Effects of internal and external corporate social responsibility on employee job satisfaction during a pandemic: A medical device industry perspective

Abstract

The outbreak of the COVID-19 pandemic has dramatically changed human lifestyles and contributed to the creation of a new normal in the business environment. This study examines the direct and indirect impacts of internal and external corporate social responsibility (CSR) practices on employee job satisfaction through organisational identification, conditional on employee age. A total of 236 valid responses were received from eight multinational medical device manufacturers in Malaysia. Partial least squares and PROCESS algorithms were employed to assess the hypothesised interactions between the predictors and criterion variables. The empirical results showed that internal CSR (i.e., CSR to employee) could significantly drive a greater sense of belonging and work satisfaction. Surprisingly, however, external CSR (i.e., CSR to community) negatively affects job fulfilment in the medical devices industry during the pandemic. Nevertheless, the findings also showed that ongoing CSR activities in the community could build organisational identification and subsequently improve job satisfaction. Conversely, CSR to environment did not statistically influence job satisfaction, either directly or indirectly. The mediating effects of organisational identification were not associated with employee age. This study provides a practical framework for effective CSR strategies amid the pandemic that can help organisations align with social responsibility, enhance their reputation, and contribute to society.

1. Introduction

Job satisfaction can drive employees' commitment and encourage positive managerial and financial performance (Chamberlain & Zhao, 2019; Zhou et al., 2018). Business units with people who are thriving at work have reported higher profits than units with miserable workers (Gallup, 2022). Conversely, low job contentment decreases the quality of work deliverables, causes higher employee turnover, and depletes organisational success. In 2021, global workers remained under the heavy burden of the COVID-19 waves. According to the State of the Global Workplace 2022, 60% of world workers are emotionally detached, and 19% are miserable or disengaged. Europe and South Asia dropped five percentage points in well-being in the past year, and the latter even reached the lowest regional score in the world, at 11% (Gallup, 2022). Correspondingly, the Workforce Hopes and Fears Survey 2022 also showed that 41% of Malaysian employees were neutral or not satisfied with their work, and 17% indicated that they were likely to switch to a new employer in the next 12 months (PwC, 2022). However, increasing employees’ job fulfilment is a complex and challenging issue.

In recent years, corporate social responsibility (CSR) has been identified as a prominent factor in predicting job satisfaction in the organisational psychology field (e.g., Du et al., 2015; El Akremi et al., 2018; Lee et al., 2012; Zhou et al., 2018). For instance, based on the data from over 85 thousand employees at 381 Brazilian companies, Barakat et al. (2016) provided evidence that CSR-oriented actions would lead to a better company image, and consequently, to greater employee satisfaction. Correspondingly, Du et al. (2015) found that even if employees have multi-faceted job needs (i.e., financial, developmental and ideological needs), CSR programs positively impact job satisfaction and reduce employee turnover intention. The systematic descriptive review conducted by Macassa et al. (2020) also showed that most empirical findings in the European context exhibited a direct and positive relationship between perceived CSR and job satisfaction. However, CSR activities can also negatively affect perceptions of a company's social responsibility due to overly intensive communication (Viererbl & Koch, 2022). Furthermore, the global crisis, such as COVID-19 pandemic, has severe implications for businesses and their coping strategies (Amankwah-Amoah et al., 2021; Bouncken, Kraus, & de Lucas Ancillo, 2022). For instance, the pandemic has accelerated the adoption of remote work (e.g., home office and hybrid/flexible work) as a way to maintain business continuity while ensuring employee safety and reducing the risk of infection (Bouncken, Lapidus, & Qui, 2022). Whether CSR practices would have similar impacts attributed to the “new normal” business environment due to the COVID-19 pandemic remains unknown. Thus, it is essential to examine the relationship between CSR-oriented activities and employee job satisfaction by incorporating intermediate factors during these new circumstances.

The notion of CSR is highly fragmented and covers arguments from various angles, disciplines, practices, and ideological stands (Barakat et al., 2016; Viererbl & Koch, 2022). CSR initiatives can range from benevolence, social or charitable campaigns, employee well-being, and community services to sustainable business practices (Du et al., 2015). This study operationalises CSR as a firm's actions or obligations to various stakeholders beyond its legal requirements (Turker, 2009a). It can be separated into two main categories: CSR towards internal and external stakeholders (Farooq et al., 2017; Hameed et al., 2016). The classification is consonant with the stakeholder theory narrative (Du et al., 2015; Gibson, 2000), which uses the designation “primary” to describe stakeholders who are closely (or formally) tied to a firm's business operation, and “secondary” to describe those who are more distantly (or informally) tied. To obtain parsimonious research, we specify internal CSR as the firm's socially responsible actions towards “employees”, whereas external CSR is the CSR-oriented activities towards the “community” and the “environment”. The selected three groups are among the most prevalent stakeholders in previous studies (e.g., Farooq et al., 2014; Farooq et al., 2017; Turker, 2009a).

The current literature demonstrates that employee job satisfaction has been investigated based on social exchange and social identity theories. While these theories are not new, this study intends to present further important information in a new pandemic reality. According to social exchange theory, organisational pride and satisfaction can be gained when employees acknowledge that their company is concerned about staff welfare, society, and the environment (Cropanzano & Mitchell, 2005). In other words, employees might feel that their contributions to the company are exchanged and ultimately returned to their community and environment. Social identity theory describes how the company's positive and pleasant social image cultivates employees' organisational identification (e.g., self-esteem and commitment) to develop greater workplace fulfilment. Thus, the researchers suggested that organisational identification mediates different CSR practices and job satisfaction. However, there is only limited research that emphasises the underlying role of corporate psychological mechanisms that mediate CSR and employee satisfaction (Jones et al., 2017), while empirical studies on their conditional indirect effects (or moderated mediation relationship) are even scarcer. Hence, the research questions of this study are: a) Do internal CSR (i.e., CSR to employee) and external CSR (i.e., CSR to community, CSR to environment) directly affect employee job satisfaction during the COVID-19 pandemic? b) Do internal and external CSR indirectly affect employee job satisfaction through organisational identification? c) Do demographic characteristics such as employee age moderate the mediation effects of organisational identification between CSR (i.e., internal and external) and job satisfaction?

