Wellness programs in times of COVID-19, perceived organizational support and affective commitment: effects on employee innovative behavior

Abstract

Purpose

Innovation is one of the most important foundations on which to create and sustain competitive advantages in companies, but at the individual level, employee innovative behavior has recently been jeopardized by the situation caused by the COVID-19 pandemic (e.g. changes in workplaces, employee interaction, motivation). This study analyzes wellness programs and actions through which organizations have tried to adapt to the new situation caused by COVID-19 and their effect on employee innovation behavior.

Design/methodology/approach

Structural equation modeling by means of the partial least squares technique was used to test the study's hypotheses after collecting survey data from Spanish companies, providing evidence that wellness programs and measures to deal with COVID-19 through perceived organizational support and affective commitment encourage employee innovation behavior.

Findings

The results suggest that efforts developed by firms focused on employee well-being to overcome difficulties caused by the pandemic strengthen innovative behaviors by means of intrinsic motivation based essentially on personal commitment. Theoretical and practical implications of the findings are discussed by the paper's authors.

Originality/value

This paper corroborates and extends previous research regarding wellness programs, perceived organization support and affective commitment. It provides a comprehensive model of relationships that predicts employee innovative behavior. It analyzes the influence of enterprise wellness programs based on protective COVID-19 measures.

Citation

Muñoz, R.M., Andrade, S.M., Peña, I. and Donate, M.J. (2023), "Wellness programs in times of COVID-19, perceived organizational support and affective commitment: effects on employee innovative behavior", European Journal of Innovation Management, Vol. 26 No. 7, pp. 23-44. https://doi.org/10.1108/EJIM-02-2022-0072

1. Introduction

Innovation is one of the fundamental elements to creating and sustaining competitive advantages in companies (McGrath et al., 1996). As innovation capabilities are based on the combination and integration of individual and collective knowledge, their development and use pose important human resource management (HRM) challenges at different firm levels (Jafri, 2010). At the individual level, a firm should be able to maintain favorable relationships with employees and motivate their dedication to the company, in order for them to become further innovative (Carmeli et al., 2006; Jong and Den Hartog, 2007), thus increasing organizational performance (Sun, 2019).

The antecedents or determinants of employee innovative behavior have been studied thoroughly in HRM literature, but there are still important gaps to be filled, as these antecedents or determinants can be affected by, among other aspects, different contexts, levels of analysis and organizational incentives (Hughes et al., 2018). While the content-based approach in HRM considers “best practices” to encourage innovative behavior (Combs et al., 2006; Wright et al., 2005), a process-based approach has recently emerged to focus on psychological processes through which employees engage with companies' HRM practices, which may result in improved outcomes (Sanders et al., 2014). This article will try to clarify an existing knowledge gap in this process-based approach to explain how employees perceive signals that lead them to behave creatively and subsequently to develop and implement new ideas in the company. To our knowledge, there is a lack of research on the study of employees' perceptions about their role in innovation when they are stimulated by HRM initiatives (Sanders and Yang, 2016). This paper focuses particularly on the concept of perceived organizational support (POS) as a determinant of these innovative behaviors when an employee has previously been subjected to stimuli. POS can be defined as an employee's feeling on how the company values his/her contribution to performance and cares about his/her well-being (Eisenberger et al., 1986). This concept reflects the employee-company relationship from the perspective of employees and can have important benefits for both employees and employers. Sun (2019) describes the antecedents of POS by grouping them in individual and organizational variables, along with interactions generated between the company and its employees. Organizational variables include working conditions, which we will focus on herein, placing particular emphasis on enterprise wellness programs (EWP) and actions to adapt the organization to the situation caused by the COVID-19 pandemic.

The Organisation for Economic Co-operation and Development (OECD, 2019) stated that, in the case of Spain, wellness programs in the workplace could prevent up to 96,000 illnesses by 2050, saving millions of euros in healthcare costs, and increasing employment and productivity by an amount equivalent to 4,000 full-time workers per year. In the last few years, many companies have become aware of this fact, and they are increasingly focusing on developing well-being activities and illness-prevention policies for their employees (Maryam et al., 2016). In this context, a major challenge for companies is to improve their employees' health and to obtain organizational benefits from their well-being and work development (Pronk, 2014). The implementation of wellness programs is also a form of corporate social responsibility (CSR) focused on employees, which can lead firms to higher levels of employee satisfaction and commitment (Mirvis, 2012).

In addition to the analysis of companies' EWPs, this paper also considers a range of measures adopted by companies as a consequence of the COVID-19 pandemic that are likely to affect POS and innovative behaviors of employees. Given the extraordinary nature and impact of the pandemic, a main objective of this paper is to clarify the existence of a relationship between anti-COVID measures implemented by companies in addition to EWPs and POS. To our knowledge, these relationships have not yet been analyzed, and the results of the study can thus be useful for managers looking to improve decision-making processes in companies especially affected by the COVID-19 pandemic, which, in most industries, has substantially changed working conditions (e.g. new tools, workplace environment) and interaction among people.

Moreover, this paper introduces the concept of affective commitment as a mediating construct between POS and innovative behavior. From cognitive collective engagement (Fachrunnisa et al., 2020) and job demands-resources theoretical approaches (Demerouti et al., 2001), we posit that high levels of POS lead employees to show high levels of affective commitment, which is reflected in their innovative behaviors. Although the relationship between POS and employee performance (mostly well-being), from a theoretical perspective of organizational support, has been widely studied in the past few years (Kurtessis et al., 2017), to our knowledge, the effect on employee innovation behaviors mediated by psychological variables such as affective commitment has not yet been analyzed. It is important that this relationship be understood by managers since the effects of the COVID-19 pandemic have created a new scenario in which innovative behaviors of employees are needed for firms to adapt working conditions and human connection effectively in order to meet new challenges (Diab-Bahman and Al-Enzi, 2020).

