings in the job insecurity literature (e.g., Guarnaccia et al., 2018;Reisel et al., 2010) including those conducted in the MENA region (e.g., , 2015 and meta-analyses (e.g., Cheng & Chan, 2008;Sverke et al., 2002). COVID-19 has blanketed the globe posing signi cant challenges to nations, organisations and individuals. Its macro impact on business has precipitated both contractions and delayed business openings, both with negative human capital implications. The persistent grip of COVID-19 on business output in the MENA region poses a disruptive force perceived by employees as a threat to job insecurity and a stressor about which little can be done. In the previous section, we reasoned how job insecurity perceptions are directly predictive of burnout and, in this section, we suggest that burnout negatively in uences job satisfaction. Employees perceiving their job as insecure are expected to be emotionally more exhausted, less motivated, and far less satis ed with work. Support for this argument is offered by theories dating as far back as Herzberg et al.'s (1959) Two-Factor Theory, as well as Deci and Ryan's (1985) Self-Determination Theory, Agnew and White's (1992) General Strain Theory and Robinson and Rousseau's (1994) Psychological Contract Breach theory. Therefore, we anticipate a negative relationship between burnout and job satisfaction. Moreover, recent empirical investigations (e.g., Capone & Petrillo, 2018;Chen et al., 2019;Lu & Gursoy, 2013;Peng et al., 2016) have reported a signi cant and negative relationship between job burnout and job satisfaction.
Based on the rationale of the theories and evidence cited above, we further expect that the effect of job insecurity on job satisfaction will be mediated through burnout, thus: H3: Burnout among service employees in the MENA region negatively predicts job satisfaction