The motivation behind this study can be described from theoretical and practical perspectives. From the academic aspect, as defined by Summers (2001), this study improves the theoretical rationale of social identity theory and social exchange theory by explaining the relationships among the latent variables in the previous research framework. Second, this research contributes to the body of knowledge by identifying and conceptualising all the latent variables under the pandemic circumstances, addressing a limitation in previous literature (Jones et al., 2017). Last, this study provides insights into the operational definition of latent variables such as CSR, organisational identification, and job satisfaction in the context of developing countries. On the other hand, the expected findings could serve as practical references for the corporations in helping them to decide whether they should retain existing CSR practices or improvise according to the developments caused by the COVID-19 pandemic.

This paper is organised as follows: Section 2 reveals previous literature reviews, the conceptualisation of critical constructs, and hypothesis development. Section 3 discusses the research methodology, respondents’ background, and the approach to data analysis. Section 4 exhibits the analysis results and discussion. Section 5 presents the conclusion and theoretical and practical implications, examines potential bias, and provides suggestions for future research.

2. Literature review

The researchers conducted a systematic literature review on CSR and job satisfaction to comprehend the discussion and ongoing debates in the study area and identify research gaps in previous empirical studies. To select core articles, specific keywords (i.e., CSR, employee retention, job satisfaction, employee well-being) were searched in academic databases, resulting in a preliminary list of 710 articles. After some filtering processes and manual review, 15 narrow-scoped articles were selected, and the main results are displayed in Table 1 . However, the literature review was not limited to these 15 core articles. The study also incorporated elements from other disciplines and perspectives to provide a more comprehensive understanding of the academic interest in the topics.

Table 1.

Prior quantitative studies on CSR and job satisfaction.

No.	Author (year)	Country and respondents	Main findings/results
A1	Barakat et al. (2016)	Brazil	CSR had a positive relationship with job satisfaction. Organisational image mediated the relationship between CSR and employee job satisfaction.
		85167 valid responses from 381 companies
A2	Brammer et al. (2007)	United Kingdom	External CSR was positively related to organisational commitment and job satisfaction.
		4712 employees from financial services companies
A3	Du et al. (2015)	USA	CSR was positively related to reducing turnover intention and job satisfaction.
		322 employees from a sizeable national leadership conference for professional women
A4	El Akremi et al. (2018)	France	CSR was positively and directly related to perceived organisational support and organisational pride. CSR was also indirectly and positively related to organisational identification, affective commitment, and job satisfaction.
		332 employed MBA students
A5	Hameed et al. (2016)	Pakistan	External and internal CSR had no direct effect on organisational identification but showed indirect impacts via perceived external prestige and internal respect.
		414 employees across five large multinational companies
A6	Kim et al. (2018)	South Korea	Economic and philanthropic CSR positively affected the quality of work life (QWL), while ethical and legal CSR was insignificant towards QWL. Employees' job satisfaction and QWL would affect the standard of their life.
		442 employees from upscale hotels
A7	Kim et al. (2020)	USA	Philanthropic and ethical CSR separately had positive relationships with organisational identification. Their indirect effects on the QWL via organisational identification were also significant. The indirect relationship between economic CSR on the intention to stay through the QWL was significant.
		203 employees from Casino Hotels
A8	Lee et al. (2012)	South Korea	Philanthropic and economic CSR had a significant positive relationship with organisational trust. Ethical CSR and job satisfaction were positively related.
		21 franchised food services
		Sample size: 276
A9	Farooq et al. (2014)	Pakistan	CSR to employee, community, and consumers positively affected knowledge-sharing behaviour and organisational identification.
		378 employees
A10	Farooq et al. (2017)	South Asia	External CSR and internal CSR separately had a positive influence on organisational identification.
		408 employees in France & 415 employees in Pakistan
A11	Raub and Blunschi (2014)	United Kingdom	Employees' awareness of CSR was positively related to job satisfaction.
		211 employees from a hotel chain
A12	Senasu and Virakul (2015)	Thailand	Perceived CSR and implemented CSR positively influenced job satisfaction and work-life satisfaction.
		1277 employees from different employment sectors
A13	Turker (2009a)	Turkey	CSR to the customer, employee, non-social, and social stakeholders significantly influenced organisational commitment. The relationship between CSR to the government and employee commitment was insignificant.
		269 business professionals
A14	Vlachos et al. (2013)	European Union	CSR-induced intrinsic values forged by charismatic leadership positively influenced job satisfaction. CSR-induced extrinsic values were not able to predict job satisfaction.
		438 employees from three world-leading manufacturers
A15	Zhou et al. (2018)	China	Perceived CSR increased organisational pride. Job satisfaction and affective commitment were influenced by organisational pride. Organisational pride mediated the relationship between perceived CSR and job satisfaction.
		268 employees in 12 companies

In brief, Table 1 provides an overview of prior quantitative studies on the relationship between CSR and job satisfaction. The table includes information on the author, year, country, and respondents, as well as the main findings and results of each study. The studies were conducted in various countries, including Brazil, the United Kingdom, the United States, France, Pakistan, South Korea, Thailand, Turkey, and China, and included employees from different sectors and industries. Overall, the table suggests that CSR has a positive relationship with job satisfaction, and that this relationship is mediated by various factors such as organisational image, organisational commitment, turnover intention, perceived organisational support, organisational pride, and work-life satisfaction. The table also highlights the importance of different types of CSR, such as economic, philanthropic, ethical, and legal CSR, and their impact on job satisfaction and other organisational outcomes.

The current study aims to address the research gaps in previous studies on CSR and its impact on employee job satisfaction. Through a content analysis of the selected past research articles, we identified two main areas where previous studies have shown inconsistencies and gaps in literature. First, the conceptualisation of CSR varies widely among researchers. Some studies have presented CSR as a single and consolidated construct (Barakat et al., 2016; Du et al., 2015; Zhou et al., 2018), while others have manifested it into multiple facets such as organisational aspects (i.e., internal and external) (Brammer et al., 2007; Farooq et al., 2017; Hameed et al., 2016), philanthropic dimensions (i.e., economic, legal, ethical, and philanthropic) (Kim et al., 2018, 2020; Lee et al., 2012) and stakeholders’ perspectives (i.e., employee, community, environment) (Farooq et al., 2014; Turker, 2009a). Even though organisational identification has been identified as a construct that can influence CSR (Farooq et al., 2014, 2017) and job satisfaction (El Akremi et al., 2018; Hameed et al., 2016), the lack of a unitary and broadly used measurement of the CSR notion has led to untested relationships between various conceptualisations of CSR, organisational identity, and job satisfaction.