The objective of this study is thus to analyze two unexplored relationships in the HRM literature: (1) the effect of enterprise wellness programs (EWP) along with COVID-19 measures on POS and (2) the mediating role that affective commitment plays in the relationship between POS and employee innovation behavior. Overall, this study develops a theoretical model integrating these relationships and testing them empirically in a sample of employees especially affected by the COVID-19 pandemic and the transformation of their workplaces. Literature typically suggests that organizational support, as perceived by employees, improves organizational commitment (Kwak et al., 2010). We thus try to contribute to the literature on HRM by providing a comprehensive model of relationships that predicts employee innovative behavior by means of the implementation of EWP programs based on protective COVID-19 measures, POS and their effects on employee affective commitment.

This paper is structured as follows. The following section presents a literature review and develops the research hypotheses of the study. The methodology section describes data collection methods and the measures of the variables. Thereafter, we explain the statistical technique used and show and discuss the obtained results. Finally, we put forward the conclusions, the key contributions and the limitations of the study, as well as suggesting new directions for future research.

2. Theoretical background and hypotheses

2.1 Employee wellness programs and their organizational adaptation in workplaces to the COVID-19 pandemic

Well-being is a complex construct that is considered by current research from two general perspectives: hedonic, which focuses on happiness and defines well-being in terms of pleasure attainment and pain avoidance; and the eudaimonic approach, which focuses on meaning and self-realization and defines well-being in terms of the degree to which a person is fully functioning (Ryan and Deci, 2001). The second approach includes characteristics closely linked to a work environment, that is, the participation in activities that lead to human growth, such as purpose in life, personal growth, self-acceptance, autonomy, mastery or positive relatedness (Ryff and Keyes, 1995). Well-being at a job level indicates a subjective perception of general satisfaction and positive feelings toward work (Keeman et al., 2017). This second perspective is adopted in this paper, as we try to delve more deeply into the organizational context surrounding employees and its improvement via HRM actions.

Julander (2014) classifies the interventions on wellness programs on three levels. The first one focuses on generating awareness, and it consists of actions/instruments such as newsletters, health fairs, health screenings, posters or flyers. The second level describes actions oriented to modifying employees' lifestyle (healthier habits) and other health-related behavior, and may include physical activity programming, health-related lectures, guidance on the correct performance of physically demanding jobs or individual and group counseling. The third level attempts to create a supportive environment through a healthy lifestyle and behavior, which may encompass the provision of resources and spaces for on-site fitness centers, health coaching, support groups, worksite cafeteria adaptation or policies reinforcing positive health-related behaviors (e.g. non-smoking facilities, options for benefit plans).

There are a wide range of empirical studies analyzing the impact of EPW on outcome indicators. These can include productivity, since it has been observed that the level and the quality of production are higher (Loeppke et al., 2008), and certain labor costs, since these are significantly reduced (Baicker et al., 2010), mainly in the case of those associated to absenteeism (van den Heuvel et al., 2005). These indicators could also be profitability (Goetzel et al., 2005), job satisfaction (Barr-Anderson et al., 2011), employee engagement (Saks and Gruman, 2011) and work environment improvements (Jack and Brewis, 2005). In a 3-year prospective study of more than 300 employers, Schwatka et al. (2018) showed the potential benefits of EWP for organizations of all sizes. If well implemented, wellness programs can thus be highly beneficial for both the company and the employee.

This paper focuses on wellness programs oriented to maintaining a safe environment for employees in workplaces, that is, the employee perception of the importance and prioritization of workplace safety (Zohar and Polachek, 2014). A meta-analysis research has already shown that a safe environment is positively linked to workplace safety behaviors, thereby improving employee performance (Christian et al., 2009). By investigating the applicability of a safe environment, Beus et al. (2017) identify seven indicators for this concept that include a company commitment to safety, safety communication, safety training, coworker safety practices, safety equipment and housekeeping, safety involvement and safety rewards. These indicators show a company's commitment to safety in the workplace.

In the last few months, countries and regions have slowed down the extension of COVID-19 quarantines, lockdowns are being lifted and business activity is recovering while vaccination levels spread all over the world. However, the current pandemic is likely to last for some time, and the possibility of community transmission in the workplace should continue to be monitored (Yuan et al., 2021). Companies need to learn an important lesson from this situation: they should find effective ways to mitigate workplace health and safety threats from a proactive and holistic perspective. Hence, the concept of a safe environment should also include practices that companies have recently developed to deal with the threats arising due to the COVID-19 pandemic. These practices, along with traditional wellness programs, should have an impact on employee-perceived organizational support, as we analyze next.

2.2 Wellness and COVID-19 programs, and perceived organizational support

Organizational support theory claims that positive or negative perceptions of employees about how a company assesses their work performance and cares about their well-being have an important impact on employee outcomes (Giorgi et al., 2020). In this regard, POS reflects an employee's feeling on how much the company cares about his/her well-being (Eisenberger et al., 1986), becoming an instrument by which employees can improve and increase their motivation (Christian et al., 2009).