Organisational citizenship behaviours
The job insecurity literature has evolved since the 1980s from its initial focus on negative effects on employee attitudes and health to subsequent organisational impacts through behavioural change. While our prior hypotheses show how job insecurity predicts burnout and reduced job satisfaction, in this section, we further reason that job insecurity also negatively in uences discretionary employee behaviour in the form of reduced organisational citizenship behaviours (OCB). OCB is de ned 'as individual discretionary behaviour that is not directly or explicitly recognised by the formal reward system, and that in aggregate promotes the effective functioning of the organisation' (Organ, 1988). Theory and empirical evidence show that job insecurity has negative implications for attitudes and behaviours. Our expectation is supported by meta-analysis (Sverke et al., 2002) that shows job satisfaction as a key antecedent to OCBs as well as a mediator (Mahmoud, Sitladin, et al., 2020) that transmits other variable effects, mainly triggered by job insecurity , into OCBs. Employees reduce discretionary contributions because they are stressed, unprepared to address the stress meaningfully, and are conserving resources by not doing things they are not obligated to do, such as helping coworkers, trying their best, and seeking training to build skills. Thus, building on the prior hypotheses, we further anticipate the effect of COVID-19 perceptions will negatively in uence OCBs through job satisfaction; thus: H4: Job satisfaction among service employees in the MENA region positively predicts OCBs.
We further anticipate that COVID-19 perceptions will have an effect on OCBs via a sequence of mediators composed of job insecurity to burnout to job satisfaction: H5: COVID-19 perceptions among service employees in the MENA region have indirect adverse effects on OCBs via a series of mediators following the order: job insecurity, burnout and job satisfaction.
A generational cohort includes an identi able group of individuals who share distinctive social or historical life events during critical developmental stages (Hernaus et al., 2014). However, the perspective about close similarities in generations has been questioned in recent research (Rudolph & Zacher, 2016). Rudolph and Zacher (2016, p. 2), for instance, suggest that "generational differences do not have an appreciable in uence on work processes and outcomes". Current developments may give some credence to this argument. In fact, economic development and globalisation make today's workplace more complex than ever before, and improved general health conditions lead to prolonged career life for employees (Yi et al., 2015), thus providing a rich context for this study. We, therefore, cautiously use the generational groupings for the purpose of conceptual consistency with mainstream sociological literature while acknowledging the theoretical limitation highlighted by Rudolph and Zacher (2016).
Baby boomers are providing an excellent opportunity for younger generations to play a more signi cant role in the workplace upon the boomers' retirement at a record rate and pace (Flippin, 2017b). As per Rudolph and Zacher's (2016) nding earlier cited, generational differences do not necessarily impact work outcomes. Their main effects lie in work values, expectations and attitudes, which could generate potential con ict and affect readiness for change (Bresman & Rao, 2017). In fact, unlike their Baby boomer managers and supervisors, who were prone to working long hours, Generation X employees tend to value work-life balance, making sure they have more time to dedicate to their families. Generation X is characterised as self-directed, sceptical and autonomous, born during a time of rapid change. While they are looking for work-life balance, they are not impressed by authority and micromanagement (Waltz et al., 2020). In contrast, Generations Y and Z -61% to 77% in a large survey (Bresman & Rao, 2017) -would engage in more aggressive paths to satisfy their leadership ambitions, often taking more signi cant risks, compared with only 57% of Generation X cohorts.
As Baby Boomers retire, Generation X employees are becoming senior staff members in the workplace, with their children (Generation Z) climbing the ranks (Seemiller & Grace, 2019). Generation X is not tech-novice as many members of this cohort share interests in social media and smartphones analogous to those of younger generations. Generation X employees tend to have different communication preferences from those of younger generations. For example, Generation Z employees prefer to use texting instead of e-mail to communicate in the workplace-a method not ideal for Generation X (Seemiller & Grace, 2019).
According to The Federal Competitiveness and Statistics Authority (FCSA), Generation Y is becoming the workforce's largest generation in the MENA region. For example, 58% and 64% of the UAE and Oman workforce were Millennials (FCSA, 2018; National Centre for Statistics and Information, 2020). Baby Boomers and early Generation X are the ancestors of Generation Y employees. Both Generation Y and Generation Z employees are digital natives. They are familiar with the internet content and nd technology to be an essential part of everyday life (Lebowitz, 2018;. Further, Generation Y is often viewed as self-con dent, connected, and agile (Taylor & Keeter, 2010).
Generation Z is the youngest generation of workers who are entering the workforce. Their use of technology and desire to engage in exible work arrangements are nearly identical to that of Generation Y (Ryback, 2016). Generation Z workers recognise the signi cance of nancial stability and realise the excitement of performing well at work. Interestingly, they focus their lives on professional success (Flippin, 2017a;. Together Generation Y and Generation Z are more ethnically diverse than any preceding generational cohorts (Bannon et al., 2011;Flippin, 2017aFlippin, , 2017b. Our study contends that understanding the generational gap in the workplace is critical as it may introduce con ict and low levels of employee and management engagement. However, when generational differences are successfully handled, a healthy workplace will be created, and employee engagement and motivation will be enhanced. The hypothesised model we test in this research examines a rst of its kind potential pandemic perceptions effects on OCBs via a sequence of attitudinal variables, namely job insecurity, burnout and job satisfaction. OCBs are important extra-role behaviours that support organisational goals. However, in a formal sense, they are not required by employers, e.g. helping colleagues, speaking favourably about the organisation to outsiders, sharing knowledge, and so forth (Organ & Paine, 1999). Therefore, we seek to understand the indirect relationship between COVID-19 perceptions and OCBs via a sequence of mediators (job insecurity, burnout and job satisfaction) while adding multigenerational complexities which may impact the relationships studied here, as suggested by De Witte (2016). This study will provide a novel understanding of the need to address generations differently in organisational response to job insecurity.
Based on the previous studies related to cognitive, attitudinal and behavioural variances across generational cohorts in the workplace (e.g., Moore et al., 2014;Stevanin et al., 2018;Twenge et al., 2010) we have witnessed rising concerns about the generational variances of emotional fragility when triggered by the perception of job stressors (e.g., American Psychological Association, 2018; Shellenbarger, 2019) especially during a pandemic time (Aratani, 2020;Davies, 2020;Lee, 2020). A recent large survey (Bresman & Rao, 2017) of over 19,000 people across several countries found striking differences between generations, which led the authors to suggest that organisations are compelled to keep those differences in mind. Such differences include differences in aspirations, values, needs and expectations. For example, with regards to the use of modern work-related technologies such as virtual reality (VR), which could signi cantly enhance one's ability to reskill or upskill themselves, Generation X showed signi cant reluctance compared with the enthusiasm shown by Generation Z and, to some extent, by Generation Y (Bresman & Rao, 2017). Also, we respond to the call  for investigating generational differences concerning the relationship between COVID-triggered job insecurity and OCB in non-Western contexts. Thus, we expect that generational differences will result in inconsistencies regarding employees' COVID-19 perceptions effects on OCBs transmitted via job insecurity, burnout and job satisfaction. We hypothesise: H6: Generation moderates the relationships between COVID-19 perceptions to OCBs among service employees in the MENA region.
To determine whether parallel mediations could exist, we hypothesise additional direct paths: COVID-19 Perceptions to Burnout, Covid-19 Perceptions to Job Satisfaction, COVID-19 Perceptions to OCB, Job Insecurity to Job Satisfaction, Job Insecurity to OCB, and Burnout to OCB. If they were found signi cant and sizable, the mediations, if statistically supported, would be deemed partial, otherwise full (Baron & Kenny, 1986;Hair et al., 2017;Nitzl et al., 2016). The research model is presented in Figure 1.