Second, the past empirical results also revealed inconsistent correlations between notions of CSR, in either single or multiple dimensions, and employee job satisfaction. Some research findings showed that CSR to community (or philanthropic CSR) has a positive effect on employee satisfaction (e.g., Barakat et al., 2016; Raub & Blunschi, 2014; Zhou et al., 2018), but it was insignificant for Lee et al. (2012). The content analysis also exhibits limited intermediated linkages examined in past studies, which encourages the inclusion of organisational identity (as a mediator) and employee age (as moderators). Additionally, the collection of information from developing economic countries (e.g., Malaysia) is scarce.

This study aims to contribute to the management discipline by providing new insights into the relationship between CSR, organisational identification, and job satisfaction during the COVID-19 pandemic. With the use of empirical data collected during this unprecedented period, this research seeks to fill the literature gaps and shed light on the ways in which CSR practices can impact employee job satisfaction in the “new normal” business environment.

2.1. Theoretical framework and hypothesis development

Job satisfaction is a complex phenomenon influenced by various factors, including pleasurable state resulting from perceived job experiences (Kim et al., 2018). Job satisfaction is determined by the employee's feelings, such as whether the job is comfortable, fun, and makes them feel important. Low job satisfaction can lead to low staff morale, disengagement from work, and reduced productivity and performance (Chamberlain & Zhao, 2019). Given the current pandemic situation, it is important to explore how CSR practices can influence employee job satisfaction.

Based on the literature review, two prominent theories have been consistently employed in explaining CSR and job satisfaction: social exchange theory and social identity theory (Du et al., 2015; Kim et al., 2020). This study assimilates both views to establish a more comprehensive framework for predicting job satisfaction. Under this broad picture, we use the social exchange perspective to study how external and internal CSR influence employee job satisfaction. Meanwhile, the social identity perspective facilitates the importance of organisational identification as a mediator between the predictors and criterion. It includes individual characteristics such as employees’ age as a moderating variable.

Social exchange theory is commonly used to explain workplace relationships and behaviours (Cropanzano & Mitchell, 2005). Social exchange behaviour can be the interchange between tangible materials and intangible sensations (Du et al., 2015; Kim et al., 2020). For instance, a company is perpetually concerned with staff well-being and provides a decent working environment. In return, employees' organisational pride and satisfaction would be higher (Cropanzano & Mitchell, 2005). Employees might feel that their contributions to the company are exchanged for their benefit. Hence, organisational efforts on CSR could induce positive outcomes and job satisfaction when employees perceive CSR as a valuable activity (Barakat et al., 2016; Zhou et al., 2018). Furthermore, CSR fosters rewarding situations to improve employees’ organisational commitment (Lee et al., 2012; Raub & Blunschi, 2014; Zhou et al., 2018) and motivates employees to support its goal.

Meanwhile, social identity theory describes that the company's positive and pleasant social image will cultivate employees' organisational identification (sense of belonging) and encourage them to perform actions aligned with the organisation's directions and targets (Du et al., 2015; Kim et al., 2020). For instance, CSR activities, such as voluntary or community service and nature conservation programmes, positively affect the organisational image and pride (Barakat et al., 2016; Zhou et al., 2018). Thus, social identity theory provides a clear explanation of the influence of CSR on forming organisational identification. Subsequently, social exchange theory further explains how the direct and indirect effects (via organisational identification) of various types of CSR determine employees' job satisfaction.

CSR is broadly defined as non-obligatory organisational policies and systems to improve long-term social, economic, and environmental benefits (Turker, 2009a). Nowadays, CSR initiatives are a part of corporate strategy to gain a competitive advantage. As discussed in the previous section, the CSR notion can vary from single to multidimensional perspectives, as shown in Table 1. Based on past arguments, this study delineates the CSR notion into internal and external aspects (Brammer et al., 2007; Farooq et al., 2017; Hameed et al., 2016). Grounded in stakeholder-oriented CSR (Farooq et al., 2014; Turker, 2009a), internal CSR is represented by CSR to employee. Meanwhile, firms’ collective conscientiousness for social and natural environmental factors has become a crucial metric. Thus, external CSR is designated as CSR to community and CSR to environment. This study anticipates that different types of CSR targeting various stakeholders might generate distinctive impacts on employee job satisfaction.

Internal CSR is the firm's social responsibility initiatives that are directly related to their internal stakeholders' (i.e., employees') physical and psychological working environment. The CSR to employee should go beyond compliance with the conventional organisational practice required by law or cultural norms (Farooq et al., 2017). The internal CSR activities should drive advantages to the employees instead of the company (Margolis & Walsh, 2003) and should be sensitive to workers' specific necessities and preferences (Turker, 2009b). During the COVID-19 outbreak, not everyone was able to work remotely, and for some essential sectors like medical and health care, employees still needed to go to work. We anticipate that CSR to employee actions can turn a pandemic crisis into an opportunity to generate a mild chain reaction for the improvement of job satisfaction through, for instance, implementing flexible working arrangements (i.e., working hours and place) to sustain a work-life balance, ensuring safety and fairness across the whole workforce, establishing online education support and career advancement opportunities, encouraging social and voluntary support activities among staff, and providing exceptional care to employees in need (e.g., employees at high risk, working parents with childcare problems). Social exchange theory describes that employees' organisational pride and satisfaction would be higher when their operating company safeguards their well-being during difficult times (Cropanzano & Mitchell, 2005). Hence, the following hypothesis is proposed:

Hypothesis 1

CSR to employee has a positive effect on job satisfaction.

External CSR represents corporate behaviours that aim to positively affect external stakeholders and extend beyond economic interest (Farooq et al., 2017). Some empirical studies (e.g., Barakat et al., 2016; Senasu & Virakul, 2015; Vlachos et al., 2013) showed that external CSR positively affects job satisfaction, but this finding is not always consistent. Thus, more advanced studies (e.g., Kim et al., 2018; Lee et al., 2012) specified external CSR into different dimensions and found distinctive impacts on job satisfaction. As discussed in the previous section, there is no unitary segmentation for external CSR, and the relevant research is still limited. This study adheres to the approach used in previous studies (e.g., Farooq et al., 2014; Farooq et al., 2017; Turker, 2009a) and designates external CSR into CSR to community and CSR to environment.

