Towards a Sustainable Leader-Follower Relationship: Constructive Dissensus, Organizational Virtuousness and Happiness at Work (HAW)

This paper examines how leaders and followers can interact positively under complex and bounded conditions. For this purpose, this paper conceptualizes and measures the concept of constructive dissensus. Constructive dissensus relates to a mutual understanding leading to a situation of coregulation. In addition, the relationship between constructive dissensus and happiness at work is examined on the basis of affective events theory. Furthermore, the mediating role of organizational virtuousness is assessed. Through structural equation modeling, a multilevel analysis was performed. Data from 130 bank branches and 606 employees were gathered. The results reveal a direct effect of constructive dissensus on employees’ happiness at work. Moreover, a partial mediation effect of organizational virtuousness was found in the relationship between constructive dissensus and happiness at work.


Introduction
Leadership has traditionally been seen as a relationship, but research assumes that the leader defines the relationship, which may be moderated by follower characteristics. The debate on relations between leaders and followers also puts the focus on how leaders and followers influence each other and achieve common meanings through co-construction of a shared reality [1]. However, business life is complex, and leadership is frequently simplified. There is a limited understanding of leader-follower interactions in terms of workplace dynamics and social contingencies. Cunliffe and Eriksen [2] focused on the moral and dialogical practice in the leader-follower relationship, where collaboration, empathy and trust should be present. Therefore, we need to study in more depth how leaders and followers might converge and whether and how they improve quality of life at work.
In the current highly competitive environment, many academics and practitioners have become increasingly interested in managing follower-leader interactions [3,4]. The convergence or divergence between leaders and followers is a crucial aspect that deserves close attention [1], in particular in banking services. However, more frequently than expected, the relationship between branch managers and their subordinates is certainly not as good as it ought to be. In this vein, it is necessary to explore the mechanisms that facilitate improving these relationships and empirically check their efficacy in terms of better attitudes and quality of life at work. Accordingly, we propose the novel concept of constructive dissensus, understood as a situation of harmony that emerges from the mutual regulation of feelings and behaviors that reduces each other's negative emotions, and we suggest that constructive dissensus might positively affect happiness at work (HAW). Furthermore, a positive mediating role of organizational virtuousness is expected in the relationship between constructive dissensus and HAW.
Leadership is attractive to the leaders because it strengthens their sense of being exclusive and special, while providing it depends on context and power [5], and leaders need followers to exist [6].
The importance of followers comes from the relational nature of leadership. A good follower is expected to trust and rely on the leader, show enthusiasm, reveal dependence and, to a certain extent, try to follow their leader's ideas and behave in a civil way towards other employees at work [6]. In some traditional views of followership, leaders are believed to build followers, who are seen as objects and are embodied in an asymmetrical relationship [7]. On the other hand, other approaches, such as the one put forward by Hersey and Blanchard [8], suggest that followers affect the level of leadership: leadership does not always have a mechanical effect, and followers are active actors that take part in leader/follower interactions. More recent research considers followers as the constructors of leadership [6] in a scenario in which they exert significant influence (either as passive or active players).
It is a fact that the relational mechanisms that explain positive experiences between leaders and followers are of increasing interest [5], but real-life organizations are complex, and leadership processes are harder to explain and manage. Followers differ in their attributes, ranging from passive to active [9], and the work context and other organizational circumstances can affect leaders. In addition, the type of tasks, skills, educational backgrounds, psychological characteristics, organizational culture, human resource policies or highly demanding goals can significantly influence leader/follower relationships.
Following Alvesson and colleagues [6], we call for reflection among leaders and followers in order to improve their understanding of one another and, as a result, develop an optimal working relationship configuration. Leader-follower dynamics are complex, and mutual comprehension of these dynamics is fundamental for any organization [10]. We argue that constructive dissensus encourages feeding off each other and working with each other [11] rather than adjusting to each other, as stated by Mintzberg. Therefore, the first objective of this paper is to conceptualize and measure the concept of constructive dissensus.