Method Research design
Our research is quantitative. It adopts a cross-sectional survey design study that supports post-empiricism assumptions and draws on social constructionism (Mahmoud & Tehseen, 2021) to form an understanding of our sample's attitudinal and behavioural tendencies.

Analysis
We adopted several indicators to assess the validity and reliability of measures employed in this study (Ghasemy et al., 2020;Hair et al., 2017). We tested the hypothetical model utilising a PLS-SEM approach method via SmartPLS 3 (Ringle et al., 2015). We conducted a path evaluation followed by multi-group analysis (MGA), building on standardised betas (β: for direct effects), un-standardised betas (B: for indirect effects) and the matching t-values by using bootstrapping, Q 2 for predictive relevance, and Cohen's f 2 to ascertain effect sizes where f 2 ≥ .02, f 2 ≥ .15, and f 2 ≥ .35 indicate small, medium and large effect sizes, respectively (Cohen, 2013). Besides, standard root mean square residual (SRMR) was used to evaluate the model t to the data (Henseler et al., 2014).

Participants
The MENA region occupies a strategically important geographic location between Europe, Africa and Asia. The population currently exceeds 450 million, or 6% of the world population (Ibrahim et al., 2020). Our study population consisted of customer service employees in the MENA region, where we randomly selected three countries, i.e., the United Arab Emirates, Egypt and Oman, out of the region's twenty countries. While customer service employees' perceptions and behavioural attitudes in the workplace have been shown to be instrumental for service industries' jobs; they play a central and integral part in the service delivery (Bordoloi et al., 2018), where employee interactions with the customer (internal or external) are directly linked to long-term economic sustainability.

Measurement of Variables
Appendix 1 shows a set of already validated measures that we utilised in this study. COVID-19 perceptions were assessed using a three-item scale that was designed for this study. We used previously validated scales reported in the work of Francis and Barling (2005) to measure job insecurity, Lang (1985) and Maslach and Jackson (1981) to measure burnout, Judge et al. (2006) to measure job satisfaction and nally Van Dyne et al. (1994) to measure overall workplace motivation, and nally, Van Dyne et al. (1994) to measure organisational citizenship behaviour. All measures were evaluated on a 5point Likert scale. Also, all of the measures, including the COVID-19 perceptions, ful lled the reliability and discriminant, convergent and construct validity criteria as exhibited in the results section.