This study is contextualised under a pandemic reality. We thus envision that the delivery mechanism of CSR to community has to be pivoted accordingly. Still, the main agendas of CSR-oriented actions towards the community, as proposed by Farooq et al. (2014), should have remained. Specifically, as the COVID-19 outbreak affects all segments of the population, and is particularly detrimental to vulnerable groups, CSR to community practices such as charity contributions, community health and well-being aid, and support for non-governmental organisations (NGO) to improve social welfare are still crucial and urgent. Meanwhile, CSR for natural environmental protection (CSR to environment) should be continuous. Many countries declared nationwide lockdowns and movement restrictions during the outbreak of the coronavirus disease. The temporary economic activity slowdown reduced air, water and noise pollution and carbon emissions in different cities and tourist spots. However, it also heavily increased the haphazard use and disposal of chemical disinfectants, facemasks and medical waste (Rume & Islam, 2020). The pollution and emissions are expected to rebound once economic activities recover, and the burden of untreated pandemic wastes could continuously endanger the environment unless decisive environmental action is taken. Firms’ CSR to environment practices, such as minimising greenhouse gas emissions and establishing ecological pollution policies, organising more nature conservation projects and awareness programs to improve and protect nature, and investing in sustainable production and consumption, are among the essential procedures (Farooq et al., 2014).

In a nutshell, external CSR actions can generate positive work attitudes as a symbol of prestige that brings benefits towards sustainable development for future generations (Kim et al., 2020). Based on social exchange theory and previous studies, we anticipate that both external CSR initiatives of CSR to community and CSR to environment will positively influence employees’ job satisfaction during the COVID-19 pandemic. Thus, this study proposes the following hypotheses:

Hypothesis 2

CSR to community has a positive effect on job satisfaction.

Hypothesis 3

CSR to environment has a positive effect on job satisfaction.

The previous empirical literature has shown an inconsistent correlation between CSR and job satisfaction. This study mitigates the literature gaps by proposing a mediator (i.e., organisational identification) and a moderator (i.e., employee age). First, we will justify and hypothesise the mediation role of organisational identification between CSR and job satisfaction. Following on from that, this study extends the intermediate relationship by involving linear moderated mediations (Hayes, 2015) in the path analysis to make the results more insightful. The rationalisation for the hypothesis on the mediating effect conditional on employee age will be exhibited accordingly. Furthermore, the outcomes can denote novel insights constructed in the pandemic era.

Corporations are often stigmatised as profit maximisation machines that ignore all their responsibilities. Indeed, no individual is willing (or inclined) to be part of an organisation that is unethical or insensitive towards genuine environmental concerns. Rather, employees would feel proud and willing to acknowledge their company's ongoing work on humanitarian and sustainability issues, coupled with the territorial impact of COVID-19. When employees perceive that their company takes responsibility for them (CSR to employee) and is accountable towards the present and future of the society (CSR to community) and the natural environment (CSR to environment), rather than just being another profit-making company, their organisational identification would improve (Kim et al., 2020; Mael & Ashforth, 1995).

Moreover, employees with high organisational identification would view company successes as their successes and clarify when someone criticises their company (Kim et al., 2020; Mael & Ashforth, 1995). Employees feel honoured to be associated with this pride and reputation, increasing their commitment to the organisation (Turker, 2009a). As elucidated by social identity theory, CSR practices can boost employees' perceived self-worth and ability to develop a greater sense of belonging to a reputable organisation (Zhou et al., 2018), subsequently providing greater job satisfaction (Karanika-Murray et al., 2015). The stronger the organisational identification, the more strongly the individual will engage with the company's vision, mission, objective, and expectations (Dutton et al., 1994). Correspondingly, as guided by social exchange theory, the intangible social prestige and pride that build employees' organisational identification are anticipated to foster positive behaviour and attitude towards job satisfaction (Kim et al., 2020). With these means, we justified the intervening role of organisational identification between CSR practices and job satisfaction. Thus, we hypothesise:

Hypothesis 4a

Organisational identification significantly mediates the relationship between CSR to employee and job satisfaction.

Hypothesis 4b

Organisational identification significantly mediates the relationship between CSR to community and job satisfaction.

Hypothesis 4c

Organisational identification significantly mediates the relationship between CSR to environment and job satisfaction.

Demographic variables such as employee age have been shown to significantly influence job satisfaction in many studies. The unprecedented COVID-19 pandemic imposed extreme environmental disruptions that affected nearly all industries and aspects of daily life (Anker, 2021). The social isolation from colleagues and the workplace during the lockdown period have influenced employees’ distress levels, job satisfaction and performance (Kumar et al., 2021; Mahmoud et al., 2022). Nevertheless, recent studies conducted during the COVID-19 pandemic still demonstrated a positive correlation between age and job satisfaction. For instance, in a study involving 624 hospitality workers, the results indicated that older participants (40+) were more motivated to work and had a much lower turnover intention than younger colleagues (Bajrami et al., 2021). Likewise, a survey established among 302 teachers revealed that older individuals had higher job satisfaction and lower stress scores during the pandemic (Thakur et al., 2022). Even though older people are at higher risk of severe illness and mortality, a study conducted during the H1N1 pandemic in 2009 exhibited that younger people (20s) were more anxious and susceptible to infection at work (Matsuishi et al., 2012).

From our review, no empirical study to date has examined how employees' age influences the mediating effect of organisational identification between CSR practices and job satisfaction. Nevertheless, based on the above narrative, we foresee that linear moderated mediation should be positively associated. Studies show that older employees seem to have better psychological resilience to the problematic pandemic challenges and are more easily satisfied with their jobs than younger employees. We anticipate that CSR practices can better develop a sense of belonging (organisational identification) and job satisfaction as employees’ age increases. Thus, the following hypotheses are proposed:

Hypothesis 5a

Employee age strengthens the mediation effect of organisational identification between CSR to employee and job satisfaction.

