Employees Perceptions about Corporate Social Responsibility—Understanding CSR and Job Engagement through Meaningfulness, Bottom-Up Approach and Calling Orientation

Abstract

This article analyses the effect of employees’ perceptions of corporate social responsibility (CSR) on job engagement, and we measure meaningfulness experienced and the role of cross-level sensemaking factors, such as the bottom-up approach and calling orientation. Drawing on qualitative data, collected among workers that had CSR implemented in their companies, our findings suggest that both calling orientation and meaningfulness influence the positive impact of the CSR perceptions on job engagement through sequential mediation. The calling orientation has an important role in this relationship because meaningfulness alone does not influence the relationship between CSR and job engagement. Additionally, employees’ perceptions of CSR positively influence job engagement. Furthermore, our research indicates that the meaningfulness experienced by workers increases in the presence of a bottom-up approach in what concerns the definition and implementation of CSR actions of the company. Overall, this research expands our understanding of how people find meaningfulness through individual experiences of CSR.

1. Introduction

We are currently witnessing an increase in the importance given to corporate social responsibility (CSR), as companies are investing more, not only in terms of money and time [1,2], but also in the diversity of programs they offer to their employees [3].

This trend towards CSR is due to the various possible outcomes for stakeholders, organizations, and institutional levels, provided by the implementation of CSR initiatives in the company [4]. In fact, it has been proven that with corporate social/environmental performance, a firm can improve its corporate financial performance; that is, by applying CSR, an organization improves its reputation and goodwill, leading to the creation of a better image among external stakeholders that ends up generating greater financial performance [5,6].

However, CSR not only creates an environment for positive results from an organizational point of view, but employees also benefit from these activities. CSR influences how employees feel about their workplace and their job. CSR increases employee engagement [7], in-role performance (workers reciprocate the positive attitudes toward socially responsible practice in their jobs) [8], organizational identification [9] and retention [8], among others. Once again, these individual outcomes also constitute a benefit for the company, as they intensify the attractiveness of the firm, providing a competitive advantage in attracting applicants/candidates [10].

Companies are intrinsically linked to work, which is an important part of human life, since it occupies one-third of our entire life [11], representing a source of belongingness and meaningfulness [12]. In this regard, we can speculate that the benefits that an employee acquires by engaging in CSR initiatives through work [13] can be great if activities are discussed in ongoing and participative dialogues which end with action [14,15]; thus, CSR can contribute to the search for meaningfulness. Meaningfulness differs from employee to employee, resulting in the amount of purpose that the worker assigns to the work. In this regard, when we refer to “meaningful work”, we are implying that the worker feels that their job provides significance and embraces positive meaning [13,16]. We call sensemaking the main process through which people attribute meaning and try to find an explanation to ongoing experiences they encounter during their daily lives, such as performing tasks or other events [17,18]. It is through this process that individuals can distinguish between the type of meaning that work makes available to them and the amount experienced. Additionally, meaningfulness at work will provide greater engagement.

The effects of CSR on employee engagement will depend on employee sensemaking and the meaningfulness employees experience due to CSR. When an employee has greater participation in the organizational practices, such as how work is designed, this will increase the person–job fit, and high meaningfulness will be experienced [19]. Consequently, the way CSR is implemented in organizations, considering a bottom-up or top-down approach, will impact employees’ sensemaking. If CSR is experienced as a bottom-up process, the meaningfulness experienced by employees will be stronger [20].

Meaningfulness and, ultimately, job engagement, will also depend on individual characteristics, or more specifically on the relationship that a worker has with his/her job. [21], developed a framework based on [22] identifying three kinds of work orientations: job, career and calling. All types of work orientations generate meaningfulness, even though the amount experienced differs among them [17,21]. Thus, it is expected that the type of work orientation will affect the relationship between meaningfulness and CSR [23], especially when an individual has a calling orientation [21].

With this in mind, this study aims to investigate whether perceived CSR impacts employee meaningfulness and engagement and whether this relationship is affected by the employees’ opinions about their participation in building CSR initiatives and policies (bottom-up approach) and by their work orientation, particularly calling orientation. In this sense, it is expected that the participation of the employees in the planning and implementation of CSR policies and practices will favor the relationship between these policies and the meaning that the employee attributes to his/her work. Additionally, it is also expected that this relationship will be positively impacted the more the employee is prone to calling orientation [21,24,25].

The study will focus on companies that already have CSR programs within their strategies. Subsequently, this study will focus on an individual-level analysis which has been neglected in the CSR field until recent years [4,26]; although studying the processes involved in CSR engagement is crucial to understand why individuals behave differently, these processes have not been widely researched [17]. Additionally, understanding the way that CSR is developed helps companies to discern pros and cons and shape programs according to the results provided; the comprehension of how work orientation can impact the positive effect of bottom-up CSR on meaningfulness can also help companies outline CSR initiatives by understanding the work motivation profile of its employees. At the same time, sensemaking offer a favorable structure for analyzing various CSR phenomena, but research in this area is expected to increase [15].

2. Theoretical Framework

2.1. Corporate Social Responsibility

CSR is a construct that considers micro and macro levels, and academics have started to focus more on the micro level of analysis [23,26], which will be the focus of this study. Corporate social responsibility (CSR) can be defined as “corporate behaviors which aim to affect stakeholders positively and go beyond its economic interest” [27,28].Even though this definition provides a general context about this term, many scholars use the following definition by [26]: “context-specific organizational actions and policies that take into account stakeholders’ expectations and the triple bottom line of economic, social, and environmental performance”. CSR focuses on many types of stakeholders, such as the government, customers, and employees, among others, considering not only financial outcomes but also non-financial ones. In addition, and even though this definition considers that actions and policies are implemented by organizations in their strategies or the daily life of their workers, it is actually the employees that generate and endorse such policies and actions [4].

CSR expands the concept of work, considering that work is not exclusively profit-focused, and produces a pathway for employees to make sense of and find meaning through work [4]. Due to this capacity for greater understanding, it is essential to study employees’ perceptions of CSR [23]. However, even though the literature has put some efforts into studying the micro level, little is known about how CSR directly influences workers [29,30,31], especially due to the way the measures of CSR are constructed, focusing only on how it is perceived by individuals external to the organization, not accounting for internal activities [32]. In this sense, we need to choose constructs that can accurately capture external and internal activities, how they are embedded in the organization, and the perception of how CSR personally affects the employees [32].

When employees perceive that the organization is treating them fairly and care for their well-being, they will respond positively towards the organization [23], increasing employee engagement and overall satisfaction. This occurs because when people feel that the company is ensuring their satisfaction and supporting them, they feel the obligation to reciprocate those behaviors [33]. Thus, CSR perceptions affect employee behavior and even performance and, thus, it is important to study them [32]. Furthermore, when employees realize that their organization also shows a caring behavior towards others through its CSR, this will trigger positive employee attitudes and behaviors, and also may play a part in employees’ sense of meaning and purpose, which will lead to employee engagement, commitment and satisfaction. Moreover, the relationship that an employee has with his work contributes to the understanding of the meaning attributed to the work [23]. This can be studied by resorting to the work orientation framework, which will be explored later.