In the context of the social exchange theory, POS can be understood as the result of an exchange between the efforts of a company and of its employees which generates a commitment to reciprocate for both parties (Sungu et al., 2019). POS, therefore, includes a series of emotional, instrumental and informative transactions from maintaining to enhancing employee well-being (Chen et al., 2009). POS constitutes a major organizational “pillar” that encourages employees to identify with the company's objectives by generating positive personal feelings (Shoss et al., 2013).

Companies have different alternatives with which to increase POS levels, such as improving their employees' working conditions, establishing participation mechanisms, using the emotional salary as an element of the remuneration policy or even creating a favorable work environment based on interpersonal relationships and family support policies (McCarthy et al., 2013; Sun, 2019; Wayne et al., 2002). Depending on their design, the implementation of wellness programs is also likely to have an impact on POS (Pronk and Kottke, 2009): these should favor employee participation, respond to specific needs, promote personal relationships and have a component of intrinsic motivation. Research on enterprise wellness programs shows that the participation in an organizational wellness program, especially if comprehensively designed (not only referring to fitness), is associated with decreased absenteeism and increased job satisfaction on the part of employees (Parks and Steelman, 2008).

Regarding the situation caused by the COVID-19 pandemic, we understand environmental safety as employee perceptions of the importance and prioritization of workplace safety through comprehensive wellness programs (Zohar and Polachek, 2014). In this research, we have adapted this definition to include employee perception about a company's actions implemented to avoid risks associated with COVID-19. Given the great social pressure generated by the pandemic, companies need to pay more attention to employees' well-being to enhance their POS, which ensures organization recovery from the effects of the crisis (Lee et al., 2021). Overall, by generating a context of well-being for employees in workplaces, these programs will have a significant influence on employee POS. We thus formulate the following hypotheses:

H1.

The implementation of wellness programs is positively related to POS.

H2.

The implementation of programs to adapt the organization to the situation caused by COVID- 19 is positively related to POS.

2.3 POS and employee innovative behavior

Rhoades and Eisenberger (2002) suggest that organizational support based on HRM programs contributes to improving employee satisfaction through higher levels of employee recognition, job security and autonomy. If employees perceive that their efforts are recognized by the company, they are likely to develop productive attitudes and carry out their tasks in a highly productive manner (Tsai, 2013). Wang et al. (2018) indicate that POS leads employees toward achieving organizational goals, enabling them to become emotionally attached to the company. POS would thus manifest itself in a positive feeling for an employee with regard to their work environment, reducing stress levels and giving him or her the opportunity to develop further creative behaviors (Eisenberger et al., 1986). POS would encourage employees to get involved in innovation activities by being more willing to exchange personal knowledge and ideas (Nonaka, 1994).

Social exchange theory explains the behavior of individuals engaged in exchange processes within a social system (Blau, 1964), suggesting that interactions between the participants within a social system represent two-way reciprocal exchanges (of something valuable) and are based on expectations of potential rewards from others (Molm et al., 2007). Social exchange theory typically distinguishes between two types of reciprocity orientation: positive, which “involves the tendency to return positive treatment for positive treatment”; and negative, which “involves the tendency to return negative treatment for negative treatment” (Cropanzano and Mitchell, 2005, p. 878). From an organizational point of view, there will be a tendency to share knowledge and ideas by an employee when he or she realizes that his or her colleagues are also sharing their knowledge, such interaction being easier when employees feel that the company is rewarding these behaviors by creating a favorable environment for them. Efforts at an organizational level to develop POS will have individual responses reflected on knowledge sharing (or at least not knowledge-hiding behaviors), allowing a company to develop innovation capabilities (Donate et al., 2022).

Social exchange theory thus suggests that innovation requires organizational support by creating “good” environments that allow people to interact and exchange resources at an individual level, with this being perceived as a natural consequence of proactive management (POS) which fosters innovative behaviors. The job demands-resources theory (Demerouti et al., 2001) is also useful as it can present POS derived from wellness programs as a resource to counterbalance job demands (e.g. stress caused by COVID-19), generating positive feelings that favor knowledge exchange. Moreover, the theoretical perspective of cognitive collective engagement (Fachrunnisa et al., 2020) suggests that personal perceptions of support policies by firms lead employees to cognitively engage with the company's strategic intent, motivating them to actively search for new ideas to solve problems and overcome existing difficulties (i.e. proactive behaviors). In this regard, Madjar et al. (2002) found that employee perception of both personal (nonwork) and organizational support contribute significantly to creative performance, with it being essential to have a purpose which leads employees to make the most of their personal capabilities by actively searching for nonconventional solutions and developing new lines of work. From these considerations we formulate the third hypothesis of this study:

H3.

POS is positively related to employee innovative behavior.

2.4 Perceived organizational support, employee innovative behavior and the mediating role of affective commitment

The perceived effect for employees of implemented HRM programs does not always concur with the real actions developed by the firm (Elorza et al., 2011; Meyer and Smith, 2000; Xi et al., 2016). Our study aims to bridge HRM and person–environment fit literature by examining the possible mediating role of employee affective commitment in the relationship between employee perceptions of a broad set of wellness practices (POS) and employee innovative attitudes and behaviors. Alfes et al. (2013) point out that positive employee behaviors are not the effect of simply perceiving the implementation of HRM practices, but that there is a bridge between perceptions and whether or how these perceptions lead employees to develop particular behaviors.

Commitment is an attitude that reveals employee identification with the company's objectives and the willingness of employees to endeavor to achieve such objectives. Meyer and Allen (1987) identified three forms of organizational commitment: (a) affective commitment (AC), which refers to the emotional connection of employees with the organization, their identification with the company and their participation in company activities; (b) normative commitment, that is, loyalty as a means of responding to the company for everything that it has provided; and (c) continuance commitment, that is, employees' awareness of losses if they were to leave the organization.