A high-quality relationship between leaders and followers is expected to generate positive outcomes [5,12]. Based on the literature on interpersonal affect dynamics [13], we argue that the dyad itself, not the individual employees, is a fundamental unit of observation [4]. An inseparable connection exists between the dyad members involving leader and follower differences in their responses to the external environment, their mutual reactions to each other and their mutual regulation of feelings and behaviors [13]. In this sense, this study suggests that constructive dissensus might improve quality of life at work, understood as HAW. HAW [14] is defined by attitudes of engagement, job satisfaction and affective organizational commitment. In building our model and developing our objectives, we rely on social exchange theory [15], which is a renowned theory for the exploration of social exchanges in the workplace. Social exchanges lead employees to evaluate the extent to which the leader/follower values their work and cares about their quality of life at work and well-being. In this vein, HAW can be considered a positive outcome stemming from the socioemotional job resources developed through social exchanges. Employees who work in close proximity will be directly influenced by helpful information, feedback and practical support derived from a high-quality relationship, thus fostering HAW. Furthermore, the social identity approach allows us to argue that both leaders and followers need to think that they are partners in a positive and shared connection, namely, a sense of shared group membership. Social identity refers to the individual's self-concept associated with their belonging to a social group [16]. In addition, affective events theory provides further support by explaining how constructive dissensus promotes HAW. Affective events theory was developed by Weiss and Cropanzano [17] and explains how emotions and moods influence positive attitudes. Under this theory, cognitive-based behaviors are the best predictors of positive attitudes. Constructive dissensus sharpens judgment and increases the flexibility of both leaders and followers in adapting to the context [18]. Bunjak and colleagues [19] argued that leader-follower congruence promoted positive attitudes. Constructive dissensus can promote the capacity for dialogue, respect and positive connections between leaders and followers. Based on the above, our second objective is to assess the effect of constructive dissensus on HAW.
A mutual understanding derived from constructive dissensus means that the leader understands his or her followers and vice versa, demonstrating citizenship, honesty and integrity. As a result, values including optimism, trust, compassion, integrity and forgiveness are expected to emerge, thus displaying the unidentified path connecting constructive dissensus and HAW. The concept of virtuousness represents the above values and can be defined as the state of excellence in human or organizational character [20]. Virtuousness prevents organizations from malpractice and misbehavior, and makes it easier for employees to pursue higher levels of individual and societal benefits [21]. While some research on leadership has focused on different effects on positive attitudes [19], the interplay of organizational virtuousness (OV) in the aforementioned relationship has surprisingly been ignored. Accordingly, as a third objective, the present study aims to take a step forward in addressing this gap by assessing the mediating role of OV in the relationship between constructive dissensus and HAW.
In short, this paper takes the critical constructionist view of leadership and extends it by proposing a theoretical model that presents a more essentialist and functionalist approach to leadership. Specifically, the objectives of this research are twofold: (1) to examine the effect of constructive dissensus on HAW and (2) to analyze the mediating effect of OV in the relationship between constructive dissensus and HAW. In contrast to placing emphasis on leadership, we suggest a "nonheroic" view of the leader-follower relationship where collective individuality and freedom via self-expression, harmony and conciliation [22] better represent reality. In this sense, we conceptualize the leader-follower relationship as a dyadic interpersonal process as a result of microlevel relational processes.
This paper is organized as follows. First, we provide a theoretical review of the concepts of constructive dissensus, HAW and organizational virtuousness. Second, we explain the methodology and present the results. Finally, we propose a discussion and future lines of research.

Constructive Dissensus, HAW and OV
For a high-quality relationship between leaders and followers, it becomes significant to have an alignment of beliefs, perceptions and cognitions. Constructive dissensus represents a marriage between leaders and followers where they live in harmony [23]. This is the "constructive" part of the concept. At the same time, criticality and reflexivity need to be present because openness is necessary to provide a broader view and new approaches to improve relationships. Therefore, critique and questioning also need to be considered as a way to promote critical debates, and this is the "dissenting" side of the concept of constructive dissensus. This can be considered a complement to the "constructive" dimension of the concept, as it is mainly focused on providing an open and enriching view of current relationships that challenges what is taken for granted in order to strengthen and improve leader-follower ties. From the above, the concept of constructive dissensus emerges as a novel concept, following a critical view of traditional leadership studies. This concept has not been examined in prior studies and, therefore, provides a promising field of research for leadership studies.
The concept of HAW was conceptually proposed by Fisher [24] and empirically checked by Salas Vallina et al. [25]. It was derived from the positive approach to human resource management to address the progressive dehumanization of organizations [26]. HAW is defined as a positive emotional state involving happy feelings towards the job itself, the job characteristics and the organization as a whole. It emerged as a response to the need for sufficiently explanatory measures of positive attitudes, as existing ones were quite narrow, and some of them overlapped with each other [27]. Based on the proposal of Harrison et al. [28], who underlined the need for a higher-order construct going beyond job satisfaction and organizational commitment, and based on the solid literature review, Salas-Vallina et al. [14,25] developed and empirically checked the HAW scale. However, only specific leadership styles have been taken into account as antecedents of HAW, and constructive dissensus might provide a broader approach to explain how and why HAW emerges.