2.2. Job Engagement

According to a global study on engagement, it has been found that only 16% of employees worldwide are fully engaged, and this percentage has decreased compared to previous years [34]. Disengagement can jeopardize companies and be very costly.

Job engagement can be defined as a “positive, fulfilling, work-related state of mind that is characterized by vigour, dedication, and absorption” [35]. In this definition, vigor means the willingness to invest energy and effort in work, dedication is related to pride and enthusiasm, and absorption refers to being fully concentrated [35,36].

Employees feel more engaged at work when they perceive that the job characteristics are challenging and meaningful to them, and this is enhanced in individuals whose qualities thrive in such environments [37]. Moreover, when employees perceive that they work for socially and ethically responsible firms, they will experience motivation and job satisfaction [38]. Therefore, CSR is one of the main agents in constructing employee engagement [1].

Indeed, in the literature, we can find examples of how employees’ perceptions of CSR initiatives may impact engagement [3]. As an example [39,40] we can say that employee’s perceptions about corporate volunteerism programs predicted that the workers would be more engaged six months after the implementation. A more broad investigation of the relationship between CSR activities and engagement [41,42,43] forecasted that engagement would be driven by CSR policies. Thus, in this sense, we expected that employees’ perceptions about CSR initiatives would generate employee engagement, announcing the following hypothesis:

H1. 

Employees’ perceptions about CSR initiatives have a positive effect on employee’s job engagement.

2.3. Meaningfulness

Meaningfulness is the amount of meaning and purpose that an individual derives from an ongoing experience that is worth something to the person [16]. Employees that experience the same situations may vary in terms of perceived meaningfulness, since something can be very meaningful for one individual, but not necessarily meaningful to another [13]. Meaningfulness is subjective, varying from an individual to another, and is not attached to a given job or organization, so universal meaning is unlikely to be accomplished or verified [16]. In order to explain the fragile nature of meaningfulness, we outline five unexpected features of meaningful work: (1) self-transcendent; (2) poignant; (3) episodic; (4) reflective; and (5) personal. In sum, these attributes denote that meaningful work is not always positive (poignant) and can wane in some instances (episodic); it may be associated with caring for others and the environment (self-transcendent), involve considering moments lived retrospectively (reflective), or be allied to certain life experiences (personal) [44].

Meaningfulness is considered a fundamental human need, and those that experience it through work may present more positive outcomes [14,20,45,46,47]. Indeed, and contrarily to what may be expected given the positive outcomes generated by having meaningful work, little research has answered the questions of “where” and “how” individuals find their work meaningful and the role of leaders in this process [25].

Meaning at work may have various sources derived from the self, other people, the work context, and/or spiritual life. It is the interconnection between these main foundations that allows an individual to fully evaluate the meaning or meaningfulness of their work [13], and when a high degree of meaningfulness is experienced, it is associated with more positive outcomes, such as job satisfaction, organizational commitment, organization identification, psychological well-being, engagement, performance, and lower levels of stress. Indeed, employees’ perceived meaningfulness is fundamental in organizations, being one of the most influential means affecting employees’ positive organizational behaviors [48].

Considering “the self”, psychologists [49,50,51] ponder that various types of attitudes, behaviors and beliefs are determined by this source. We can divide this field into three domains of research: (1) values, (2) motivations, and (3) beliefs about work, which are key valuable sources of meaning. Undeniably, the way individuals see themselves, that is, how they evaluate and orient themselves towards their own work, is vital in assigning meaning to work [13]. Additionally, the work orientation of an individual (which will be developed further below) is a sensemaking factor that affects and precedes the meaningfulness experienced. It is speculated that those with calling orientation, which is related to “the self”, experience more meaningfulness [1].

Pondering about “other persons”, as well as the interactions and relationships between other persons or groups in the workplace or outside of it, can impact the meaning attributed to their work. These other persons can be (1) co-workers, (2) leaders, (3) groups and communities, or (4) family [13]. The degree of influence a person is subject to and his/her sensitivity to that influence can impact meaningfulness [52].

The “work context” is also another player in providing a source of work meaning. The way the work is conducted, and the environment surrounding it impact individuals in terms of both perceptions of meaning and meaningfulness. In this sense, the (1) design of job tasks, (2) organizational missions, (3) financial circumstances, (4) non-work domains, and (5) the national culture in which the work is conducted are factors that help define terms in the investigation. Thus, the CSR actions developed by the organization and the processes used by the company to define CSR activities may affect meaningfulness at work. Thus far, it is difficult to fully cover these concepts, since in the modern world the context in which work is inserted is constantly changing [13]. We are witnessing at the moment with the COVID-19 pandemic that the performance of jobs needs to be updated to this new reality. Thus, the linkage between the context of work and meaningfulness needs to be investigated and keep up with changes.

Lastly, “spiritual life” is an underlying source of meaning, since people turn to their spirituality in order to try to find their plausible purpose in life [53,54]. Nevertheless, religion is still a taboo topic to be discussed in the workplace, even if some people are gaining confidence in speaking out about such a theme with other workers [55]. Thus, religion contributes significantly to how individuals manage their work lives [56].

Based on the presented perspectives, we can present our second hypothesis that considers that CSR actions, either within or outside of organizations, and the way they are perceived, have a vital impact on employees, which may be translated to a general benefit for society, leading to employees’ experiencing meaningfulness at work.

H2. 

Employees’ perceptions of CSR initiatives positively affect employees’ meaningfulness at work.

As highly referenced in the literature, employee engagement can be driven by meaningfulness [57], even though the relationship between the two concepts is still very restricted [58]. People crave to make a difference and be unique, which will, of course, have major consequences for work. The world has changed, and the aim of work is no longer to uniquely perform a certain task, but it is an opportunity to display our personality, identity, and individuality [1]. It is in this climate that workers demand meaningful work because it will make work and life engaging [58]. Additionally, the perceptions of the employees about CSR impacting meaningfulness can offer an additional source of engagement for employees at work that goes beyond the simple characteristics of the job [3], so we can consider the following hypotheses:

H3. 

The meaningfulness experienced will have a positive impact on employees’ job engagement.

H4. 

The meaningfulness experienced by employees will have a mediation effect on the relationship between CSR and job engagement.

2.4. Sensemaking

Sensemaking is “the process through which people work to understand issues or events that are novel, ambiguous, confusing, or in some other way violate expectations.” [59]. Some authors [60] have already emphasized that “Sensemaking has many distinct aspects—comprehending, understanding, explaining, attributing, extrapolating, and predicting, at least. […] What is common to these processes is that they involve placing stimuli into frameworks (or schemata) that make sense of the stimuli”. Another definition of sensemaking is provided by [18] simply, as the “making of sense”. The author emphasizes that with this process, the agents try to answer questions of how it is constructed, what is constructed, why, and with what effects. Thus, in other words, the researcher stresses that sensemaking is the procedure in which individuals attribute plausible and possible meaning to ongoing experiences.