The AC dimension has been widely explored and found to be consistently associated with employee relevant outcomes (Meyer et al., 2002). Previous research has found that AC is strongly related to prosocial behaviors, due to its roots in the identification with and internalization of organizational values. When employees feel valued, they have a predisposition to reciprocate with greater commitment, through attitudes that sometimes exceed job requirements (Demerouti et al., 2001). Hence, positive behaviors arise, which are reflected in highly creative attitudes to problem-solving (Ficapal-Cusí et al., 2020).

In this regard, Battistelli et al. (2019) indicate two main reasons due to which commitment can have an impact on innovative work behavior through the improvement of knowledge-sharing processes. First, affectively committed employees identify strongly with an organization's values and goals, which is reflected in extra efforts being made to complete tasks and activities in their work (Meyer et al., 2004), being able to actively explore, share and implement new ideas to assist the firm in achieving its objectives (Vinarski Peretz and Carmeli, 2011). Second, commitment involves positive affective experiences (Battistelli et al., 2013) that can make it easier for employees to develop individual innovation (De Dreu et al., 2008). Therefore, positive affection fosters the creation of novel ideas and solutions by enabling flexible and divergent thinking, encouraging knowledge search and implementing creative ideas (George and Zhou, 2007). Moreover, Scott and Bruce (1994) suggest that innovative behaviors require favorable work climates oriented to developing employee engagement, while Janssen (2001) found a positive relationship between employee perceptions of fair rewards and innovative work behaviors.

Therefore, affective commitment fosters a sense of belonging and is generally related to an employee who is emotionally attached to the organization. Such individuals exhibit greater capabilities to be involved in the activities of an organization and are always ready to put in extra effort beyond their duty toward the achievement of the organization's goals. Consequently, employees who have a greater level of commitment toward their organization are likely to come up with creative solutions to work-based problems, and as such show a greater tendency toward innovative behaviors (Jafri, 2010). These arguments lead us to formulate the fourth hypothesis of this study:

H4.

Employee affective commitment is positively related to innovative behavior.

From the development of H3 and H4, we establish a new connection consisting in a mediating effect of affective commitment in the relationship between POS and perceived employee innovative behavior. Studies about employee engagement claim that highly committed employees invest great efforts in seeking new ways of accomplishing their tasks or in changing and improving their environment (Ramamoorthy et al., 2005). Commitment leads to proactive behavior, proactive knowledge search and learning goals, which can be considered as innovative behavior (Christian et al., 2011). But affective commitment is created from the efforts made by firms to develop a positive context (i.e. workplace), which leads employees to perceive that there is strong organizational support from the organization's managerial team. Employee affective commitment is thus assumed to be generated from perceived efforts in managerial actions affecting employee well-being (Shore and Shore, 1995).

From the job demands-resources perspective, Schaufeli and Bakker (2004) place emphasis on the “active and favorable feeling” of personal commitment, so employees are energetic and excited about work and are ready to exert reasonable efforts to accomplish their work-related goals and objectives (Macey and Schneider, 2008). Previously, we have hypothesized the direct effect of POS on innovative behavior, but POS would also be connected to innovative behavior via affective commitment. By generating cognitive engagement reflected in affective commitment, POS would indirectly encourage the development of employee innovative behavior. The hypothesis is formulated as follows:

H5.

Employee affective commitment mediates the relationship between perceived organizational support and innovative behavior.

The model of this research is shown in Figure 1. We suggest that wellness programs and measures implemented by the company to prevent COVID-19 infections influence POS, which generates higher levels of employee affective commitment and, as a result, an increase in the level of an employee's innovative behavior.

3. Methodology

3.1 Data collection and sample

Data to test the hypotheses were collected by means of a survey in 2021. The questionnaire was designed following an extensive literature review, including questions, among others, referring to HRM practices, wellness programs, employee commitment, engagement, organizational support and innovation. We targeted our research on firms that had developed programs of health and well-being for employees before the COVID-19 pandemic, as they seemed more willing to implement measures related to protecting their employees from the effects of the pandemic. We thus sent questionnaires to customer firms of Andrade Fitness limited Co., a company based in Ciudad Real, the capital of the Ciudad Real province in Spain (75,000 inhabitants), whose main activity is focused on providing wellness services to companies. After contacting with the companies (54), we explained to them the objectives of the research, and 24 firms agreed to collaborate in the research, giving us access to the e-mails of their employees (total employees, 874). The companies were SMEs (minimum employee number, 5; maximum, 125). After sending them the questionnaire (electronic version), we collected 137 valid responses (at least 80% of the questionnaire completed). Table 1 displays the descriptive statistics for the study sample.

A Harman test was applied to the questionnaire variables to assess the existence of common variance bias for the data set. An exploratory factor analysis (principal components with a varimax rotation) was performed by considering the five main constructs of the model (wellness programs, COVID-19 programs, POS, affective commitment and innovative behavior), with results showing the existence of five factors with eigenvalues above 1, explaining 65.76% of the total variance. As only 32.3% of the total variance is explained by the first factor, common variance does not seem to be a significant concern for the research (Podsakoff and Organ, 1986).

Finally, and to check for sample representativity, we ran T-tests to analyze differences between respondents (the sample) and the total target of firms, which showed nonsignificant results both for size (t = 0.723; p < 0.87) and age (t = 0.801; p < 0.71).