A key element of positive organizational research is the concept of virtuousness. Organizational virtuousness refers to enablers that enhance and preserve hope, optimism, gratitude, forgiveness, compassion and similar virtues [27]. Although studies on business ethics or citizenship behavior exist, virtuousness is singular and different from these concepts. The importance of virtuousness comes from its definition of what is good to and for human beings, so neglecting its investigation hinders our understanding of positive organizational outcomes, such as HAW. Organizational virtuousness is related to actions with human effects, including improved positive attitudes and behaviors, with moral goodness, which represents what is worthy of cultivation [27], such as love or wisdom, with an honorable and moral component. In addition, virtuousness comprises social betterment and creates social value that goes beyond the instrumental desires of the individual. It means that virtuous attitudes or behaviors are not oriented toward obtaining external benefits, but they are their own reward [28].

The Relationship between Constructive Dissensus and HAW
Over the last years, leader-member exchange theory has been acknowledged as a central approach for exploring organizational relationships. According to this theory, positive interactions between leaders and followers result in several benefits for both individuals and organizations [29], including improved job attitudes. There is evidence that leader-member exchange relationships are predisposed to generate affectively tinged connections between leaders and followers. However, to date, research remains limited, and further studies are required [30]. We argue that research needs to pay further attention to the emotional effects of improved leader-member relationships.
When the leader and follower interact in such a way that they mutually reduce each other's negative emotions, a situation of coregulation emerges [2]. In this sense, constructive dissensus [6] shifts the attention from studying how leaders influence followers to examining the emergence of a dyadic interaction as a result of a relational process. Such a constructive relationship is illustrated by Burns [31], who underlined the fact that leaders and followers could raise one another to higher levels of morality and functioning. Constructive dissensus aims to connect with organizational and social realities, highlighting leaders' and followers' self-and mutual understanding, and offers clues to evaluating how employees behave in the leader-follower relationship.
We argue that constructive dissensus might directly affect HAW in the banking context. Fisher [24] suggested the need to differentiate between three perspectives for happy feelings: (1) the job itself, (2) the work characteristics and (3) the organization as a whole. In this sense, Fisher defined HAW as a three-dimensional higher-order concept to be measured by means of engagement, job satisfaction and affective organizational commitment. Later, Salas-Vallina and Alegre [14] implemented and checked the psychometric properties of HAW and proposed a measurement scale considering the aforementioned dimensions.
Engagement involves high levels of energy and positivity related to the capacity to be passionate about work. It captures feelings of special motivation at work. Banking employees who performed their tasks in a climate of positive interactions would show increased vigor and dedication at work. Job satisfaction represents objective evaluations related to feelings about working conditions, such as salary, promotion opportunities and work-life balance. Higher-quality relationships imply healthier and better working conditions, as employees are less likely to suffer from stress and pressure, which are typical psychological and physical consequences for banking employees. This is because coworkers and/or the leader provide support and continuous feedback. Affective organizational commitment refers to employees' special connection with an organization. When employees connect with each other, they feel involved in a working group, thus fostering feelings of belonging, which create a sense of security. This feeling of group membership has a major role in explaining the emergence of HAW, as a positive social identity appears in individuals who perceive that their motivations and positive attitudes are psychologically interchangeable [32]. This is because banking employees typically compete aggressively against each other. However, if they collaborate and cooperate, providing mutual support, defensive and negative feelings turn into positive ones. In sum, a positive connection between the leader and the follower is expected to improve positive attitudes of engagement, job satisfaction and affective organizational commitment, as reducing the friction in the leader-follower relationship will increase their quality of life and positive perceptions at work.
Affective events theory provides an additional explanation for the connection between constructive dissensus and HAW. The leader-follower relationship is likely to result in affective interactions [30] that foster positive attitudes. We suggest that constructive dissensus acts as a cognitive and emotional event, resulting in increased HAW levels. As argued by Cropanzano and colleagues [30], by integrating affective events theory with leader-member exchange theory, we can better explain the critical role played by the positive emotional experiences between leaders and followers. Constructive dissensus could be seen as a resolution strategy that transforms tension into a more manageable situation [32]. The leader-member exchange has been linked to engagement, following affective events theory [33]. Therefore, the potential antecedents of HAW may be related to constructive dissensus since the relationship with the leader includes an emotional component [34] and is important for positive attitudes [35]. To our knowledge, no previous research has examined the effect of the positive interplay between leaders and followers on HAW. Therefore, we suggest the first hypothesis: Hypothesis 1 (H1). Constructive dissensus has a direct and positive effect on HAW.