According to this definition, it is possible to deduct that sensemaking is more likely to be found when we face complex and very challenging situations, or when employees need to make decisions about tasks that were assigned to them and have to perform them, which can be impactful to stakeholders [1]. Hence, these situations are associated with CSR initiatives, where it is necessary that workers’ actions and reactions regarding activities are not completely related to the company’s core objectives and values, but to departmental projects more complex in nature that, ultimately, will affect their surrounding environments, such as the community and the planet.

Based on this, we need to study how certain events are a trigger for sensemaking. In this regard, sensemaking occurs when cues take place, such as issues, events, or situations where we are not certain about the meaning of those cues. They are ambiguous and uncertain, entailing the misperception of the world and disturbing normal people’s ongoing flow and how they act. These will occur when we denote differences between expectations and reality, even though unexpected events will not necessarily induct sensemaking. Thus, it will only be possible if it complies with these conditions: the discrepancy between what is expected and what is experienced is sufficiently large and important, leading individuals or groups wanting to know more about what is going on and what should they do [59].

Due to this, sensemaking can shape the perception of our place in the world, changing the mode of how we face the various experiences that we partake at work [4]. According to this, and as suggested by [18], identity constitutes an important component of the sensemaking process. This incorporates a more vital role when we undertake a micro-level analysis, because “when individuals collectively select a certain interpretation of some experience, they are at the same time selecting a particular identity for themselves” [61].

Some authors [4] denominate “sensemaking factors” as the underlying mechanisms explaining meaningfulness, dividing them into three categories representing three levels of analysis: (1) intraindividual, (2) organizational, and (3) extra-organizational. Through the interaction between these mechanisms, we can examine why and how the individuals that face the same experiences can deduct and interpret different CSR understandings and meanings.

Top-Down and Bottom-Up Processes

Even if the search for meaningfulness is an individual process, since it is the worker who can evaluate the meaning that certain situations provide him/her, sensemaking, on the opposite side, is not exclusively an individual process. It can be a result of the social process, influenced by other persons and organizations’ contexts [4]. In this case, we will focus on organizational-level sensemaking factors in more detail regarding how CSR is planned and implemented in a company, considering a top-down or bottom-up approach.

It is believed that for CSR to have a meaningful impact, it must be incorporated at all levels of the organization and must be seen as something important that belongs to the strategic management process [62]. We can have a simultaneous top-down and bottom-up approach; the top-down approach is related to the identification of the needs of stakeholders and the further integration of CSR with internal management systems, and it constitutes a more systematic process to communicate and translate CSR from top managers or CSR teams to all levels of the company; the bottom-up approach consists of consulting the needs of the stakeholders through the creation of indicators that link to community needs. In this sense, it is a systematic way to engage stakeholders in the decisions that will be taken regarding CSR and later incorporated into the company [62]. These authors advise the use of both processes to plan the best practices to be implemented, aiming to apply the stakeholders’ communication and feedback and act accordingly.

In the current world, the approaches to CSR are mostly top-down [63], due to the fact that normally the senior management team, CEO and president are the ones responsible for defining and creating the organization’s CSR strategy [64]. When the process of sensemaking is influenced by the management team, we are witnessing the “sensegiving” process [4], (i.e., after an interpretive process of making sense of an organization’s internal or external environment and changing the strategy’s vision, it is disseminated to all stakeholders. By doing this, we are sensegiving, but employees can have different reactions to this process: accepting, questioning or rejection [65]. Thus, naturally, most processes of sensegiving were found to be top-down, where the role of the employees is minimal or even null, which will affect their sensemaking process [59]. Thus, when CSR is implemented by only focusing on a top-down process, the initiatives diffused through the company can be perceived as an extra-role task, which could lead to stress and lower well-being [4] due to the high pressure that can be created from accomplishing not only the work’s objective but also fulfilling the CSR activities’ objectives, with high job demands [66]. In fact, and based on the reactance theory when CSR is forced upon workers, they will act on the contrary in order to prevent being influenced and avoid feeling that their freedom could be endangered [67].

Trying to offset this matter, a bottom-up approach can serve as a guideline to put into practice CSR initiatives to obtain the most favorable results for all players. This approach is often referred to as social intrapreneurship, corporate social entrepreneurship, issue selling or job crafting [4], where the stakeholders are called to expose their opinions and thoughts about those activities to further be implemented in the company and to be linked with their necessities [62].

Issue selling and job crafting are related to the sensemaking process [4]. Issue selling is when employees try to influence the organization by influencing others to change the procedures and initiatives to balance economic and social components. In this sense, workers try to “sell” what can be undertaken differently to induce a change in the issue [68]. Job crafting is the action that employees partake in in order to shape and redefine their jobs [69]. Workers resort to this process to find meaningfulness at work because, if work is designed via a top-down approach, the need for meaningfulness is often not fulfilled [70]. With this, CSR is included in the job, constituting an opportunity for employees to express themselves and make a difference to the society and environment. Additionally, if those programs are important to them, the employee–company relationship will be strengthened, redirecting those positive outcomes towards their job performance [71]. Thus, we can consider the following hypothesis:

H5. 

If a bottom-up approach of CSR is implemented in a company, the relationship between CSR and meaningfulness experienced by employees will be stronger.

2.5. Work Orientation

The relationship a worker has with work and the organization diverge according to individuals’ intentions or attitudes towards work, constituting what we call “work orientation” [21]. Various work orientations allow us to interpret thoughts, feelings and behaviors towards work, helping to provide an explanation for how people see their work, but also, for reasoning about how they shape their jobs [69].

In view of trying to summarize these intentions, some authors [22] describe three types of orientation towards work, reviewed by [72,73], based on their observation of the work in the United States. Further developed by [21], the framework encompasses three work orientations: (1) job, (2) career, and (3) calling.

Firstly, job orientation is when people perceive their work as a mere job, in which they can obtain material benefits such as money, excluding other kinds of fulfilment or desires for other types of rewards. In this sense, we can argue that the work is a means to an end, such an end being focused on financial terms [21]. Some authors [11] refers that when this occurs, normally people’s ambitions and interests are not allied with the work domain, but outside of it, with the job only a tactic to acquire resources to enjoy other things.

In contrast, those with career orientation have a strong connection with the company, since the fulfilment of their ambitions depends on it [11]. Employees are deeply invested in performing the best they can, so they can climb the various stairs to move from one position to a better one. In this regard, advancement in the positional ladder of the organizational structure, which comes with more prestige, status, and increased pay, is the main goal of these workers [21]. Some authors [22]. highlight that promotions bring more self-esteem, increased power, and higher social standing, which is enough to satisfy their desires.

Lastly, when individuals have a calling orientation, we are implying that they perform a certain job not to obtain financial means or for advancements, but for fulfilment. Thus, work is no longer a means to an end, but an end in itself, guiding the worker in believing that he is contributing to a greater good and making the world a better place [21]. In this sense, they cannot separate the work from their life, because the search for meaning depends on it [11]. The origin of the term calling is related to religious purposes, because of the belief that people have been “called” and chosen by God to perform ethically and socially significant work [74,75]. In modern times, the linkage between calling and religion has fallen and only considers the contribution to a better world [76].