3.2 Measures

3.2.1 Wellness and COVID-19 programs

Companies' wellness programs were measured using a 13-item scale adapted from Wang et al. (2018) and Sonnentag and Pundt (2016). These authors describe activities implemented at worksites by companies regarding physical activity, a healthy diet and health care, not including public health programs on these aspects (e.g. “the company frequently organizes sport events for the employees,” “the company usually provides information and training related to healthy habits”). Items ranged from 1 (totally in disagreement) to 5 (totally in agreement).

Regarding COVID-19 measures, we adapted the scale developed by Beus et al. (2017) to measure safety commitment, which includes eight items representing actions implemented to avoid infections in the workplace. Participants were asked to give their opinion about safety in the workplace and managers' behaviors regarding COVID-19 (e.g. “the company is putting a strong emphasis on workplace safety during the pandemic”; “employees receive sufficient safety equipment such as face masks, hydroalcoholic gel, temperature monitoring, etc.”). Items ranged from 1 (totally in disagreement) to 5 (totally in agreement).

3.2.2 Perceived organizational support

POS was measured using a reduced version of the scale developed by Eisenberger et al. (1986), which included 10 items. The original scale took 36 items into account, but we considered only 10 for reasons based on the context of the research (i.e. SMEs, Spain, the COVID-19 situation). Our scale is unidimensional and showed high internal reliability, meaning that the use of a shorter format does not cause any validity problems (Rhoades and Eisenberger, 2002). The items try to reflect the employee belief that the company values his/her contributions and cares about their well-being (e.g. “the company tries to take care of its employees' well-being,” or “the company pays attention to its employees' interests and objectives”). Items ranged from 1 (totally in disagreement) to 5 (totally in agreement).

3.2.3 Affective commitment

Affective commitment was measured using the scale proposed by Meyer and Allen (1987) for organizational commitment, which includes three dimensions of commitment: affective, continuance and normative. We only used the affective commitment dimension, which comprises seven items which explore employee attachment, identification and emotional links to the company (e.g. “I feel that I am a part of this company”; “I would like to be in this company all my life”). Affective commitment is the most influential dimension of organizational commitment, as shown in previous studies (Harness et al., 2018; Meyer et al., 2002). Items ranged from 1 (totally in disagreement) to 5 (totally in agreement).

3.2.4 (Perceived) employee innovation behavior

Innovation behavior was measured by using a 4-item scale adapted from the scale developed by Carmeli et al. (2006). The items reflect the perceived ability of an employee to perform his/her work while solving problems creatively, it being essential that he/she develops skills that can be important in his/her future work (e.g. “I tend to propose innovative solutions to problems that arise in the course of my duties,” “I use a wide variety of sources/types of information to find an innovative solution to a problem”). As we put the questions directly to employees about their innovative attitudes, this is a perceptual scale, rather than an objective measure. Items ranged from 1 (totally in disagreement) to 5 (totally in agreement).

4. Results

The partial least squares technique for structural equation modeling (PLS-SEM) was used to test the hypotheses and analyze direct, indirect and mediating effects in our model. PLS-SEM is a statistical multivariate analysis tool that allows the researcher to model latent constructs, even when non-normality conditions for data are present and the sample size is small (Hair et al., 2012). In this research, the data set fulfills the condition that the sample size should be at least 10 times larger than the largest number of structural paths directed toward any construct (Chin et al., 2003). Descriptive information and correlations of the study variables are shown in Table 2.

PLS-SEM method was selected due to the following reasons: (a) it is an appropriate method in the early stages of new theory or if the study has a predictive nature (Hair et al., 2014), (b) it allows the researcher to design different causal relationships to be analyzed (Astrachan et al., 2014), (c) it is an appropriate method when the sample size is small (Henseler et al., 2015) and (d) it is well suited to models analyzing complex relationships (Hair et al., 2012).

The software applied was SmartPLS 3.3. PLS, based on an iterative algorithm to obtain the weights used to build linear combinations of the observed indicators as proxies for all the constructs in the model. The analysis of the research model must follow two steps. First, the measurement model is examined to assess the reliability and validity of the theoretical constructs, and, second, the structural model is estimated in order to examine the (path) associations hypothesized in the research model (Hair et al., 2014).

4.1 Measurement model

An indicator is considered adequate when it shows a loading (λ) above 0.7 for its respective construct. The loadings could also be considered if they are higher than 0.6 and significant (Benitez-Amado et al., 2015). To maintain convergent reliability, we deleted those indicators with loadings below 0.6. We used Cronbach's α, Rho_A and composite reliability (CR) (Table 3) to test construct reliability, all of which should be above the standard threshold of 0.7 (Fornell and Larcker, 1981). Convergent reliability was then checked by examining the loadings of the common factor constructs and the AVE of each latent variable. All the loadings were above the recommended threshold of 0.5 (Hair et al., 2014) (Table 3).

When evaluating discriminant validity, Fornell and Larcker (1981) suggest the criteria that considers that the square root of the AVE of a latent variable should be higher than the correlations between the other latent variables and the construct itself. As Table 3 exhibits (with the main diagonal showing the square root of the AVE of each construct), discriminant validity is supported in our model, since this condition is fulfilled for all the considered constructs.

Discriminant validity was analyzed first by comparing the loadings of the indicators of each latent variable with cross-loadings, meeting the recommended condition. We also checked the Fornell-Larcker matrix (Table 4), which consisted of a comparison between the AVE of the factors and the square of each estimated correlation between those same factors (Fornell and Larcker, 1981), showing discriminant validity. Finally, we calculated the ratios between the heterotrait-monotrait correlations, which resulted in a matrix containing values below 0.9 (Table 4).