The Mediating Role of OV in the Relationship between Constructive Dissensus and HAW
The Latin root of the word "virtuousness" is "virtus", meaning "excellence". Aristotle understood virtue as the desires and actions that create personal and social good. More recently, virtuousness has been defined as the search for the maximum ambitions [36]. Virtuousness is what humans aspire to be when they are at their best condition or state [37]. A virtuous character is a fusion of virtues, personality traits and values that foster excellence [38] and, in turn, promote optimism. A virtuous organization involves organizational conditions where "good habits", desires and actions are undertaken, supported, cultivated, diffused and perpetuated at individual and organizational levels [37]. Optimism, trust, compassion, integrity and forgiveness are common attributes of virtuous organizations. Virtuous organizations enable virtuous practices on the part of their members. Organizational virtuousness is associated with three essential attributes: moral goodness, human impact and social betterment. Moral goodness refers to what is good in itself, with no expected benefit (love, wisdom or fulfillment). However, reality shows that most organizations, and particularly banks, tend to exploit human resources for profit [5].
We suggest that OV would help to explain how and why constructive dissensus fosters HAW. In banking, although employees tend to work individually, they need team support to address problems and challenging objectives. On the one hand, a banking branch requires employees to perform specialized tasks. For example, an employee manages cash and serves visitors to the bank, while others focus on private banking, small businesses or medium-sized firms. On the other hand, tasks are interdependent, as any employee may provide the initial service to any customer (face-to-face or by telephone). Furthermore, when an employee is away visiting clients, other employees have to provide service to the client. Therefore, the nature of this work requires employees to trust each other and to collaborate with one another for achieving organizational goals. To this end, values such as trust, optimism, compassion or forgiveness are central. An employee serving the client of another employee might show opportunistic behaviors, such as talking negatively about a peer or stealing the client. In the short term, this could benefit the selfish and unethical employee, but in the medium and long term, it undermines confidence and creates division and mistrust, thus harming not only positive attitudes but branch performance. For all these reasons, OV at the branch level is fundamental and might reveal why constructive dissensus leads to HAW.
Based on social exchange theory, we posit that constructive dissensus has a positive effect on OV. The nature of constructive dissensus encourages leaders and followers to improve social interactions. A social exchange arises when the leader works for an improved relationship with followers, who positively respond with corresponding attitudes and behaviors. Furthermore, leadership emerges strongly from the system of interpersonal relations [29]. In other words, informal contacts and strong personal ties are strengthened when both the branch manager and his or her followers positively contribute to improving their relationship. Constructive dissensus means that both leaders and followers are willing to interact and share their points of view, and it provides the necessary openness and optimism. For example, conscientiousness and openness are considered to be virtuous traits [39]. Reflection in leaders and followers constructively aligns them because it clarifies comprehension, moving towards a common understanding that implies a relationship of trust. Followers that put themselves in the shoes of the leader, care about each other and show acts of concern generate an interactive reflection with the leader, which embodies compassion and forgiveness. In a recent study, Booker and Dunsmore [40], in line with previous psychological research, revealed an interesting effect of compassion on subjective well-being. When the leader understands his followers and demonstrates citizenship and a capacity for dialogue, this enables honest communication and integrity with the follower. In fact, people who behave dishonestly experience unpleasant tension. These ties imply that leaders and followers know each other better [41] and help leaders and followers bridge the gap between their conflicting points of view [42], thus enhancing optimism. Efforts to find common ground are essential for both banking leaders and followers. Not only branch subordinates but also branch leaders feel more secure and empowered when they feel connected to each other.
We also build on social identity theory to strengthen the connection between constructive dissensus and OV. The social identity salience from constructive dissensus leads individuals to feel that their motivations and positive attitudes are interchangeable with those of coworkers sharing the same identity [43]. According to social identity theory, individuals share self-defining values, norms and goals with other employees of a group. Constructive dissensus fosters mutual understanding and shared beliefs and emotions, which creates a common identity among branch members. It is the effort of both the branch manager and his or her subordinates that creates mutual understanding. Then, they feel that the branch challenges, objectives and goals affect every branch employee, and it is necessary to pull in the same direction. However, social identity also enables individuals to agree with each other on issues regarding their work group. Furthermore, a shared social identity leads to increased motivation and working collaboratively. Accordingly, social identity salience can strengthen positive attitudes and behaviors, such as trust, collective action [44], optimism, gratitude or compassion, namely, OV.