Sampling techniques and sample size
The data analysed in this study were part of an ongoing participants recruitment process in three countries in the MENA region consisting of Egypt, Oman and the United Arab Emirates. That part represented data collected from the beginning of April to the end of June 2020. The primary approach to nding participants was through LinkedIn (a professional social network). We hired professional surveyors who set the search lter criteria to identify participants from the target population. Aiming for a sample size that would contain considerable numbers of each generational cohort, we approached more survey participants through other online social networks (e.g., Facebook, WhatsApp, etc.). The survey did not include a question identifying the participants' country of residence. All the participants were made cognizant of the purpose and procedures of the investigation. They were noti ed that they were permitted to ask questions, raise concerns about the study, or leave the study at any point. Agreement to participate in the survey was part of the questionnaire sent to the participants. Since the survey was conducted online, the signatures of the participants were not acquired. Responses to the survey were anonymous, and all participants were advised that their responses would be con dential. It took around ten minutes to ll out the questionnaire made available in Arabic and English. Given that all measures used in this study were initially constructed in an English-speaking Western context, all measures were double-translated (English to Arabic and then Arabic to English). This is the established procedure to eliminate meanings that are not intended in a foreign translation of an English scale . After this procedure, further endeavours were made to establish the face validity for the measures. The scales were shared with Arabic-speaking academics and practitioners in human resource management to assess the Arabic phrasing in the questionnaire items and suggest corrections where needed. Besides, the questionnaire was piloted to a convenience sample of thirty Arabic-speaking employees in Egypt to assure the understandability of the measures' items. As a whole, our study yielded 578 responses. We computed the generation variable by recoding participants' age using the generational thresholds identi ed in previous research.

Common method bias
Before moving on to the path and multi-group analyses, we ran Common-Method Bias (CMB) tests, which are required when using perceptual, self-report measures from a single survey (Podsakoff et al., 2003). The inner variance in ation factors (VIFs) values (see Table 1) were all less than 3.3 (Kock, 2015). Hence, there were no CMB issues identi ed.

Measures validity and reliability
We calculated the Heterotrait-Monotrait Ratio of Correlations (HTMT), and they are presented in Table 2. We found that all of the correlations had values less than .9 implying that all of the measures had a satisfactory discriminant validity (Hair et al., 2019). Table 3 indicates that all the constructs had average variance extracted scores (AVEs) higher than .5 (Fornell & Larcker, 1981), composite reliability values (CRs) between .7 and .9 (Hair et al., 2017) satisfying the convergent validity and reliability criteria for all of the measures (Hair et al., 2019). Moreover, Table 2 shows that the Variance In ation Factor values for all the measures' items were less than 5 offering evidence that collinearity is not a crucial issue (Hair et al., 2019).

Sample description
The nal sample was composed of three generational groups, i.e., Generation X (32%), Generation Y (45%) and Generation Z (23%). Fifty-eight per cent of our sample were male. A signi cant number of our participants were educated to a university degree (41%) and single (55%). Appendix 1 shows the descriptive statistics of the variables under investigation clustered into generational groups. Appendix 2 demonstrates the descriptive statistics (produced using SPSS version 26) of the path model constructs for the whole sample and each generational group.

Measurement Invariance of the Composite Models
Before examining the moderating role of generation, we ran Measurement Invariance of the Composite Models (MICOM), as suggested by Henseler et al. (2016), to test whether both con gural invariance and compositional invariance were established. As indicated earlier, we employed a variance-based approach; thus, according to (Hair Jr et al., 2018), con gural invariance was veri ed by default. Moreover, our permutation test (see Table 7) returned p-values greater than .05. This led us to accept the null hypothesis (the constructs' original correlations are non-signi cantly different from 1), offering evidence of compositional equivalency and supporting the viability of conducting multi-group analysis (Hair Jr et al., 2018).