4.2 Structural model

VIF (variance inflation factor) values (Table 4) suggest that collinearity is not a major concern for the structural model (Hair et al., 2014). Next, we analyzed the significance of the structural relationships by calculating the path coefficients through the PLS algorithm and bootstrapping (5,000 sub-samples, n = 137).

The results of the structural model analysis are provided in Figure 2, in which path coefficients are shown, along with their significance levels. The direct and indirect effects are shown in Tables 5 and 6. Overall, the results of the statistical model suggest that all the hypotheses are supported. EWP has a positive and significant effect on POS (β = 0.257, p < 0.01), so the more a firm promotes wellness programs, the higher the level of POS is. Similarly, the COV variable representing measures adopted to prevent COVID-19 has a positive and significant impact on POS (β = 0.468, p < 0.01). H1 and H2 are, therefore, supported. A strong and significant relationship is also found between POS and IB (β = 0.525, p < 0.01), so H3 is also supported. A significant relationship is also found between AC and IB (β = 0.216, p < 0.01), with H4 also being supported.

Regarding the mediating effect, we proceed as follows: first, we ran the model without indirect effects, that is, we only consider a direct effect of POS on IB. The path coefficient was 0.661 (p < 0.01). Second, we ran the model including direct and indirect effects by considering AC as a mediator between POS and IB. The indirect effect was positive and significant (β = 0.136, p < 0.05). The path coefficient of POS on IB was reduced to 0.525, with p < 0.01. As the indirect effect is significant and the direct effect drops substantially, this suggests the presence of a mediating effect. Moreover, there is an increase in the coefficient of determination from 0.437 to 0.465. Additionally, we calculated the VAF (variance accounted for) value, which indicates the weight of the indirect effect between two variables in relation to the total effect (of the relationship), which was 21%. This value indicates that a partial mediating effect exists in the considered relationship (Nitzl et al., 2016).

To determine the goodness of fit of the model we calculated the standardized root mean square residual (SRMR). Values for both the estimated and saturated model are 0.071 and 0.086, respectively, which are below the threshold of 0.10, established by Williams et al. (2009). As the orientation of this paper is basically predictive, we analyzed R² values and performed the Stone-Geisser test (Q²) to determine the relevance of structural relationships and the entire model. Values for R² were 0.390 (POS), 0.396 (AC) and 0.472 (IB), with p < 0.01. Following Falk and Miller (1992), these values should be above 0.1 for it to be considered that the model shows predictive validity. Moreover, Chin (1998) points out that positive values for Q² also reflect predictive validity. Calculated values were 0.239 (POS), 0.246 (AC) and 0.251 (IB), so we can consider that the structural relationships and the model show satisfactory predictive capacity.

Finally, nonlinear effects were examined for all the established relationships by means of quadratic functions (normal and inverted U-shaped relationships). All these relationships were nonsignificant (p > 0.1), so the established model that shows linear relationships would seem to be appropriate for our research.

5. Discussion and conclusions

The COVID-19 pandemic has brought major challenges for companies and their employees due to its broad extension and scope (Sinclair et al., 2020). In addition, the pandemic has extended the importance of health and safety to all industries, so actions on the part of companies worldwide to fight against the virus and maintain employee well-being are of utmost importance, as they affect the entire society.

This study has tried to explore relationships between variables that improve employee well-being in the workplace in difficult times for companies, industries and the entire society. In so doing, firms behave ethically and play an essential role as a social agent cooperating for the expected general welfare of the society (Carroll, 2016). H1 and H2 were formulated following the assumption that ethical firms are trying to improve employee well-being as a response to highly stressful work conditions and to the physical and psychological personal conditions of employees in the wake of the COVID-19 pandemic. Both hypotheses have been corroborated: when these practices (in terms of wellness and COVID-19 protection programs) are properly implemented, there is a recognition by employees of the company's efforts to behave fairly and by giving them the opportunity to carry out their work under reasonably good conditions.

The results are consistent with previous studies that have examined the impact of wellness programs on employees' perceptions of workplaces conditions, such as the safety environment (Giorgi et al., 2020; Zohar and Polachek, 2014). Previous literature on psychological conditions of employees highlighted the fact that wellness programs are a resource that counterbalances the demands of a job (Demerouti et al., 2001), compensating stress and burnout generated during the execution of tasks. As we have previously explained, POS includes a series of emotional, instrumental and informative transactions from maintaining to enhancing employee well-being (Chen et al., 2009). Thus, as expected, wellness and COVID-19 measure programs generate positive feelings on the part of employees, which are reflected in POS.

The next question we tried to address was whether POS has positive effects on employee behavior (H3). More specifically, we focused on innovative behaviors, as innovation is a fundamental issue for companies in an increasingly dynamic and global environment (Petit and Teece, 2021), and we found that a positive relationship exists between POS and perceived innovative behavior (H3). These findings are consistent with the organizational support theory. Wellness programs implemented by companies benefit both the employee and the company when employees are able to perceive a significant organizational support for their well-being (Keeman et al., 2017). Employees who perceive that their company cares about them develop innovative behaviors as a consequence of their engagement and commitment (Rockstuhl et al., 2020). The perception of an organization that provides well-being leads to creativity generation, which can provide solutions to overcome difficulties and problems, and to employees actively seeking opportunities and contributing more freely to the exchange of knowledge and ideas. The practical implementation of new ideas will also be perceived by the employee as tolerated by the firm on the basis of “signs” of organizational support and justice (Shore and Shore, 1995). From an individual perspective, one of the main assumptions of this paper has been that POS will channel employee efforts toward innovative behaviors via cognitive engagement with the company (Fachrunnisa et al., 2020).