Work orientation is a sensemaking factor that divides meaningfulness into three types that can be experienced, as mentioned above (job, career and calling). In all situations, the worker experiences meaningfulness, although the meaning changes ([1]. Thus, in this sense, in all kinds of jobs, it is possible to view the work as a job, career or calling [21], finding within the same occupation individuals with all the three kinds of work orientation [11]. Theoretically, calling orientations should be considered as antecedents of meaningfulness, since those with callings are provided with a sense of meaning and purpose when at work [13,21], therefore, it is expected that CSR will have a stronger echo if individuals are concerned with having ethically and socially significant work. Thus, we can consider the following hypothesis:

H6. 

The calling orientation will positively impact the relationship between CSR and meaningfulness.

Keeping in mind what was mentioned previously with the direct and indirect relationships between the concepts, as well as the formulated hypotheses, we put together two conceptual frameworks, as shown in Figure 1 and Figure 2. The main difference between them is that, in Model 1, the calling orientation is moderating the relationship between CSR perceptions and meaningfulness, whereas in Model 2, the calling orientation is not moderating the relationship, but mediating it, which in consequence influences job engagement by having sequential mediation.

With these conceptual schemes, we can observe not only the hypotheses individually, but most importantly how variables interact with each other. We want to analyze if there is mediation through meaningfulness (between CSR perceptions and job engagement), and if two moderators—the bottom-up approach and calling orientation—intensify the relationship between CSR perceptions and meaningfulness in Model 1 or, on the other hand, if this relationship is mediated or moderated by calling orientation. Additionally, we will verify which models have a higher impact on job engagement.

3. Methodology

A quantitative study was carried out to test our models. To accomplish this, an online questionnaire was developed to collect data, voluntarily and confidentially. The questionnaire was aimed at employees that work on companies that have CSR initiatives implemented in their strategy, which restricted the number of people that could participate in the survey. The surveys were distributed in English and Portuguese through online Google Forms links. The Portuguese version (PTV) was shared via social networks such as Facebook and LinkedIn, whereas the English version (ENGV) was on the Amazon Mechanical Turk (Mturk), which allowed us to have no geographical restrictions by distributing it globally.

In order to control for common method bias which can prejudice the relationship between two variables, leading to incorrect perceptions and affecting hypothesis tests, we decided to apply temporal separation between the measurement of the main (CSR) and dependent (job engagement) variables. This also helps to reduce the respondents’ fatigue due to a very long survey [77]. In this sense, both versions of the questionnaire were divided into two moments (Moment 1 and Moment 2). However, even with the application of the technique mentioned, it is important to verify the existence or not of bias in our data collection and to what extent it may be a problem. To do that, we resort to Harman’s one-factor (or single-factor) test, which involves analyzing to what extent one single factor explains the majority of the variance of the results. Using the SPSS Software and applying this test to both databases together (ALL’s version), we conclude that a single factor only accounts for 32.21% of the total variance, justifying that the possibility of verifying this bias is reduced.

For Moment 1, we received a total of 563 responses. Among them, only 549 (97.51% of the total submissions) could be considered as we had 14 responses that were duplicated (using the same code), which were considered invalid. In this way, the second part of the questionnaire was only sent to those whose responses were considered valid, obtaining a response rate of 60%. We had a total of 326 responses for Moment 2, 155 of them were from the Portuguese version (PTV) and the remaining 171 were from the English version (ENGV). These responses underwent an analysis and refinement and, in the end, we had 322 valid answers.

To analyze and validate the theoretical models proposed, we opted to use structural equation modelling based on partial least squares structural equation modelling (PLS-SEM) for estimating the parameters [78,79,80].

3.1. Measures

Corporate Social Responsibility Perceptions will be evaluated using the 17-item scale developed by [28]. This scale evaluates the perception that an employee has about the CSR initiatives and the organization responsibilities toward the various stakeholders by establishing four dimensions: (1) society, environment, future generations, and NGOs; (2) employees; (3) customers; and (4) government. It has obtained a Cronbach’s Alpha of 0.9279. A Portuguese version of the Corporate Social Responsibility Scale (CSRS) was used by [81] and obtained a Cronbach’s Alpha of 0.93, which is similar to the one observed in the original investigation. Responses will be assessed on a five-point Likert scale where 1 indicates strongly disagree and 5 indicates strongly agree, which means that the higher the score, the more the employees perceived CSR actions taken by their employers.

The bottom-up approach will be based on the work of some authors [82,83]. These scales were adapted to fulfil the objectives proposed in this investigation. The original Cronbach’s alphas [82] were 0.93 and 0.87, according to the subscales of strategic renewal behavior and venture behavior, respectively. Additionally, in the research developed by [83] the scale obtained a Cronbach’s alpha of 0.93 and 0.81, according to the subscales for building collaborative structures and participative management, which are more consistent with the objectives of our investigation. For obtaining the Portuguese version of this scale, we proceeded to a translation of the items and back-translation to guarantee that the measures maintained the same meaning. A five-point Likert scale similar to the previous one was used.

For calling orientation, we will use the model developed by [11] resorting to the 18 true–false propositions. The scale was translated into Portuguese by [84], obtaining a Cronbach’s alpha ranging between 0.701 and 0.754. However, in this study, it will be assessed by resorting to a five-point Likert scale.

Meaningfulness will be measured using the work and meaning inventory scale (WAMI) [85].The scale was used before [85] and the Portuguese version of this scale was adapted by [86], and obtained an internal consistency of 0.94. Once again, we will use a five-point Likert scale.

Job engagement will be measured based on the 9-items Utrecht work engagement scale (UWES-9) [87], which is divided into three dimensions: vigor, dedication and absorption. The scale already has a Portuguese version, which obtained a Cronbach’s alpha of 0.95 [88]. The responses to the items will be assessed on a five-point Likert scale.

3.2. Sample

We had a total of 322 valid answers.

Table A1. Sociodemographic and professional characterization of the sample.

Characteristics	Responses	Frequency	Percentage
Gender	Female	146	45.3%
	Male	176	54.7%
Age	≤20 years	1	0.3%
	21–40 years	236	73.3%
	41–60 years	77	23.9%
	>60 years	8	2.5%
Country of Origin	Portugal	141	43.8%
	USA	42	13.0%
	Brazil	26	8.1%
	India	69	21.4%
	Others	44	13.7%
Level of Education	Primary school	11	3.4%
	Secondary school	38	11.8%
	Bachelor’s degree	170	52.8%
	Master’s degree	100	31.1%
	Doctoral degree	3	0.9%
Employment Status	Part-time	20	2.5%
	Trainee	15	6.2%
	Employed with a fixed-term contract	69	21.4%
	Employed with a non-fixed contract	158	49.1%
	Self-employed	52	16.1%
	Other	8	2.5%
Activity Sector	Agriculture, farming of animals, hunting, forestry, and fishing	14	4.3%
	Mining and quarrying	3	4.3%
	Manufacturing	38	11.8%
	Electricity, gas, and water	10	3.1%
	Construction	11	3.4%
	Wholesale and retail trade	28	8.7%
	Transportation and storage	11	3.4%
	Accommodation and food service activities	5	1.6%
	Financial and insurance activities	57	17.7%
	Real estate activities	5	1.6%
	Education	32	9.9%
	Human Health and social work activities	32	9.9%
	Other sectors	76	23.6%
Organization Dimension	<10 workers	51	15.8%
	10–50 workers	92	28.6%
	51–250 workers	64	19.9%
	>251 workers	115	35.7%
Seniority in the organization	<1 year	36	11.2%
	1–2 years	64	19.9%
	3–5 years	81	25.2%
	6–10 years	47	14.6%
	>10 years	94	29.2%