Multigroup Analysis
To assess the path model invariance across generational cohorts, we ran a multi-group analysis (MGA). Using the t-values associated with the multiple comparisons and reported in the parametric tests, we found the paths: job insecurity to burnout, burnout to job satisfaction, and job satisfaction to OCB were substantially not invariant across Generations X, Y and Z. Figure 2 illustrates the paths moderated by generational differences. More detailed, Table 8a shows that generation Y (β y = 0.57; P < .01) levels of burnout tend to be more susceptible to variations in perceived job insecurity than any other generational group in our sample. Only younger generations, Generation Z (β z = -0.638; P < .01) and Generation Y (β y = -0.374; P < .01) tend to experience slumps in job satisfaction as the levels of burnout are ampli ed. This means that Generation X (β x = -0.066; P = .638) is considerably less likely to experience a lack of job satisfaction than younger generations because of burnout. Job satisfaction amongst Generation X (see Table 8b) seems to block the cascading effects of COVID-19 perceptions on OCB (B = -0.003; P = .621), which is not the case for younger generations where COVID-19 perceptions damages extend to job satisfaction (B y = -.083; P < .01; B z = -.012; P <.01) and OCB (B y = -.042; P < .01; B z = -.064; P <.01). This implies that younger generations are substantially more vulnerable to the indirect effects of COVID-19 perceptions than Generation X. With the relationship between COVID-19 and job insecurity not moderated by generation whilst the remaining three paths are either directly or indirectly moderated by generation, we arrive at the conclusion that H6 is partially supported.

Discussion
Our study empirically examined a path model between COVID-19 perceptions and OCBs via job insecurity, burnout and job satisfaction as mediators. Our ndings revealed that a sequence of three mediators: job insecurity, burnout and job satisfaction, mediates the relationship between COVID-19 perceptions and OCBs. As expected, we found a positive correlation between COVID-19 perceptions and job insecurity, job insecurity and burnout, a negative correlation between burnout and job satisfaction. Job satisfaction was found to be positively correlated to OCBs (supporting hypotheses 1-5). Furthermore, our model con rmed that COVID-19 perceptions exerted a negative indirect effect on OCB. This means the perceptions of COVID-19 will elevate burnout by triggering job insecurity. Increases in the levels of workplace burnout will decrease OCBs via eroded job satisfaction amongst employees.
Our ndings expand upon prior research that examined broad macro in uences like wartime conditions on job insecurity.
We then further investigated the moderating impact of generational differences on the whole path from COVID-19 perceptions to OCBs amongst customer service employees in three countries in the MENA region. Our generational analysis reveals the existence of some signi cant cross-generational variations. Generation X employees are shown to be more likely to increase their involvement in OCBs when they are satis ed with their job; this is not as discernible as it seems amongst younger generations-both Generation Y and Generation Z workers. Our results suggest that this can be a result of the hindrance role of job satisfaction which blocks the cascading effects of burnout amongst Generation X. The intensity of the connection between increased levels of COVID-19 perceptions and lower engagement in OCBs is ampli ed amongst employees of younger generations compared to those of the older Generation X. When generational variations are understood and successfully managed, particularly for customer service employees, increased employee engagement and improved motivation will lead to a more balanced and healthier workplace, which in turn translate into customer satisfaction and loyalty as suggested by the service pro t chain model .