We have also found a positive relationship between affective commitment and innovative behavior (H4). Employees actively search for new solutions to solve problems and different perspectives to approach difficulties when they are motivated and empowered (Alfes et al., 2013). The results are consistent with the existing literature, which highlights the connection between commitment and processes of development and implementation of new ideas by employees (e.g. George and Zhou, 2007; Janssen, 2001; Ramamoorthy et al., 2005). Affective commitment is reflected on an emotional attachment to the organization, which gives the employee an interest in the continuous improvement of the firm. Thus, affective commitment involves employee well-being when the organization obtains “good” results, so employees will be interested in searching for new innovative approaches that benefit the firm (i.e. reducing costs, improved customer service, improving product image).

Moreover, and considering the former relationships, we formulated a hypothesis proposing that affective commitment exerts a mediating effect on the relationship between POS and innovative behavior (H5). The mediating effect is partial, meaning that POS leads to employee innovative behavior by itself and through affective commitment. Our findings corroborate previous research in which employee affective commitment is a key channel through which POS might increase employee innovative actions (Alfes et al., 2013; Macky and Boxall, 2008; Ogbonnaya and Messersmith, 2019; Sanders and Yang, 2016). The indirect effect helps to explain employee innovative behavior, so both perceived organizational support and affective commitment should be HRM objectives for a firm when it is trying to create a positive context in which to improve innovation capabilities (Ficapal-Cusí et al., 2020). According to these results, managers could enhance a company's employee innovative behavior through POS and affective commitment, and POS could be increased by enhancing wellness and anti-COVID measures. These findings are important contributions of this paper to organizational psychology and HRM literature.

5.1 Theoretical and practical implications

The concept of innovative behavior derives from the so-called adaptive performance, that is, an individual's ability to adapt to dynamic work environments. A theoretical contribution of our study is that adaptive performance indicates that employees are facing uncertainty and unpredictable work circumstances arising from organizational restructuring, crisis situations or emergencies (Neal and Griffin, 1999). This construct includes as a dimension employee innovative behavior, which reflects the ability of an employee to solve new problems and develop creative solutions to handle unexpected or atypical situations. We have considered this variable especially appropriate to the situation originated by COVID-19.

The present research has examined employee discretionary behavior based on creative actions toward problem-solving. It includes a wide set of actions such as the ability to create new ideas, put them into practice and search for creative ways to develop them. We focus on the mutual gain perspective that considers shared benefits for both the company and employee when employees feel a positive balance between job demands and resources (Demerouti et al., 2001). The conclusion is that POS is positively linked to employee affective commitment, which, in turn, is linked to employee innovative behavior. This is an important theoretical contribution from a psychosocial labor risk point of view, as it identifies perceived organizational support as a stress-reducing factor that has an important influence on employee behavior and employee commitment to the company's strategic intent and objectives based on innovation.

We have implicitly assumed an ethical positioning as wellness programs and measures to protect employees from COVID-19 are voluntarily implemented by firms. From a stakeholder's perspective, employees are considered to be a relevant group whose emotional needs should be catered for based on their personal well-being (Trevino and Nelson, 2021). Benefits for both the firm and its employees are reciprocated when employees respond favorably, thus reinforcing ethics and helping the organization through a “virtuous circle” process that occurs when the company performs well financially, which encourages it to dedicate further financial resources to other worthy endeavors, such as wellness programs (Trevino and Nelson, 2021). Innovation is the result that is finally derived from well-being reflected in POS and affective commitment. The catalyst behind this process is likely to be well-being associated with organizational ethics (Valentine, 2014, p. 5). We thus contribute to HRM literature by explaining innovative behaviors that derive from practices implemented by firms from a CSR point of view.

This research shows the importance of employee wellness programs in terms of the practical benefits enjoyed by companies and individuals. The planning of these practices shows that the company is interested in its employees' well-being, which, in turn, can generate valuable returns for the organization. Wellness encompasses more than mere physical aspects. It also includes emotional and intellectual aspects, which must be incorporated into wellness programs developed by companies. Employees perceive these signals as a company's concern for them, and they will reciprocate through commitment and innovative attitudes. Companies should therefore pay attention to the way in which these HRM practices are implemented and communicated to employees (Sanders and Yang, 2016). It is thus necessary for managers and employers to recognize how their employees perceive HRM. When they feel these practices as consistent and consensual, HRM makes sense to them and will lead to improved attitudes and behavior. If employees do not understand the intentions of management, the effectiveness of these types of practices in developing affective commitment, and, as a result, innovative behavior, can diminish or may even disappear.

This study also contributes to the body of literature interested in the consequences of large-scale traumatic events for companies and employees. The pandemic has exposed many potentially dangerous business practices for employee health. Given the rapid changes that the highly contagious coronavirus has generated, in the future, managers should show a proactive attitude in planning actions related to health and safety at work (Yuan et al., 2021). We have learned that healthier employees are less likely to have a dangerous infection if they are infected by the virus. Enterprises with a culture of well-being are likely to have a healthier and more resistant workforce. Thus, a feeling of personal responsibility will be generated, which will encourage teamwork, engagement and collaboration (Fabius and Phares, 2021). As the results of this study have shown, the company commitment to the safety of their employees is positively related to important results and behaviors in the workplace.