Hypothesis 5b

Employee age strengthens the mediation effect of organisational identification between CSR to community and job satisfaction.

Hypothesis 5c

Employee age strengthens the mediation effect of organisational identification between CSR to environment and job satisfaction.

A summary of the development of these hypotheses is shown in Fig. 1 .

Fig. 1.

Summary of hypotheses development.

3. Methodology

This study employed a quantitative approach to address the research questions. As displayed in Fig. 1, the research framework included three independent variables: CSR to employee (representing internal CSR), CSR to community and CSR to environment (constituting external CSR); one dependent variable: job satisfaction; one mediating variable: organisational identification; and one moderating variable: employee age. The partial least squares-structural equation modelling (PLS-SEM) approach was used to assess the measurement model and analyse the path relationships of the latent variables. Then, the PROCESS algorithm was used to examine the moderated mediation relationships between CSR practices and job satisfaction through organisational identification, conditional on employee age.

Due to the limited time and financial costs, instead of a pilot test that required at least 30 pilot participants, this study attentively adopted all the survey instruments from peer-reviewed journal articles, which had undergone rigorous scale development procedures. A pretesting assessment was established to mitigate measurement error before the full-scale data collection. This ensured that the questionnaire was easy to understand, and the respondents could follow the instructions correctly. Six individuals from the target respondent pool participated in the pretesting. After a few simple amendments, the refined survey instrument was utilised for data collection. All the latent variables in the framework were reflective measurement models and had high internal consistency and reliability, with Cronbach's alpha exceeding the threshold value of 0.70. Specifically, the six measurement items of CSR to employee were adopted from Turker (2009a), with Cronbach's alpha of 0.90. On the other hand, the items for CSR to community and CSR to environment were from Farooq et al. (2014), with Cronbach's alpha of 0.79 and 0.91, respectively. The organisational identification measurement scale was adopted from Mael and Ashforth (1995) and Farooq et al. (2014), with Cronbach's alpha of 0.74 and 0.87, respectively. The four items for job satisfaction were adopted from Feng and Savani (2020), with Cronbach's alpha of 0.94. Moreover, respondents' demographic characteristics, such as age, were included as the moderating variables. The CSR to employee, CSR to community, CSR to environment, and organisational identification questionnaire were developed using a five-point Likert scale (1 = strongly disagree to 5 = strongly agree). In contrast, job satisfaction items used a seven-point Likert scale. Different scores increased the respondents' sensitivity and assisted in mitigating the common method bias (Podsakoff et al., 2012; Whitley & Kite, 2013). The latent variables and their measurement items are shown in Table 2 . Subsequently, PLS-SEM was used to analyse the direct and mediating causal relationships between the predictors and criteria, followed by the PROCESS algorithm on the moderated mediation effect.

Table 2.

Research constructs and items.

Latent variable	Items
a)Internal CSR – CSR to employee (CEP) (Turker, 2009a)
CEP1	Encourage voluntary activities.
CEP2	Skills training and career development.
CEP3	Take care of employees' needs.
CEP4	Promote work and life balance.
CEP5	Distributive justice.
CEP6	Provide educational support.
b)External CSR – CSR to community (CCO) (Farooq et al., 2014)
CCO1	Charity contributions.
CCO2	Support NGOs for social welfare.
CCO3	Take care of social well-being.
c)External CSR – CSR to environment (CEV) (Farooq et al., 2014)
CEV1	Improve and protect nature.
CEV2	Create a better life for future generations.
CEV3	Minimise environmental pollution.
CEV4	Target sustainable growth.
d)Organisational identification (OI) (Farooq et al., 2014; Mael & Ashforth, 1995)
OI1	Feeling when someone criticises the organisation.
OI2	Interest in the organisation's image.
OI3	Treat the organisation as their own
OI4	Want the organisation to be successful.
OI5	Likes the organisation to be praised.
e)Job satisfaction (JS) (Feng & Savani, 2020)
JS1	Meaningful job.
JS2	Like the work.
JS3	Find pride in the job.
JS4	Enjoy the job.

3.1. Research context and sampling population

The COVID-19 outbreak has enormously impacted the economy and business environment. The COVID-19 pandemic hit Malaysian shores in January 2020, resulting in the government having to issue a Movement Control Order (MCO) that restricted the movement of persons unless they were engaged in essential services (e.g., food and beverages, health care and medicine, and security and safety defence). Some industries, such as tourism and aviation, have experienced massive losses and downturns due to the continuous MCOs and nationwide lockdowns. Contrastingly, other sectors, such as technology and medical devices, received colossal and unprecedented market demands, and some hit new record highs on the stock markets. This study chose the medical device industry, which is among the best performers globally and in Malaysia's economy. Under the Eleventh Malaysia Plan 2016–2020, the medical device industry has been identified as one of the most dynamic and high-potential industries. According to the Ministry of International Trade and Industry (MITI), around 30 multinational companies (MNCs), including a few of the world's largest medical device companies from Europe, contribute to Malaysia's offshore manufacturing operations.

The list of medical device companies was obtained via MITI and the Association of Malaysian Medical Industries (AMMI) websites. Next, targeted companies' information was acquired from their official websites. The survey was restricted to staff at the managerial or professional level, such as engineers, executives, supervisors, managers, and officers, to ensure that the respondents were exposed to their company's overall direction and CSR initiatives. The quantitative data was collected through an online survey to avoid physical contact. A total of 236 valid responses were received from March to April 2021. As the sample size exceeds the minimum threshold value (103) requested by the G* Power programme (effect size 0.15, significant level 0.05, power 0.80), the analysis generated a statistically significant result.

The sampling population represents a diverse demographic of multinational medical device companies in Malaysia. As shown in Table 3 , from the total 236 valid responses, 39.8% respondents were male and 60.2% were female. Nearly half of the respondents (49.1%) were aged 31–40, followed by those aged 41–50 (22.5%), 21–30 (17.4%), and 51 and above (11.0%). The survey was restricted to employees at the managerial or professional level. As a result, eleven (4.7%) participants held director or senior management positions, 146 (61.9%) were at executive, engineer, or officer level, 56 (23.7%) held managerial positions, and 9.7% held other positions. In terms of department or function, 26.3% of respondents were from the engineering department, while the remainder were from the quality (20.3%), supply chain and planning (15.7%), production (7.2%), finance (7.2%), human resources (3.4%), and other (19.9%) departments. The largest group in terms of working experience comprised those with 6–10 years’ experience (27.5%), followed by 11–15 years (19.5%), 16–20 years (17.8%) and 1–5 years (15.7%). Almost all (97.5%) respondents were educated beyond high school level. This figure included diplomas (6.4%), undergraduate (73.3%) and graduate (17.8%) certificates. These indications are associated with the standard qualifications necessary for a particular position in this field.