Practical implications
The workplace in contemporary organisations requires that researchers study differences that are attributable to demographic factors such as generational composition. Firms seek practical guidance that will help stimulate optimal performance and sustain valuable discretionary behaviours such as OCBs. We need research that identi es patterns of consistency or points to divergent patterns among different generations.
The current study found that COVID-19 perceptions can be classi ed as a crisis, much as that of wartime-like crisis, and this causes indirect harm to employees' willingness to perform non-formal requirements of the work (Mahmoud & Reisel, 2015;. Interestingly, we found that COVID-19 perceptions had a similar positive association with job insecurity across the three generations. This result further supports the idea that no one is safe concerning the perceived threat to their job security when triggered by COVID-19; even the greater propensity of Generations Y and Z to contemplate paths to leadership and willingness to take risks does not shield them from threat to job security. However, the subsequent attitudinal and behavioural consequences of job insecurity were found to vary across generations. Our results suggest that younger generations, especially the Millennials, are more vulnerable to developing burnout as a result of pandemic-caused job insecurity than other generations. This vulnerability may be ampli ed in the MENA cultural context where greater credence is accorded to age and seniority, meaning potentially more protection or job security for Generation X cohorts (Litwin, 2008) and high unemployment rates for younger people. This result is consistent with the current discourse on employment, where Millennials are being used metaphorically to depict employees' vulnerability to the growing precarity of work amidst the COVID-19 pandemic. For instance, Filipovic (2020) subtitles her article 'we're all Millennials now' to describe the deteriorating work conditions like job insecurity, low pay and working from home alongside the adverse emotional rami cations that have been introduced by modern technology and virtual workplaces. Such pandemic-time consequences have not been limited to the Millennials but also affected the workforce from all generations (Filipovic, 2020).
Another important nding is that, unlike Generations Y and Z, Generation Xers' COVID-19 indirect effects onto OCBs were found non-signi cant. This nding accords with our analysis which revealed that job satisfaction was not signi cantly related to burnout amongst Generation X employees suggesting that job satisfaction buffers the trickling effects of COVID-19 perceptions onto OCBs through the sequence of mediators-job insecurity, burnout and satisfaction. Moreover, this further explains why satis ed Generation Xers are substantially more likely to perform OCBs than younger generations. Besides, these results are in keeping with previous observational surveys about mental health across generations. For example, A recent report (Moody's Analytics, 2019) revealed that as Millennials age, the generation is witnessing much faster declines in mental and physical health than those from Generation X. Besides, the report warned that unless proper management or treatment are developed and implemented, the Millennials could see a 40% surge in mortality rate in comparison with the Generation Xers of the same age (Moody's Analytics, 2019).
When considering generational cohorts, it becomes increasingly important to develop workplace communication strategies aimed at reducing con ict among employees by adopting a motivation perspective of communication. There is, therefore, a distinctive need for managers to interact with employees in ways that would position work as a positive experience. That would help in protecting employees' job satisfaction and, consequently, their levels of engagement in OCBs as an outcome of controlled burnout, according to our results. This implies that generationally diverse organisations should pay attention to the consequences of the variations in the communication tools they use. In other words, employees of different generations may have contrasting views of what constitutes a healthy communication environment (Mehra & Nickerson, 2019). Clearly, with COVID-19 having led many organisations to adopt virtual workplace environments, generational differences amongst employees must be considered as a matter of crucial concern for these organisations. For instance, when Generation X employees favour e-mailing over other communication tools, they might feel unhappy with Generation Z employees' preference for texting (Seemiller & Grace, 2019) which can elevate the chances of organisational con icts, particularly as the traditional workplace environment has been replaced by largely virtual ones almost overnight during the pandemic.