5.2 Research limitations and future directions

These conclusions should be considered cautiously, as this paper is not free of limitations. First, we have used self-reporting measures that are prone to generating common-method bias. However, the chosen variables show a perceptional character which allows the use of employee self-reports without distorting the results. Second, the study focuses only on employee innovative behavior as a measure of employee performance. Additional variables representing other measures could be incorporated into the proposed model to reveal a broader picture of the consequences of wellness programs on perceived organizational support and affective commitment. Third, the cross-sectional design constricts the scope of the results, as causal effects are not realistically shown, and therefore a longitudinal study could be developed as a future work. Fourth, the proposed model may be contingent on the national or even local culture, so it would be advisable to replicate it in different countries or regions.

Figures

Figure 1

Research model

Figure 2

Structural (path) model

Table 1

Descriptive statistics (sample)

		N	%
Gender	Female	78	57%
	Male	59	43%
Organizational tenure	1–5 years	22	16%
	6–10 years	37	27%
	11–15 years	35	26%
	16–20 years	26	19%
	More than 20 years	17	12%
Education	Below secondary education	6	4%
	Secondary education	64	47%
	Bachelor's degree	67	49%
Managerial level	Operative worker	34	25%
	Sales representative	27	20%
	Senior or middle manager	34	25%
	Technician	21	15%
	Administrative assistant	21	15%
Industry (respondents)	Manufacturing	24	18%
	Commerce	36	26%
	Services	56	41%
	Construction	21	15%

Table 2

Descriptive statistics (constructs)

	Descriptives		Correlations (Pearson)
	Mean	SD	EWP	COV	POS	AC	IB
EWP	2.54	1.16	1	0.43**	0.45**	0.47**	0.49**
COV	3.75	0.97		1	0.58**	0.48**	0.49**
POS	3.36	0.87			1	0.62**	0.65**
AC	3.27	1.02				1	0.53**
IB	3.73	0.80					1

Note(s): ** Significant with p < 0.01

Table 3

Construct reliability and convergent validity (measurement model)

Construct	Range of loadings	Cronbach's α	rho_A	CR	(AVE)
EWP	0.656–0.870	0.901	0.902	0.921	0.594
COV	0.638–0.883	0.879	0.884	0.910	0.630
POS	0.717–0.849	0.906	0.912	0.925	0.640
AC	0.643–0.852	0.907	0.919	0.927	0.646
IB	0.713–0.774	0.745	0.747	0.839	0.566

Table 4

Discriminant validity and Variance Inflation Factor (VIF)

Construct	Fornell–Larcker criterion					Heterotrait–monotrait ratio (HTMT)				VIF
	EWP	COV	POS	COM	IB	EWP	COV	POS	COM	IB
EWP	0.771
COV	0.434	0.794				0.471
POS	0.460	0.580	0.800			0.500	0.641			1.679
AC	0.472	0.481	0.630	0.804		0.525	0.528	0.683		1.669
IB	0.498	0.487	0.661	0.547	0.752	0.608	0.592	0.795	0.643

Table 5

Structural model analysis: direct effects

	Path coefficient	T-statistics	p-value	Confidence interval
				5%	95%
EWP → POS	0.257	3.524	0.000	0.143	0.384
COV → POS	0.468	6.075	0.000	0.336	0.591
POS → AC	0.630	10.176	0.000	0.525	0.729
POS → IB	0.525	5.575	0.000	0.377	0.695
AC → IB	0.216	2.094	0.018	0.041	0.369

Table 6

Structural model analysis: indirect effects

	Path coefficient	T-statistics	p-value	Confidence interval
				5%	95%
POS → AC → IB	0.136	2.050	0.020	0.027	0.238

Appendix Items of the research

Employee wellness program (EWP)

1. I feel I can keep fit in my current job.

2. The company expects employees to be physically active.

3. The company places great emphasis on employees being physically active.

4. The company provides training and information on healthy lifestyle habits.

5. Ideas about physical exercise are exchanged at work.

6. The company encourages healthy eating habits.

7. Sometimes the company organizes sport events for the employees

8. The company usually provides information and training related to healthy habits

COVID-19 program prevention (COV)

1. The company has placed strong emphasis on workplace safety during the pandemic.

2. The company has given a lot of importance to safety at work in this pandemic situation.

3. The company has kept us informed regarding safety regulations.

4. The company has ensured that employees receive adequate training in relation to the pandemic.

5. The company has encouraged employees to be actively involved and participate in safety issues in relation to COVID-19.

6. The company has facilitated teleworking by providing the necessary means to do so.

Perceived organizational support (POS)

1. The company tries to take care of its employees' well-being.

2. The company pays attention to its employees' interests and objectives.

3. The company values the contribution of its employees

4. The company cares about employee satisfaction.

5. The company tries to make our work as interesting as possible.

6. The company considers the opinions of its employees.

7. The company is always willing to help its employees.

Affective commitment (AC)

1. I feel that I am a part of this company

2. I would like to stay in this company all my life

3. I really feel as if this organization's problems are also mine

4. I feel a strong sense of belonging to my company

5. I feel emotionally attached to my company

6. I feel like part of the family at my company

7. My company has a personal meaning for me

Innovative behavior (IB)

1. I frequently propose innovative solutions to problems that arise in the course of my duties.

2. I use a wide variety of sources/types of information to develop an innovative solution for a problem.

3. I frequently develop new methods to solve unexpected problems.

4. In my department, colleagues rely on me to help them to find solutions to work problems.