In turn, truthful and authentic followers that show consideration towards the leader and actively participate in task performance facilitate a positive connection with the leader. Bright and colleagues [15] argued that leaders with a virtuous character are able to produce social betterment.
Social exchange theory is a suitable framework for explaining why OV might foster HAW. Employees experiencing positive working conditions that benefit their psychological states and their performance are expected to replicate it with positive attitudes, such as HAW. Furthermore, as argued above, social identity theory explains why people's behavior is underpinned not by personal identity but by a sense of mutual group membership or social identity. Virtuousness has been linked to customer retention, innovation, quality, turnover and profit margins [13]. Optimism, a characteristic of OV, has been proved to be a strong positive antecedent of well-being [45]. This is interesting because virtuousness in banking branches might predict HAW. Social trust has been shown to significantly affect subjective well-being [46], and trust in leadership has been proved to affect positive attitudes [47]. Helping behaviors (compassion) have been found to predict positive emotions, including happiness [48]. Compassion increases meaning [49] and, in turn, happiness [50]. Yukl [51] (p. 331) understands integrity as "honesty and consistency between a person's espoused values and behaviour", which has been connected to trust [52]. Finally, exploring the relationship between forgiveness and happiness has also been suggested as an area for research, arguing that forgiveness could have a positive impact on happiness [53].
Employees tend to treat organizations in the same way they are treated by their organizations by adapting their attitudes in an appropriate way. Positive judgments of OV lead employees to feel proud to work in their organization [54] and to increase their positive attitudes, such as affective commitment [55]. As employees feel more valued, they give everything they have got, making them more affectively attached to their jobs [56]. Accordingly, positive experiences at work may lead to pleasant emotional associations with the workplace, strengthening positive attitudes [26]. As Cameron and colleagues [29] argued, OV engenders positive emotions. Therefore, in light of the above, our second hypothesis is as follows: Hypothesis 2 (H2). OV mediates the relationship between constructive dissensus and HAW. Figure 1 shows the proposed theoretical model. more affectively attached to their jobs [56]. Accordingly, positive experiences at work may lead to pleasant emotional associations with the workplace, strengthening positive attitudes [26]. As Cameron and colleagues [29] argued, OV engenders positive emotions. Therefore, in light of the above, our second hypothesis is as follows: H2. OV mediates the relationship between constructive dissensus and HAW. Figure 1 shows the proposed theoretical model.

Study Context
The banking sector is an essential element of the economy in most countries, playing an intermediary function between investors and agents needing funds. In this sector, the role of the branch manager, acting as a leader, is a very appealing option. Banking is a particularly interesting area of research for leadership studies, as its leaders have played a central role during and after the financial crisis. In this sector, middle managers (branch leaders) are expected to be able to challenge, influence and change followers. However, leadership might be conditioned by the context. For example, circumstances vary considerably from one bank branch to another, and therefore, a universal leadership style will not always work. This is the case of middle managers led by authoritarian (or participative) superiors, who may hinder (or facilitate) their leadership tasks. At the same time, followers have diverse degrees of motivation, education and attitudes towards work, which facilitates dissensus. Furthermore, reality shows that banking objectives demand unachievable goals under ruthless working pressure [8], which creates disagreements and tension between leaders and followers. This has led us to rethink the virtue of specific leadership styles, which may not be the solution to a universal mode of organization. In this respect, a high-quality relationship among banking branch employees becomes central for both individuals and organizations.
Our target population was employees working in bank branches specialized in medium-and large-sized companies. We focused on these bank branches because the relationship between leaders and followers was deemed to be of absolute importance in this complex context. Bank branches are structured in working teams, including a branch manager, an associate manager and the sales force. They are small working units in which employees work in especially close proximity, and thus, good linkage is essential. This study focused on branch leaders and associate managers.

Study Context
The banking sector is an essential element of the economy in most countries, playing an intermediary function between investors and agents needing funds. In this sector, the role of the branch manager, acting as a leader, is a very appealing option. Banking is a particularly interesting area of research for leadership studies, as its leaders have played a central role during and after the financial crisis. In this sector, middle managers (branch leaders) are expected to be able to challenge, influence and change followers. However, leadership might be conditioned by the context. For example, circumstances vary considerably from one bank branch to another, and therefore, a universal leadership style will not always work. This is the case of middle managers led by authoritarian (or participative) superiors, who may hinder (or facilitate) their leadership tasks. At the same time, followers have diverse degrees of motivation, education and attitudes towards work, which facilitates dissensus. Furthermore, reality shows that banking objectives demand unachievable goals under ruthless working pressure [8], which creates disagreements and tension between leaders and followers. This has led us to rethink the virtue of specific leadership styles, which may not be the solution to a universal mode of organization. In this respect, a high-quality relationship among banking branch employees becomes central for both individuals and organizations.