Table 3.

Respondents’ profile (n = 236).

Profile	n	%
Gender
Male	94	39.8
Female	142	60.2
Age
21-30	41	17.4
31-40	116	49.1
41-50	53	22.5
51 and above	26	11.0
Position
Director/Senior management	11	4.7
Manager	56	23.7
Executive/Engineer/Officer	146	61.9
Other professionals	23	9.7
Department/Functions
Engineering	62	26.3
Quality	48	20.3
Production	17	7.2
Human resource	8	3.4
Supply chain and planning	37	15.7
Finance	17	7.2
Others	47	19.9
Working experience
1–5 years	37	15.7
6–10 years	65	27.5
11–15 years	46	19.5
16–20 years	42	17.8
21–25 years	19	8.1
26 years and above	27	11.4
Education level
Secondary	6	2.5
Diploma	15	6.4
Degree	173	73.3
Postgraduate	42	17.8

3.2. Validation and reliability

The reliability and validity of the measurements are shown in Table 4 . The average variance extracted (AVE) for all latent variables was more than 0.50, with a range between 0.522 and 0.837. The loadings of three indicators, CEP1 (0.573), CEP2 (0.613), and OI2 (0.664), were slightly lower than the threshold value of 0.70. Nevertheless, they were higher than 0.40. As the AVE for all the related latent variables was higher than 0.50, the items were retained for their contributions to the content validity (Hair et al., 2017). Composite reliability (CR) was used to assess the internal consistency reliability. All variables showed values above the threshold of 0.70, ranging from 0.866 to 0.954, as displayed in Table 4. Heterotrait-monotrait ratios (HTMT) were used to assess discriminant validity. No collinearity issue existed among the predictors, as all of the variance inflation factor (VIF) values were below the limit of 5 (Hair et al., 2017). All construct pairs had HTMT values lower than the threshold value of 0.85 (Henseler et al., 2015), ranging from 0.357 to 0.820. Hence, the discriminant validity of the variables was supported by the HTMT results.

Table 4.

Reliability and validity measures.

Indicators	Construct	Convergent validity		Internal Consistency Reliability		Collinearity statistics
		Loading (>0.70)	AVE (>0.50)	Cronbach's Alpha (>0.70)	Composite Reliability (>0.70)	VIF (<5.0)
JS1	Job satisfaction (JS)	0.892	0.837	0.935	0.954	3.110
JS2		0.927				4.266
JS3		0.929				4.164
JS4		0.911				3.511
OI1	Organisational identification (OI)	0.702	0.572	0.812	0.869	1.473
OI2		0.664a				1.409
OI3		0.754				1.554
OI4		0.820				1.837
OI5		0.830				2.005
CEP1	CSR to employee (CEP)	0.573a	0.522	0.815	0.866	1.290
CEP2		0.613a				1.395
CEP3		0.816				1.993
CEP4		0.749				1.578
CEP5		0.796				1.834
CEP6		0.755				1.680
CCO1	CSR to community (CCO)	0.863	0.733	0.819	0.892	1.836
CCO2		0.866				1.832
CCO3		0.841				1.797
CEV1	CSR to environment (CEV)	0.763	0.687	0.848	0.898	1.625
CEV2		0.842				1.844
CEV3		0.839				2.190
CEV4		0.868				2.432

^aNote: No item was removed from the measurement (Hair et al., 2017).

4. Results

Table 5 shows the descriptive statistics and correlation coefficients of the variables. The correlation results reviewed all hypothesised and non-hypothesised relationships between all the variables. The three components of CSR, namely CSR to employee, CSR to community and CSR to environment, correlated positively. They also correlated positively with organisational identity and job satisfaction. In the matrix, the highest positive correlation was gained between CSR to employee and CSR to environment (0.684). In contrast, employee age only correlated with CSR to community, but not with other constructs. The data indicated the absence of any multicollinearity problem, as the absolute correlation coefficient among constructs was less than 0.70.

Table 5.

Descriptive statistics and correlation matrix (N = 236).

Variable	Mean	Std. Dev.	1	2	3	4	5
1. CCO	−0.00016	1.002221
2. CEP	0.00005	1.002180	0.599**
3. CEV	−0.00002	1.002166	0.684**	0.604**
4. JS	−0.00006	1.002048	0.314**	0.555**	0.369**
5. OI	0.00000	1.002176	0.507**	0.606**	0.442**	0.544**
6. AGE	−0.00003	1.002130	−0.114*	−0.056	−0.013	0.049	−0.029

Note: CEP = CSR to employee; CCO = CSR to community; CEV = CSR to environment.

OI=Organisational identification; JS = Job satisfaction; AGE = Employee age.

Correlation is significant at *p < 0.05 and **p < 0.01.

This study refers to three statistical analysis standards to decide whether to reject the null hypothesis: p-value (<0.05), t-statistic (>1.65 or >1.96) and confidence intervals (+/+ or −/−) (Hair et al., 2017). The statistical findings of path coefficients are shown in Table 6 . The path coefficient of CSR to employee (CEP) to job satisfaction (JS) was statistically significant, with a mean of 0.409 (p = 0.000). Thus, this result supports hypothesis H1. Unexpectedly, CSR to community (CCO) also had a significant but negative impact on JS, with a mean of −0.167 (p = 0.004). The finding contradicted hypothesis H2. Thus, H2 is not supported. The path coefficient for CSR to environment (CEV) towards JS was insignificant, as the p-value was larger than 0.05. Hence, H3 is not supported. Meanwhile, the mediation effects of organisational identification (OI) were significant for hypotheses H4a and H4b, with means of 0.172 (p = 0.001) and 0.081 (p = 0.004), respectively, but not for H4c. Thus, H4a and H4b are supported, but H4c is not. The results indicate that the employee age (AGE) index values for moderated mediation effects were insignificant. The findings revealed that AGE did not significantly influence the mediating impact of OI between CEP and JS, CCO and JS, and CEV and JS, respectively. Thus, H5a, H5b and H5c are not supported.