(Appendix A) details the sociodemographic and professional characteristics of the sample, and we can observe that it is more or less balanced with regard to the gender of the participants. The vast majority of our respondents are from Portugal (43.8%), followed by India (21.4%), the USA (United States of America) (13%) and Brazil (8.7%), among others (representing 13.7%). Most respondents were of adult age, with 73.3% of participants aged between 21 and 40 years old, followed by 23% between 41 and 60 years old. These age results explain the fact that most of the respondents had a higher education, with 52.8% having a bachelor’s degree and 31% having a master’s degree.

Most of the contributors reported a stable employment situation, working mainly in the areas of financial and insurance activities (17%) and manufacturing (11%), as represented in our sample. Most of the participants work in companies with more than 251 workers (35%), followed by companies with 10 to 50 workers (28%), organizations with 51 to 250 workers (19%) and with less than 10 workers (15%). In addition, we verify no tendency for seniority in the organization; 29% of the respondents had been working for more than 10 years in their organization and a lower percentage of representation had been working in the organization for a period time of less than 1 year (11%).

4. Results

4.1. Exploratory Factorial Analysis (EFA)

The bottom-up approach constitutes a measure that results from the junction of two scales previously elaborated, resulting in a new scale. Due to this adaptation, an EFA for this scale was conducted so that the items proposed could be validated. The principal component analysis (PCA) method was chosen to perform the EFA and estimate the communalities. We performed these two tests for the English and Portuguese versions, where we obtained the values for the KMO criterion of 0.936 and 0.953, respectively, which according to the literature (Marôco, 2014) are very good values. Bartlett’s sphericity test gave us a value of 867.771 (ρ < 0.001) for ENGV and 1338.352 (ρ < 0.001) for PTV, which are both considered as good values, meaning that there is a significant correlation between the items. Next, we focused our attention on the analysis of the individual communalities of the items and, in our two versions, all the items had a value for the individual communalities greater than 0.5, so no item was eliminated. Next, we evaluated the number of factors to retain; we were able to extract only one factor that explained 61%, for ENGV, and 76%, for PTV, of the total variance. This extraction followed the assumptions made when we joined and adapted the two scales, together with evaluating the bottom-up approach when choosing, deciding and applying the CSR initiatives. Finally, to analyze the internal consistency of this factor, we resorted to Cronbach’s alpha; with the minimum acceptable value of 0.7 [89], we obtained for the ENGV and PTV the values of 0.92 and 0.96, which are considered as very good [89].

4.2. Measurement Invariance

To evaluate if we could join our two databases (ENGV and PTV), we tested if there was measurement invariance for the two models proposed in this study. To accomplish this, we used the procedure to access the measurement invariance of composite models (also known as MICOM) when using PLS-SEM [78,90]. The procedure created by [90]. involves three steps, which each must be verified to conclude that there is measurement invariance between the two databases: (1) configural invariance; (2) compositional invariance; and (3) the equality of composite means and variances.

The first step, configural invariance, is of qualitative nature, since with it we entail that an item has been equally specified for both groups [90], and since we apply indistinguishable indicators in each measurement model with equal data treatment (coding in the same way, for example) and identical algorithm settings, we do not have configural invariance. We created Model 1 and Model 2 in order to perform a permutation test with an acceptable minimum of 500 subsamples based on the two groups defined [90]. If the value of c (correlation between the items’ scores using the weights obtained from each group) exceeded the five percent quantile, we concluded that there was compositional invariance, but the results were very similar for both models, so we concluded that we had compositional invariance. Finally, the third stage is to confirm that measurement invariance consists by proving the equality of the composite mean values and variances. Relying once again on permutation tests, we investigated if there were differences between the means and the variances between the variables’ observations values and we concluded that we had measurement invariance, so we could join both models with the two databases (ALL’s version).

4.3. Model Evaluation

4.3.1. Model 1

Concerning the reliability of Model 1, all the observed variables have an outer loading above 0.5, which is considered significant, and many present a loading higher than 0.7, which indicates that the variables are well defined and have indicated reliability.

Concerning internal consistency and composite reliability, Cronbach’s alpha and composite reliability present values above 0.8 for all latent variables. Based on the results for the convergent validity, the variable calling is the only one whose value is not above the reference value of 0.5, even though it is considered reasonably close (0.409). Convergent validity for calling may be adequate since AVE is considered as a very conservative measure and CR alone is sufficient to prove convergent validity [91].

We observe that, for all pairs of variables, except for calling–meaningfulness, the values for HTMT are below 0.9, proving discriminant validity. One of the reasons for the HTMT between calling and meaningfulness of 0.934 (close to the minimum cut-off of 0.9) is due to the conceptual similarity between these two constructs. Due to this reason, we will check for HTMT inference criteria by running the bootstrap routine. The HTMT inference states that there is discriminant validity for two variables if, using the complete bootstrapping test, we obtain HTMT significantly different from 1, which means that at a 5% significance level, the confidence interval must not contain the value one ([92]. By performing the test and observing the results, we can conclude that the two constructs are empirically distinct for HTMT inference since the upper limit of the confidence interval is less than one (0.975).

Table 1. Indicators for reliability, convergence and discriminant validity—Model 1 and 2.

Latent Factors	Alpha	CR	AVE	1	2	3	4	5	6
PES	0.934	0.934	0.545
PCG	0.885	0.884	0.607	0.818
Bottom-Up	0.946	0.946	0.662	0.736	0.450
Calling	0.805	0.805	0.409	0.462	0.320	0.556
Meaningfulness	0.913	0.912	0.537	0.359	0.274	0.412	0.934
Job Engagement	0.913	0.912	0.539	0.352	0.286	0.404	0.883	0.811

presents the values obtained for accessing discriminant validity through the HTMT criterion, but also the values for Cronbach’s alpha, CR and AVE.

4.3.2. Model 2

We start by assessing that the reliability of the results obtained for the outer loadings of the manifest variables are the same as the ones verified in Model 1, and once again we conclude that the reliability of the indicators is verified since all the values are greater than 0.5, and many of them have a value higher than 0.7 (significant and well defined).

The same is observed for Cronbach’s alpha, composite reliability, convergent validity and divergent validity (Table 1), as we obtain the same scores for all the tests. In this way, the reasoning applied above for Model 1 is applied to Model 2, proving that the constructs are reliable and have convergent and divergent validity.

4.4. Descriptive Statistics and Correlational Analysis

After the factorial analyses, an assessment was made of the descriptive and of the bivariate correlations between the model variables with the ordinal sociodemographics through the calculation of correlation coefficients, in order to understand the relationships between them and their importance for the participants.