Our target population was employees working in bank branches specialized in medium-and large-sized companies. We focused on these bank branches because the relationship between leaders and followers was deemed to be of absolute importance in this complex context. Bank branches are structured in working teams, including a branch manager, an associate manager and the sales force. They are small working units in which employees work in especially close proximity, and thus, good linkage is essential. This study focused on branch leaders and associate managers.

Sample and Measurement
The population chosen for this study was composed of a total of 418 branches, which included 418 branch leaders, 418 associate managers and 2090 sales staff working at the five major banks in Spain and Italy; 130 branch leaders and associate managers (from 130 branches) responded to the questionnaire, representing a response rate of 31.10% at the managerial level. 606 employees returned valid questionnaires (29% from the total population). Questions related to constructive dissensus were addressed to branch leaders and associate managers, as they were more aware of their personal attitudes and behaviors towards each other. The questions about organizational virtuousness and happiness at work were addressed to branch leaders, associate managers and sales staff. We assured all participants that their responses would be strictly anonymous and would only be used for academic purposes. Women represented 54% of the total sample, and the average age of respondents was 36.45 years of age.
To measure constructive dissensus at the branch level, we developed a measurement scale consisting of nine items. Items were based on Alvesson and colleagues' [6] reflection about questions that take a critical view of what one is doing and promote an open and constructive debate. Leaders and associate managers were asked questions such as, "I am truly reflexive concerning my work situation, views and beliefs", ranging from one ("totally disagree") to five ("totally agree"). The items are listed in Appendix A. A pool of experts from five different universities was asked to review the measurement scale and discuss each item. Later, exploratory factor analysis was carried out, with the selected nine items loaded into a single factor. The Cronbach's alpha for this scale was 0.96. An inter-rater reliability (rwg(j)) test was also conducted to assess the level of inter-rater agreement for constructive dissensus within the work groups. The median rwg(j) of constructive dissensus across 130 branches was 0.98, suggesting acceptable consensual validity (James et al., 1993). Intraclass correlations ICC (1) and ICC (2) results for constructive dissensus were 0.29 and 0.58, respectively. Both the ICC and rwg(j) results suggested that aggregation at the branch level was justified [57].
To measure OV at the branch level, we used Cameron and colleagues' [29] measurement scale, consisting of five dimensions (social optimism, trust, compassion, integrity and forgiveness) and 15 items. We asked branch leaders, associate managers and employees questions such as, "Employees trust one another in this organization", in which respondents had to report the degree to which the statements were false (one) or true (five). The Cronbach's alpha for this scale was 0.90. Inter-rater reliability (rwg(j)) and ICC tests were also undertaken to assess the level of inter-rater agreement for constructive dissensus within the work groups. The median rwg(j) of organizational virtuousness across 130 branches was 0.89, suggesting acceptable consensual validity (James et al., 1993). ICC (1) and ICC (2) results were 0.42 and 0.74, respectively, suggesting a justified aggregation of data at the branch level.
To measure the SHAW (shortened version of HAW) at the individual level, we used Salas-Vallina and Alegre's [14] measurement scale, consisting of nine items asked to every employee (including branch leaders, associate managers and lower job positions) (i.e., "At my job, I feel strong and vigorous") ranging from one ("totally disagree") to five ("totally agree").
Due to the nature of this multilevel study, we used both branch-level and individual-level controls. At the branch level, we controlled for branch age and firm performance. Firm performance included an average of three performance indicators: profits, sales growth rates and market share. At the individual level, we controlled for age, gender and task performance. Task performance was measured using Williams and Anderson's [58] six-item scale (i.e., "I adequately complete assigned duties"). Cronbach's α for this scale was 0.88.
Psychometric properties of the measurement scales were as follows.
Six criteria were used to check the three constructs of the proposed model: construct dimensionality; content, convergent and discriminant validity; composite reliability; and average variance extracted (AVE) [59].