Table 6.

Hypothesis testing.

Hypothesis	Relationship	Mean	Std Dev	t-value (>1.65)	p-value(<0.05)	Confidence Intervals		Supported?
						Low	High
H1	CEP→JS	0.409***	0.099	4.143	0.000	0.241	0.570	Yes
H2	CCO→JS	−0.167**	0.062	2.651	0.004	−0.271	−0.065	Opposite
H3	CEV→JS	0.080	0.080	1.000	0.159	−0.050	0.218	No
H4a	CEP→OI→JS	0.172**	0.052	3.246	0.001	0.094	0.260	Yes
H4b	CCO→OI→JS	0.081**	0.031	2.647	0.004	0.038	0.141	Yes
H4c	CEV→OI→JS	0.007	0.031	0.186	0.426	−0.042	0.060	No
H5a	AGE*CEP→OI→JS	−0.048	0.036	1.317	0.094	−0.112	0.005	No
H5b	AGE*CCO→OI→JS	0.002	0.032	0.073	0.471	−0.048	0.055	No
H5c	AGE*CEV→OI→JS	0.023	0.043	0.044	0.330	−0.039	0.099	No

Note: CEP=CSR to employee; CCO=CSR to community; CEV=CSR to environment.

OI=Organisational identification; JS = Job satisfaction; AGE = Employee age.

Significant level *p < 0.05, **p < 0.01 and ***p < 0.001.

The model's predictive power of endogenous constructs was examined using the coefficient of determination (R²). The mean R² of job satisfaction was 0.430 (43.0%). Therefore, JS is considered to be substantially explained by the predictors, as the R² has exceeded 26% (Cohen, 1992). The predictors also substantially explained organisational identification, with an R² value of 0.428 (42.8%). Predictive relevance (Q²) was utilised to assess the predictive capability of the endogenous latent variables. The Q² value for job satisfaction was 0.310, while the Q² value for OI was 0.391. Since the Q² of both variables are larger than zero (Hair et al., 2017), all endogenous constructs in the model have substantial predictive relevance. According to the exogenous effect size f² standard proposed by Kenny (2018), the variable that contributed a large effect size f² on JS was CEP (f² = 0.138), followed by CCO, with a medium effect size (f² = 0.022), and CEV, with a small effect size (f² = 0.005). The effect size analysis is consistent with the results from the hypothesis testing.

5. Discussion

As discussed in the previous section, this study aims to contribute to the management literature in the work satisfaction field based on social exchange and social identity theories in a new pandemic reality. Specifically, we sought to understand how the direct impact of different CSR types influences employees' job satisfaction in the medical industry during the pandemic. Furthermore, we know very little about the intermediary mechanisms between CSR and employee satisfaction (Jones et al., 2017). Based on previous studies, this study also examined the indirect effect of CSR on job satisfaction through the mediation of organisational identification (e.g., El Akremi et al., 2018; Farooq et al., 2014; Farooq et al., 2017) and moderated mediation of employees’ age (e.g., Bajrami et al., 2021; Matsuishi et al., 2012; Thakur et al., 2022). In the following, we discuss the analysis findings for the direct and indirect effects.

5.1. Direct effects

The findings showed that internal CSR (CSR to employee) positively affects job satisfaction, aligning with most of the review findings in Europe (Macassa et al., 2020). When the company is concerned about its employees’ needs and benefits, in exchange, the staff will demonstrate favourable behaviour, such as job satisfaction, towards their company and job. Hence, in line with social exchange theory, the request for employers to be vigilant towards their workers should be much higher in unprecedented circumstances (Cropanzano & Mitchell, 2005). The employees would appreciate it if the company could advocate for “people before profit” during a healthcare crisis. However, we cannot compare this result with relevant studies such as the works of Hameed et al. (2016) and Farooq et al. (2017), which also coined the CSR to employee variable, as they did not include job satisfaction in their research framework.

The analysis has shown several unexpected results regarding the relationship between external CSR and job satisfaction. The influence of CSR to community was significant but in the opposite direction to past empirical results (e.g., Barakat et al., 2016; Brammer et al., 2007; Vlachos et al., 2013). The backfire from the employees might be due to external CSR, which increases the employees’ workload. During the pandemic, the workload of employees in the medical sector was comparatively heavy. External CSR to the community might also be a source of anxiety, as Coronavirus is still widespread in the region. In addition, some respondents might opine that the company should reserve more capital instead of spending it on window dressing or greenwash (Lea Cassar, 2018) during uncertain business circumstances.

Nevertheless, CSR to environment showed no statistical impact on employees’ job satisfaction. The results did not agree with the original hypothesis, possibly due to two main reasons. First, environmental awareness is not mainstream among the general public, including professionals in the medical industry. There is a deficiency of long-lasting national environmental policies and investment in this country. Second, the finding implies that the respondents might not perceive ecological conservation as an essential responsibility of a private company, especially during threats such as a pandemic crisis. Some respondents might not be interested in or pay attention to such “public relations” activities. However, further research may need to clarify the reasons behind this.

5.2. Indirect effects

Based on the results, organisational identification demonstrated three types of mediation (Hair et al., 2021). First, it was a “complementary or partial mediator” between CSR to employee and job satisfaction, given that its indirect and direct effects pointed in the same direction. Second, it was a “competitive mediator” between CSR to community and job satisfaction, given that the indirect and direct effects were significant but pointed in opposite directions. Last, it showed “no-effect non-mediation” between CSR to environment and job satisfaction, in that neither the direct nor the indirect effect was significant.

From the viewpoint of social identity theory, organisational identification should be an essential intermediate variable that could justify several of the inconsistencies between CSR and job satisfaction arising in previous studies (e.g., Kim et al., 2020; Mael & Ashforth, 1995; Turker, 2009a). The hypotheses have been statistically supported for antecedents such as CSR to employee and CSR to community. A sense of belonging and organisational pride would develop when the company continuously practises social responsibilities towards its employees and society. Hence, the organisational identification driven by the two CSR initiatives would be transferred into higher job satisfaction, as illustrated by social exchange theory and past research (e.g., Kim et al., 2020; Mael & Ashforth, 1995).