Research implications and limitations
We examined the impact of employee's COVID-19 perceptions on the extent of employee's engagement in OCBs and found that it is not invariant to the generational differences. This study contributes to our understanding of how times of pandemic can affect OCBs as we tested a path model linking COVID-19 perceptions to OCBs via employee's job insecurity, burnout and job satisfaction as mediating factors. Another key research nding is the acknowledgement of generational issues in the workforce, which can engender major uctuations in behaviour in terms of the responses to increased COVID-19 perception from different generation groups.
The study showed that workers among the Generation Y and Generation Z cohorts unsurprisingly have a greater propensity to withdraw from OCBs if they have apprehension about job insecurity, which causes them to develop burnout symptoms at a faster pace compared to Generation X cohorts. Generally, the individual and organisational consequences of higher job insecurity and burnout tend to be less of a concern for workers within Generation X. An explanation for this may be linked to the buffering role of job satisfaction amongst Generation X workers as well as the fact that younger generational cohorts can be more vulnerable to job stressors and hence be at higher risk to disengage in OCB. Besides, younger generations show more exibility and willingness to engage with changing job contexts and relocating, aiming for more secure and stable jobs (e.g., Filipovic, 2020). Most chief nancial o cers in the MENA region (e.g., 39% in the UAE) continue to see adverse changes in sta ng and layoffs (ILO, 2020). Unemployment in Egypt has risen to 9.6% in the second quarter of 2020, compared to 7.5% in the same period the previous year, and it is thought that the COVID-19 crisis may amplify poverty rates in the country to 40 per cent from 30 per cent (Middle East Monitor, 2020). The extent of job insecurity could continue to rise during the period of the pandemic. Based on our ndings, it is anticipated that this would have negative rami cations for individuals, societies, and governments.
A signi cant endeavour for organisations is to strive to retain the experience of their Generation X employees. These have shown loyalty and have a wealth of skills (Neal, 2019) that could be deployed in various ways. For example, they could act as potent external consultants due to their inside knowledge of the organisation. What is needed is to reassure Generation Xers about their continuous importance to organisational continuity. This way, organisations will strike a balance between nurturing Millennials and rewarding Generation Xers, who will continue to ensure the retention of organisational knowledge (Neal, 2019). In the context of COVID-19, human resource management (HRM) has an important role to play in providing counselling, advice and guidance to all employees to ensure they do not disengage. HRM also has a responsibility to design work and contract patterns that create room for exibility. For example, the recent furlough strategy (reduction in hours or pay and temporary layoffs) adopted by many companies during the COVID-19 pandemic was geared to keeping the employees connected and ready for re-hiring (Siegel, 2020).
While this research has successfully shown that generational differences lead to variations in the hypothesised relationships, it has, nevertheless, a particular limitation about the probability that such differences may also be generated by the age or status of the corresponding generational cohort. Therefore, we suggest that further research should address this matter. Recent research on generational differences in the workplace ) encourages looking at whether or not Generation Z employees' attitudinal and behavioural patterns might turn out within the next two decades to be more analogous to Generation Y's in the 2010s. Alternatively, whether Generation Alpha (entering the job market in 2030) would demonstrate the same characteristics as Generation Z today. If yes, that would suggest that such variations in organisational attitudes and behaviours were age-caused rather than generational. If not, that would earn the generational cohort's impact more legitimacy in triggering attitudinal and behavioural differences in the workplace.
This study sample description was limited by the absence of a question identifying the respondents' current country of residence. Although the selection of three countries was aimed primarily for better representation for the MENA region, however, including that question would have allowed for richer pro ling for our sample. Further, the limitation of the study with regards to data collection solely from the MENA region warrants a re-examination of the effects of COVID-19 effects on job attitudes and organisational outcomes and the moderating role of generation in additional geographical settings. It is also recommended that future researchers examine multi-cultural, multi-gender, and multi-industry re nements to the model. That will further enrich our understanding of employees' contextual performance and discretionary behaviours for different genders and across several different employment sectors. Moreover, the countries' economic development, levels and speed of technology adoption and advancement embedded in service jobs might all moderate the complex dynamics of the impact of pandemic-triggered job insecurities on OCBs among a diverse workforce. Future studies may be able to integrate some or all these factors to explore potential moderations in our model.
One thing that might produce response bias in the answers obtained and analysed in this study is the employment of selfreporting data (Mahmoud & Grigoriou, 2019). We acknowledge this limitation because this type of data, tacitly, suggests that the respondents have the same comprehension or interpretation of the survey items . The research is also limited by the cross-sectional design, so causal effects are implied but not formally evaluated.
Therefore, future research is encouraged to collect longitudinal data to understand the real effects and causal relationships better. However, employing a longitudinal design to evidence causality has been ampli ed (Spector, 2019). It only offers limited advantages over the cross-sectional design in most cases in which it is used (Spector, 2019). Additionally, consistent with Tharenou et al. (2007), discoveries made out of cross-sectional research can still be interpretable and legitimate if a robust theoretical basis is embraced. Moreover, cross-sectional design for data collection has received backing for scholarly inquiries in wartime-like circumstances (e.g.,  and other contexts of extreme environments (e.g., .
Finally, while the new measure of COVID-19 perceptions, developed for the current study, is assessed in this paper. It would require further con rmatory investigations equally for construct validity and predictive validity, especially when employed in contexts outside the boundaries of the MENA region. Future studies on the validation of the concept of COVID-19 perception and its measure are therefore highly endorsed.
-Data available on request due to privacy/ethical restrictions -The authors declare no competing interests.
-No nancial support was received for this research. epidemic in Wuhan, China. Journal of pain and symptom management.
Yi, X., Ribbens, B., Fu, L., & Cheng, W. (2015). Variation in career and workplace attitudes by generation, gender, and culture differences in career perceptions in the United States and China. Employee Relations, 37 (1)           Path model analysis