Table 2. Descriptive statistics and correlational analysis.

Variables	Descriptive Statistics and Correlation Coefficients
	Mean	SD	1	2	3	4	5	6	7	8	9	10	11	12
1. Gender	0.547	0.499
2. Age	35.106	10.447	−0.159 **
3. Education	3.143	0.764	0.022	−0.143 *
4. Dimension	2.755	1.105	0.102	−0.052	0.153 **
5. Seniority	3.307	1.368	−0.100	0.544 **	−0.193 **	0.096
6. CSR	3.752	0.801	0.025	−0.042	0.032	0.038	0.133 *
7. PSE	3.632	0.853	0.052	−0.096	0.034	0.033	0.071	0.979 **
8. PCG	4.040	0.833	−0.041	0.099	0.018	0.042	0.203 **	0.865 **	0.744 **
9. B	3.430	0.906	0.126 *	−0.087	0.104	−0.073	0.007	0.648 **	0.694 **	0.411 **
10. WOCA	3.650	0.732	0.057	−0.023	0.143 **	0.051	0.019	0.383 **	0.401 **	0.268 **	0.485 **
11. M	3.820	0.698	−0.021	−0.012	0.156 **	0.027	0.056	0.320 **	0.328 **	0.241 **	0.383 **	0.801 **
12. JE	3.746	0.718	0.060	−0.009	0.193 **	0.054	0.023	0.321 **	0.325 **	0.251 **	0.376 **	0.755 **	0.740 **

N= 322 *. Correlation is significant at the 0.05 level (2-tailed). **. Correlation is significant at the 0.01 level (2-tailed). Notes: Gender: 1 = Female; 2 = Male. Education: 1 = primary school; 2 = secondary school; 3 = bachelor’s degree; 4 = master’s degree; 5 = doctoral degree. Dimension: 1 = less than 10 workers; 2 = 10 to 50 workers; 3 = 51 to 250 workers; 4 = more than 251 workers. Seniority: 1 = less than 1 year; 2 = 1 to 2 years; 3 = 3 to 5 years; 4 = 6 to 10 years; 5 = more than 10 years. CSR = perceptions of corporate social responsibility; PSE = CSR perceptions related to aspects of society and employees; PCG = CSR perceptions related to customers and government; B = bottom-up; WOCA = calling orientation; M = meaningfulness; JE = job engagement.

contains the results collected on the statistics software, with the indication of mean, standard deviation (SD) and the correlation coefficients between the variables. It is interesting to highlight that, for a 99% confidence variable, all the model variables are statistically significant and positively associated with each other, meaning that the variables vary in the same direction (if one increases/decreases, the other will also increase/decrease). In a general way, in social and human sciences, we consider that the correlation between two variables is very strong when the absolute value of r (|r|) is equal to or higher than 0.75 (|r| ≥ 0.75) [89]. Thus, we can affirm that the following pairs of variables are very strong positively significantly correlated: CSR-PSE (r = 0.979), CSR-PCG (r = 0.865), WOCA-M (r = 0.801) and WOCA-JE (r = 0.755). In addition, we can also identify a strong positively significantly correlation (0.5 ≤ |r| < 0.75) between the variables: CSR-B (r = 0.648), PSE-PCG (r = 0.744), PSE-B (r = 0.694) and M-JE (r = 0.740). The other pairs of model variables not mentioned present weak (|r| < 0.25) and moderate (0.25 ≤ |r| < 0.5) correlations.

It is also important to emphasize that, at a 99% confidence level, age is significantly and negatively associated with education (rs = −0.143) and significantly and positively associated with seniority (rs = 0.544). At the same significance level, education is positively and significantly associated with organization dimension (rs = 0.153), WOCA (rs = 0.143), M (rs = 0.156) and JE (rs = 0.193), except for seniority (rs = −0.193), which is negatively and significantly correlated. At this confidence level, seniority is positively and significantly associated with PCG (rs = 0.203). Finally, at a 95% confidence level, gender is significantly and positively correlated with B (rs = 0.126) and seniority with CSR (rs = 0.133).

4.5. Hypotheses Testing

Model 1

The structural model was estimated through the bootstrapping test. The assessment of the first three hypotheses (H1, H2 and H3) obtained the results provided in

Table 3. Hypotheses validation—path coefficients—Model 1.

Hypotheses	Path	β	t-Value	p-Value	Result
H1	CSR–Job Engagement	0.093	1.951	0.052	Rejected
H2	CSR–Meaningfulness	0.016	0.401	0.689	Rejected
H3	Meaningfulness–Job Engagement	0.720	18.503	0.000	Confirmed

. All three hypotheses investigate the direct effect of one variable on another, and if the effect is positive or negative and statistically or statistically non-significant. All the relationships were positive, meaning that an increase in one variable implies an increase in the other variable (0 < β < 1). However, at a 95% confidence level, only the relationship between meaningfulness and job engagement is statistically significant with a t-value of 18.503 (p-value = 0.000), high above the reference value of 1.96, confirming the hypothesis H3. The relationship between CSR and job engagement and between CSR and meaningfulness is statistically non-significant, as the p-values are higher than the 5% significance level (0.052 and 0.689, respectively), leading to the rejection of the hypotheses H1 and H2, respectively.

4.6. Mediation Analysis—Model 1

To study the mediation role, it is important to consider the indirect effects that may occur between the variables of CSR and job engagement with meaningfulness as the mediator. The results obtained are summarized in

Table 4. Hypotheses validation—specific indirect effects/mediation—Model 1.

Hypotheses	Path	β	t-Value	p-Value	Result
H4	CSR–Meaningfulness–Job Engagement	0.012	0.389	0.697	Rejected

. The indirect effect of CSR on job engagement mediated by meaningfulness does not exist, since the relationship between CSR and job engagement, although positive (β = 0.012), is statistically non-significant at a 95% confidence level, with a p-value of 0.697 (t-statistics of 0.389). Thus, the meaningfulness experienced by employees does not mediate the relationship between CSR and job engagement, rejecting hypothesis H4.

4.7. Moderation Analysis

Lastly, we focus on the moderating effects to the relationship between CSR and meaningfulness to verify if bottom-up (H5) and/or calling (H6) are moderators of this relationship. To prove the existence of moderator effects, the interaction between the moderator and the independent variable (CSR) must be statistically significant, which means that the direct effect of the interaction on meaningfulness must have a t-value equal to or greater than 1.96 (or p-value < 0.05). To add the moderating effects of bottom-up and calling, we define CSR as the independent variable and meaningfulness as the dependent variable in both cases, originating new variables bottom-upMO and callingMO.

Table 5. Hypotheses validation—path coefficients—moderation—Model 1.

Hypotheses	Path	β	t-Value	p-Value	Result
H5	Bottom-UpMO–Meaningfulness	0.062	1.750	0.081	Rejected
H6	CallingMO–Meaningfulness	0.025	0.665	0.506	Rejected

aggregates the results obtained of the direct effects, where we can affirm that the relationships between bottom-up and CSR and calling and CSR are positive since both beta coefficients are positive (β > 0). However, at a 95% confidence level, both interactions are statistically non-significant, leading to the rejection of H5 and H6.