Remarkably, the influence of CSR to community on job satisfaction was only positively correlated when mediated by organisational identification. In other words, if the employees do not identify as part of the organisation or the CSR activities have no relationship, the impact on job satisfaction is entirely different or even hostile. Thus, employees' involvement and participation in CSR initiatives, such as community service and volunteering work, are essential. Nonetheless, social mobilisation is challenging to execute during a pandemic. Lastly, CSR to environment did not significantly impact organisational identification and job satisfaction among the employees. We presume that this might be due to this country's low environmental awareness and public ignorance. Specifically, Malaysia only earned a score of 35 and placed 130th among 180 countries in the Environmental Performance Index 2022 (Wolf et al., 2022). The findings from our analysis revealed that employee age has no significant effects on the mediation effects of organisational identification between the three CSR types and employee job satisfaction. Even though previous studies, such as those conducted by Bajrami et al. (2021) and Matsuishi et al. (2012), suggested that older workers might have lower stress scores and be more motivated to work than their younger colleagues, the moderating effect of age was not statistically significant. During the pandemic period, the indirect impact of internal and external CSR on job satisfaction via organisational identification was not associated with the employee's age.

6. Conclusion

To conclude, this study has made five significant conclusions. First, CSR to employee directly and indirectly impacts employee job satisfaction. Second, surprisingly, CSR to community negatively influences job satisfaction during the COVID-19 pandemic. Third, long-term and persistent CSR to community could build organisational identification and improve job satisfaction. Thus, critical community support from companies is still needed, but must be carefully planned and executed. Fourth, CSR to environment does not statistically affect job satisfaction in the medical device industry, neither directly nor indirectly. Last, during the pandemic era, employees’ age is not associated with the mediating effect of organisational identification among CSR types and job satisfaction.

6.1. Theoretical contributions

This study sheds light in the area of theory in several ways. Firstly, this research assimilates two prominent views, namely social exchange and social identity theories, to improve the interpretation of phenomena influencing employee job satisfaction during the COVID-19 pandemic (Leidner, 2020; Summers, 2001). Integrating views from different theories offers a more comprehensive framework and improves the remaining theoretical rationale, especially in a “new normal” business environment. Generally, the social exchange view facilitates the relationships between internal and external CSR and employee satisfaction from a broader perspective. Meanwhile, the social identity theory guides the influences of intermediaries (i.e., organisational identification, employee age) on the linkages between CSR types and job satisfaction. This research has evidenced and improved the remaining links with theoretical justifications and contributed to the body of knowledge of both theories.

Secondly, the study has broadened the theoretical boundaries by contextualisation (Suddaby, 2014; Whetten, 1989) of the selected constructs (i.e., CSR to employee, CSR to community, CSR to environment, organisational identification, and job satisfaction) conforming to the context of the medical industry in a developing country (i.e., Malaysia). As discussed, the CSR notion is highly fragmented and comprises various arguments. This study is consonant with the stakeholder theory narrative from some recent studies (e.g., Farooq et al., 2017; Hameed et al., 2016) that specify a firm's CSR practices into internal (or primary stakeholder: employee) and external (or secondary stakeholder: community and environmental) dimensions. This research challenges the conventional belief that assesses CSR-oriented activities using a single or consolidated dimension.

Thirdly, the research adds value to the theory development by mitigating literature gaps (Leidner, 2020; Summers, 2001) in previous studies. This study provides empirical evidence and theoretical interpretations for the findings that shed light on the inconsistent results (i.e., the relationship of CSR-oriented activities with job satisfaction) and untested theoretical relationships (i.e., the mediating effect of organisational identification, moderated mediation of employee age) in the past research. Lastly, even though the study does not introduce a new measurement tool, it extends the existing important scales and contributes to the enrichment of psychometric properties (Summers, 2001). Utilising the second-generation analysis technique of PLS and the PROCESS algorithm approach provides a more systematic and comprehensive understanding of the current research domain.

6.2. Practical contributions

The practical implications of this study are primarily focused on a mechanism for developing CSR strategies during the pandemic era. For a company to be socially responsible, it must first be accountable to its employees. The findings have shown that companies concentrating on internal CSR (i.e., CSR to employee) while the pandemic virus circulates would significantly increase employee job satisfaction. Corporations can turn a pandemic crisis into an opportunity to boost staff fulfilment. Top executives or CSR managers may refer to the measurements (Table 2) and the illustrations in this paper while drafting their CSR initiatives. Secondly, external CSR is crucial to support the community during the pandemic crisis but should be limited and carefully executed. High market demand and headcount control at the workplace have caused unprecedented pressure and heavy workloads, especially in the medical industry. External CSR activities would increase employees’ work burden and risk of virus exposure. In addition, companies should promote external CSR genuinely rather than as just another public relations show or greenwashing to prevent an adverse effect on the public and employees.

6.3. Limitations and future research

This research has potential biases and limitations. Firstly, common method bias might affect the findings, since the predictors and criterion data were collected from the same source (respondents). Thus, the survey has been kept specific and straightforward and ensured the participants' anonymity to mitigate the method variance. Using various measurement scales for different variables is one of the ex-ante approaches (procedure control) to minimise common method bias (Podsakoff et al., 2012; Whitley & Kite, 2013). Besides, the variance inflation factor (VIF) was below 5 (Hair et al., 2017), indicating that the common method bias was not significant. Secondly, nonresponse bias might affect the generalisability of the results. This research assessed the nonresponse bias in the data by establishing a multiple group analysis to compare respondents’ first and fourth quartiles (Armstrong & Overton, 1977). The path coefficient differences for all correlations were insignificant (p-value >0.05). Hence, the nonresponse bias was not significant in this study. Thirdly, the research findings are limited to the medical device industry in Malaysia. The results might differ in other sectors or in the same sector in other regions. Therefore, the proposed conceptual model should be tested in various research settings to increase its generalisability. Lastly, intermediating variables that clarify the relationship between CSR and job satisfaction are still lacking. Future empirical studies might include constructs such as distributive justice and organisational citizenship, to further enhance the framework.