Dimensionality assessed whether the factorial structure was appropriate for configuring each construct. We checked dimensionality by using structural equation modeling through confirmatory factor analysis, which enabled us to establish a priori the latent variables and the connection between them and the observable variables [58]. The results revealed that all the estimated parameters for the three scales were sufficiently high and statistically significant [60], revealing a very good fit [61]. When checking the scale to measure constructive dissensus, the nine items of the original scale were found to be significant, charging over its one dimension. The scale used to measure organizational virtuousness showed that the 15 items of the original scale presented significant loading, revealing five discrete subfactors. The scale used to measure happiness at work revealed nine significant loadings on its one dimension. Table 1 shows the correct fit of the constructive dissensus, OV and HAW. The p-value of the chi-square statistic was below the 0.05 significance level for all the scales, and the other indicators showed the correct dimensionality of the constructs. The validity guaranteed that the scale captured what it aimed to capture. We examined content, convergent and discriminant validity.
Content validity allowed us to affirm that the items reproduced the construct and that they were easy to respond to. The procedure used to select the measurement scales provided support to content validity. The constructive dissensus items were taken from Alvesson's [6] academic work. Both the OV scale and the SHAW scale have been thoroughly reviewed in the literature. Hence, we can assure the content validity of the three scales.
Convergent validity assessed whether the measure used had a high correlation with other measures assessing the same concept. We checked the Bentler-Bonet Normed Fit Index (BBNFI) indicator obtained in the Confirmatory Factor Analysis (CFA). The values in the BBNFI lay above the recommended value of 0.90 and the factorial loads for each construct were above 0.4 [61], confirming convergent validity.
We also examined discriminant validity for the OV scale by comparing two models for each scale, one of which was estimated by constraining the correlation to one. We demonstrated the discriminant validity of the OV scale using pairwise analyses through confirmatory factor analysis between all dimensions. Then, we estimated another model, setting the correlation between these dimensions and unity. The results revealed that the model fit better when the correlation between dimensions was different from unity, therefore demonstrating discriminant validity. Additionally, we checked discriminant validity for the constructive dissensus and the SHAW scales, and the results indicated that the correlation coefficients were significant and below 0.9, which also ensured discriminant validity. Following the Fornell-Larcker criterion, we also checked that the average variance extracted (AVE) of each latent variable was higher than the squared correlation between the constructs [61], again confirming discriminant validity.
To assess reliability, the R-squared (R 2 ) values were confirmed to be within the recommended range (above 0.50), and composite reliability values were above 0.70 [62]. Hence, we could confirm the reliability of the measurement scales.

Setting and Procedure
This research utilized quantitative analysis. We employed a multilevel structural equation modeling technique, or MSEM [63]. The proposed model reflected the individual differences between branch-level outcomes, and MSEM enabled us to treat upper-level variables as outcomes. In traditional multilevel models, upper-level variables cannot be included as outcomes [63]. Therefore, to test this effect, hierarchical linear modeling had to aggregate individual-level predictors and check these aggregated individual differences and higher-level outcomes. However, this approach significantly reduced the sample size.

Results
Descriptive statistics are presented in Table 2. Constructive dissensus was significantly correlated to both OV (r = 0.06, p = 0.02) and HAW (r = 0.09, p = 0.03). By contrast, branch age was negatively related to OV (r = −0.05, p = 0.01) and HAW (r = −0.08, p = 0.00). Financial performance was positively related to OV (r = 0.08, p = 0.00) and HAW (r = 0.23, p = 0.00). Age was negatively related to OV (r = −0.01, p = 0.00) and HAW (r = −0.02, p = 0.00). Gender was positively related to OV (r = 0.06, p = 0.00) and HAW (r = 0.10, p = 0.01). Finally, task performance was positively related to OV (r = 0.06, p = 0.00) and HAW (r = 0.12, p = 0.00). Two MSEM analyses were performed with M plus software [64]. Table 3 shows the results. In the first analysis (Model 1), constructive dissensus revealed a significant impact on OV (β = 0.92, p = 0.02) and HAW (β = 0.57, p = 0.04). These results led us to support Hypothesis 1. A one-unit increase in constructive dissensus involved a 0.92 unit increase in OV and a 0.57 unit increase in HAW. In the second analysis (Model 2), we added a link from OV to HAW. Model 2 showed that the link from OV to HAW was positive and significant (β = 0.71, p = 0.00). A one-unit increase in OV resulted in a 0.71 unit increase in HAW. In Model 2, the positive and direct effect of constructive dissensus on HAW increased (β = 0.78, p = 0.04), lending support to Hypothesis 2.
In sum, constructive dissensus had a direct and positive effect on HAW, and OV only partially mediated the relationship between constructive dissensus and HAW, thus strengthening this relationship.