Model 2

Taking into consideration the above mentioned, we estimate the structural model by performing the bootstrapping test to determine the Beta coefficients. We will start by evaluating the confirmation or rejection of the hypotheses H1, H2 and H3 which investigate the direct effect of one variable on another. The results obtained (

Table 6. Hypotheses validation—path coefficients—Model 2.

Hypotheses	Path	β	t-Value	p-Value	Result
H1	CSR–Job Engagement	0.094	2.082	0.038	Confirmed
H2	CSR–Meaningfulness	0.039	0.984	0.326	Rejected
H3	Meaningfulness–Job Engagement	0.720	19.268	0.000	Confirmed

) offer support to confirm the hypothesis H1, with a positive effect of CSR on job engagement (β = 0.094), and H3, representing a positive effect of meaningfulness on job engagement (β = 0.720). However, they do not support the H2 hypothesis (p-value ≥ 0.05), with a non-significant positive effect of CSR on meaningfulness (β = 0.039).

4.8. Mediation Analysis—Model 2

Now, we will examine the indirect effects that may occur between the variables CSR and job engagement, mediated by meaningfulness (H4) and/or the sequential mediation of calling and meaningfulness (H6). We can see in

Table 7. Hypotheses validation—specific indirect effects/mediation—Model 2 (H4).

Hypotheses	Path	β	t-Value	p-Value	Result
H4	CSR–Meaningfulness–Job Engagement	0.028	0.935	0.350	Rejected

the indirect effect of CSR on job engagement mediated by meaningfulness does not exist, as even though there is a positive relationship between CSR and job engagement (β = 0.028) it is statistically non-significant, with a p-value of 0.350. Due to this, the hypothesis of a mediation effect of the meaningfulness experienced by employees on the relationship between CSR and job engagement is rejected.

In the same way, we try to understand if there is an indirect effect between CSR and job engagement, mediated by both calling and meaningfulness (sequential mediation). Based on the results (

Table 8. Hypotheses validation—specific indirect effects/mediation—Model 2 (H6).

Hypotheses	Path	β	t-Value	p-Value	Result
H6	CSR–Calling–Meaningfulness–Job Engagement	0.230	6.048	0.000	Confirmed

), we state that there is an indirect effect between CSR and job engagement, mediated by calling and meaningfulness, because there is a positive relationship (β = 0.230) between CSR and job engagement, which is statistically significant, presenting a p-value = 0.000 (t-value = 6.048). We can conclude that the hypothesis of a possible sequential mediation of calling and meaningfulness on the relationship between CSR and job engagement (H6) is confirmed.

4.9. Moderation Analysis

Finally, we studied the moderator role of bottom-up on the relationship between CSR and meaningfulness (H5). As stated before, to prove the existence of moderator effects, the interaction between the moderator bottom-up and the independent variable (CSR) must be statistically significant on meaningfulness.

Firstly, we add to the latent variable meaningfulness the moderating effect of bottom-up, by selecting CSR as the independent variable and meaningfulness as the dependent variable. The bottom-upMO represents the interaction and is calculated by a two-stage approach (product of CSR and bottom-up). The results in

Table 9. Hypotheses validation—path coefficients—moderation—Model 2.

Hypotheses	Path	β	t-Value	p-Value	Result
H5	Bottom-UpMO–Meaningfulness	0.094	2.082	0.038	Confirmed

reveal that there is a direct effect of bottom-upMO on meaningfulness, indicating a statistically significant interaction between CSR and bottom-up at a 95% confidence level (p-value = 0.038). The interaction is positive, having a beta coefficient of 0.094.

For a deeper analysis of the interaction effect, the interactions were plotted at three conditional values: low (bottom-up mean-1SD), mean (bottom-up mean) and high (bottom-up mean + 1SD). From the interpretation of the plot slope graph (Figure 3), we can state that the relationship between CSR and meaningfulness changes direction based on the bottom-up approach. For low values in bottom-up (considering low participation of the employees), CSR is negatively associated with meaningfulness, and for high values in bottom-up (considering high participation of the workers), CSR is positively associated with meaningfulness. In addition, the direct effect of bottom-up and CSR on meaningfulness (without the interaction effect) are not statistically significant, presenting t-statistics of 0.047 and 0.927 (

Table 10. Path coefficients—moderation (main effects)—Model 2.

Path	β	t-Value	p-Value	Result
CSR–Meaningfulness	0.039	0.927	0.354	Not significant
Bottom-Up–Meaningfulness	0.002	0.047	0.963	Not significant

), respectively, since they average out resulting in a total effect close to 0.

5. Discussion

The results obtained by testing the structural models 1 and 2 through the SmartPLS, disclose different conclusions for the formulated hypotheses when considering calling as a moderator (Model 1) or as a mediator (Model 2). Thus, we can yield the conclusion that, for achieving the aim of this investigation, Model 2 offers a more solid and structural explanation of the effects of CSR perceptions on job engagement considering meaningfulness, bottom-up approach and calling orientation. For Model 1 we could only confirm that the meaningfulness experienced will have a positive impact on employees’ job engagement (H3). Due to this, and based on the results obtained, we will discuss Model 2.

The study of Model 2 shows that CSR perceptions have a positive impact on job engagement (H1) and, although weak (β = 0.094), this direct relationship is significant. This result supports the existing literature that suggests that when employees perceive that the company is caring for their general well-being and the world, they reciprocate this behavior in terms of better job engagement [33]. Therefore, CSR proves to be one of the agents in building job engagement [1] and employees’ enthusiasm [15]. However, we must take into account that the relationship is weak; this could be due to the variables that are interacting in the model, but also the lack of a variable affecting this relationship that can play a vital role in increasing the connection, for example, job satisfaction or motivation. Indeed, CSR may lead to motivation and job satisfaction, which engender job engagement [38]. At the same time, we must consider that CSR is an ongoing project, often shaped by credible images and actions that might serve as motivators in the daily practices of individuals [15].

Contrary to what the literature suggests, the relationship between employees’ CSR perceptions and meaningfulness (H2), although positive (and with a significant correlation), was not significant, and therefore not confirmed. The fact employees perceived that the company values its various stakeholders may impact the way employees derive meaningfulness [23], which is in line with a deontic perspective of organizational justice [93]. However, the effect of other sources of meaningfulness may be lacking to account for a better explanation of this relationship [13], which may be a reason that the relationship was not verified in this study. As discussed further regarding H5, our study shows that procedural factors, such as the approach used in implementing CSR practices, intervene in this relationship.

This investigation also reveals that meaningfulness has a positive impact on job engagement (H3). This significant direct relationship is in accordance with the literature, as as meaningfulness is highly referred to as a driver for employee engagement [57]. Meaningfulness is a way through which individuals feel a sense of purpose and belonging [12]. It is the feeling that they are in the right place that allows individuals to experience meaningfulness at work, which further leads individuals to feel more engaged [4,12] and empowered to use their own judgement [15]. This is in line with the self-concept theory, which posits that the meaningfulness experienced at work usually implies an increase in the level of motivation and, in consequence, an increase in job engagement [94]. This means that if workers do not experience meaningfulness at work, it will jeopardize job engagement [95]. Thus, it is important to ensure that an individual considers their work as meaningful.

On the other side, if we analyze the results obtained with meaningfulness as the mediator between CSR perceptions and job engagement, we verify that this mediation does not exist, rejecting H4. The use of meaningfulness as a mediator variable has been tested among academics [1,23]; however, as far as we know, this is the first time that it has been used to examine the relationship between employees’ perceptions of CSR and job engagement. While it has been speculated that perceptions of employees about CSR impact meaningfulness and can offer an additional source of engagement of employees at work [3], this investigation does not confirm it. This could be due to the influence of the other variables (calling and bottom-up) that could be affecting this mediation, as they can better explain the effects of CSR on job engagement.

Adding calling orientation as a mediator in sequential mediation (CSR perceptions–calling–meaningfulness–job engagement) (H6), we come to the conclusion that there is a mediation effect between CSR perceptions and job engagement, where the indirect effect is medium (β = 0.230). To the best of our knowledge, this is the first time this sequential mediation has been tested, demonstrating that perceptions about CSR can positively create a higher calling orientation [96]. In fact, the way an individual perceives CSR actions can have many outcomes, affecting not only “the self” but also the work context. These perceptions will directly stimulate calling orientation, as they are associated with fulfilment of the value employees derive from work [96]. It is this fulfilment that will in turn positively affect the meaningfulness experienced, as those with calling are provided with the sense of meaningful work [13,21]. As stated previously, the positive impact on meaningfulness will generate more engaged workers [12]. Thus, this reinforces the theory about the sources of meaning and the theory of self-concept theory, showing that they are connected, as the effects of CSR and calling orientation will increase the meaningfulness (sources of meaning) [13], and meaningfulness positively impacts job engagement (self-concept theory) [94]. Finally, the impact of a bottom-up approach on the definition and implementation of CSR activities was tested in the relationship between the employees’ perceptions about CSR and the experienced meaningfulness (H5). Bottom-up was found to have a moderator role in the relationship between the two variables. Our results suggest that the variable representing a bottom-up approach presents statistically significant evidence (p-value = 0.038), even if weak (β = 0.094), of moderating the relationship between CSR perceptions and meaningfulness. It is important to note that, as referenced previously, the relationship between the two variables changes direction based on the bottom-up approach. This means that when we have a low participation of employees in CSR initiatives, indicating a low presence of the bottom-up approach in a company, the effects of CSR perceptions will dimmish the meaningfulness experienced, and generally the literature indicates that the possibility for employees to change norms and practices related to CSR may be limited [15]; however, if the bottom-up approach has a high presence in the company, meaning that workers have high participation on the definition of CSR initiatives, the employees’ perceptions about CSR will increase the meaningfulness experienced by the workers, as well as employees’ strong engagement, and indeed take leading roles in influencing CSR [15].

This result proves what has been conjectured, that when employees participate in CSR and its construction, they are more likely to find meaning, but on the other hand, if they do not participate, they are less likely to develop meaning [4]. Once again, as far as we know, this was the first time that this interaction effect was tested, contributing to the literature of understanding the influence of CSR initiatives on workers’ engagement at work. Moreover, this reinforces the argument that by being engaged at work the employee finds a way of reciprocating the organization’s efforts, which ensures that employees support and participate in CSR, according to social exchange theories [8,33].

6. Conclusions

In addition to filling the gap in the literature on the effects of CSR perceptions on job engagement, considering experienced meaningfulness and various levels of sensemaking, and also how variables relate to each other, our work proves the relationship between meaningfulness and job engagement supported by the literature. Furthermore, additional evidence emerges: the effects of CSR perceptions, impacting both calling orientation and meaningfulness, constitute another source of job engagement, as already proven by the literature by the direct relationship between perceptions about CSR and job engagement. Moreover, the bottom-up approach to the definition of CSR actions impacts the relationship between employees’ CSR perceptions and the meaningfulness experienced by them.

In terms of theoretical contributions, based on some suggestions [4], the present investigation included the analysis of cross-level interactions of sensemaking factors, namely, calling orientation (intraindividual level) and the bottom-up approach (intraorganizational level), to evaluate the effects of CSR perceptions on meaningfulness, and subsequently on job engagement. Moreover, we also contribute to the development of a new measure for examining the bottom-up approach, capturing the extent to which employees are involved in the definition of CSR activities and their implementation process. This new scale was adapted from the other two scales [82,83] that captured the degree of participation of employees in managerial decisions but, to the best of our knowledge, no scale evaluates specifically the bottom-up CSR of companies. Lastly, we contribute to the topic of the implementation process of CSR in companies. Our study helps the development of the literature, proving what was speculated: when employees have an active role in the definition of strategies regarding CSR, perceptions regarding the initiatives will be more positive since they contribute to the actions’ creation and development, increasing the meaningfulness experienced by the employees. Ultimately, this meaningfulness experienced will generate a more engaged workforce [3].

In terms of practical contributions, understanding how CSR actions undertaken by a company are perceived by employees and how these perceptions affect their job engagement is an important input for organizations [97] in order to help them shape programs that will not negatively affect workers’ job engagement and arise the “greenwashing” feeling [4]. With this in mind, our main contribution relies on the fact that, if employees perceive that CSR actions towards various stakeholders are positive, and not only for public relations purposes, employees will be prone to have a calling orientation, which will increase the meaningfulness experienced, resulting in more engaged workers. In this sense, it is important that organizations pay attention to how workers perceive and react to the CSR actions undertaken by the company, as these perceptions can result in more satisfied engaged workers. We might also identify impacts in autonomy which can empower employees and strengthen their competence related to CSR, meaning that views and practices over time become closer and might powerfully anchor CSR in the organization.

In terms of limitations and future research, future studies that increase the sample size to obtain a more representative sample of the population and allow a better understanding of sociodemographic and professional characteristics could be interesting, for example, age, level of education or seniority, and employees’ perceptions about CSR and job engagement.

In this study, we used two cross-level sensemaking factors to evaluate how they affect the relationship of the effects of employees’ CSR perceptions on job engagement. There are a great variety of sensemaking factors, and it would be interesting to consider the way other factors influence or add other factors to the models of our research and evaluate the way they interact with the variables.

Moreover, our study focuses on positive experiences with corporate social responsibility and the way it positively affects job engagement. Another suggestion for future research is about assessing the negative side of the perceptions and experiences with CSR, which is not broadly explored in the literature [4]. Additionally, instead of examining only the bottom-up approach in the company, the role of both implementation processes could be considered at the same time, in order to analyze which may have a higher impact.

Finally, our study has aggregated employees’ perceptions about CSR initiatives to various stakeholders (society in general, employees, customers and government). It would also be interesting to isolate the perceptions about CSR towards internal stakeholders from those towards external stakeholders, and in this way evaluate which of them play a more important role in the performance of the individual.