Discussion
The aim of this article has been to examine the direct effect of constructive dissensus on HAW. In addition, this article was to check the mediating role of OV in the relationship between constructive dissensus and HAW. It must be emphasized that constructive dissensus is a novel concept that has been measured and conceptualized. Building on social exchange and social identity theories, the proposed theoretical model has been empirically checked, which has provided a valuable contribution to the literature on leadership and well-being at work.
High alignment between leaders and followers in banking means that they share meanings and efforts for a high-quality relationship. In this paper, followers' and leaders' self-understanding (constructive dissensus) has been seen to improve quality of life at work, measured by means of HAW. In addition, OV strengthened the effect of constructive dissensus on HAW.
First, this study shows that in banking, the leader-follower relationship is a complex phenomenon influenced by a broad range of internal and external factors [3,6]. In this sense, this piece of research highlights the importance of considering the interactions that occur in the leader-follower relationship in the banking branch context. Accordingly, this paper conceptualized the leader-follower dyad by means of constructive dissensus, a concept that emerged from the microlevel relational processes in the leader-follower connection. What seems evident is that leader-follower encounters can improve each other's quality of life at work as long as they reflect on their relationship in a constructive way. In light of this and despite the fact that leaders and subordinates have different objectives, traits, training, priorities and personal life situations, there is still hope: constructive dissensus could improve HAW, thus confirming H1. In other words, despite the competitive banking climate, it is possible to enrich and resolve branch manager-subordinate relationships by means of constructive dissensus. Leadership theory is therefore developed through affective events theory, where cognitive and emotional events at work might explain positive attitudes.
Second, OV has something to say on the relationship between constructive dissensus and HAW. The main idea is that constructive dissensus made for better organizations, thus influencing OV, which, in turn, positively impacted HAW. Our understanding of OV is based on the work of Cameron and colleagues [29]. According to these authors, virtuousness is associated with optimism, trust, compassion, integrity and forgiveness. Branch leaders and associate managers who showed constructive dissensus positively influenced OV. This means that bank branches can improve their working attitudes and behaviors through constructive dissensus. An improvement means benefits for both individuals (positive attitudes such as optimism and trust, namely, a better quality of work life) and for organizations (for example, supporting coworkers or compassionate behaviors). Past research revealed that traits such as helping others and organizational stewardship promote organizational virtuousness [65]. Then, a direct and positive effect is expected between constructive dissensus and OV.
Undoubtedly, OV can be considered a job resource, and as such, it is expected to affect positive attitudes. Job resources facilitate job development and result in positive consequences for both employees and organizations [66]. Virtuousness is related to happiness [67]. Employees adapt their attitudes according to how they feel they are treated by the organization. Accordingly, OV might lead employees to feel gratitude for working in their organization [68], and this might increase their well-being. Our results reveal that constructive dissensus positively affected OV, and OV positively influenced HAW in such a way that OV partially mediated the relationship between constructive dissensus and HAW. Such an outcome is central, as the banking industry is characterized by poor well-being levels, and this research provides a useful model that can benefit both individuals and organizations. To date, scarce research has explored the benefits of improving HAW in banking services [14].
From a practical perspective, human resource managers should promote reflection among branch leaders and associate managers, yet reality shows that it is not common, with reflection occurring spontaneously among branch leaders and associate branch leaders in order to survive. Leaders work in situations of imbalance in which they have the maximum responsibility of the branch on their shoulders, yet they do not have the corresponding power to make decisions. From this starting point, dissensus arises and generates major problems, which can result in negative outcomes (burnout, unethical behavior, toxic attitudes, etc.).
In sum, this paper conceptualized and measured constructive dissensus, demonstrating its positive impact on HAW by means of OV. In this way, this paper developed leadership theory by connecting the leader-follower dyad to a favorable organizational context and positive attitudes. This research could provide guidance and coherence for future research on human resource management innovation, as our proposed model constitutes a positive framework for anchoring the study of classical and emerging concepts in the development of more reflective employees.
Future research might consider a more complex model by integrating other variables. It would allow for developing a novel approach to explain the complexity of the relational processes between leaders and followers. Variables including the work context, skills, educational backgrounds, psychological characteristics, organizational culture or human resource management policies could be introduced into the proposed model.

Conflicts of Interest:
The author declares no conflict of interest. The funders had no role in the design of the study; in the collection, analyses or interpretation of data; in the writing of the manuscript; or in the decision to publish the results. 9.
I and my manager are flexible and tolerant with each other.