Background
Occupational research has traditionally focused on reducing factors that cause stress and might lead to disease and infirmity, but an increasing emphasis on positive organizational behavior has shifted attention to individuals’ strengths and healthy functioning and the workplace conditions that facilitate them [
1‐
3]. Mental well-being and work engagement are two desirable, positive states of mind that help individuals to better function in the workplace [
4].
Both mental well-being and work engagement have individually received much attention. Based on a definition from the World Health Organization (WHO), an individual in a state of mental well-being “realizes his or her own potential, can cope with the normal stresses of life, can work productively and fruitfully, and is able to make a contribution to her or his community” [
5]. Mental well-being is widely considered to have both hedonic (i.e., feeling good) and eudaimonic components (i.e., functioning well) [
6,
7]. Employees with low mental well-being, for example, are more likely to be less productive and have more days of sick leave [
8]. Moreover, in Germany mental disorders have been identified as one of the most common reasons for days of sick leave (16.6%) [
9] and the most common health-related reason for early retirement (43%) [
10]. Work engagement, on the other hand, is commonly defined as a work-related state of mind that is positive and fulfilling and not focused on a single object, event, or person [
11]. It encompasses vigor (e.g., having high levels of energy, mental resilience, persistence), dedication (e.g., having a sense of significance, enthusiasm, inspiration, pride, challenge), and absorption (e.g., being fully concentrated and deeply engrossed, feeling time flies by, having difficulty to detach from work) [
11]. Work engagement is associated with greater life satisfaction [
12,
13], happiness [
14], and better health outcomes [
13,
15]. Other studies identify associations between work engagement and greater job satisfaction, better in-role and extra-role job performance, and lower absenteeism [
13,
16‐
19]. Additionally, a meta-analysis has found that work engagement is related to better business outcomes such as customer satisfaction, productivity, profit, employee turnover, and accidents [
20].
Occupational researchers have argued in favor of considering both states concurrently [
1,
21,
22]. Even though the research listed above indicates that both states are relevant for both employees and employers, mental well-being is a state that focuses on life as a whole and is therefore thought to be particularly important for employees, while the work-related nature of work engagement makes it also especially relevant to employers [
21]. Combining the employee perspective on well-being and health with the employer perspective on productivity has the potential to offer mutual benefit [
1]. For example, conceptual work suggests that an indicator characterizing both mental well-being and work engagement might be a better predictor for the success of organizational interventions (e.g., coaching) than monetary outcomes such as return of investments [
22]. Grant (2012) [
22] reasons that when both states are considered simultaneously, they offer a more direct and holistic view of what most interventions intend to address – that is, improvements in the behaviour or state of employees that should in the long run lead to several organizational benefits. Other work suggests that a narrow focus on only one of these factors in organizational interventions as an intermediate outcome measure will limit the more distal organizational benefits of said interventions [
4]. Indeed, their cross-sectional study indicates that mental well-being and work engagement are related states that simultaneously better explain variations in a common outcome of interest in organizational research: employee productivity [
4]. In general, previous literature and established models such as the job-demands resources model suggest that in the workplace both health-related and motivational processes operate to influence not only employees but also organizational performance indicators [
23‐
25].
Mental well-being and work engagement are positively associated with one another [
26,
27], however, studies that address both states empirically are relatively sparse [
28]. Even fewer studies have argued, how a concurrent consideration can be transformed to provide a single indicator that characterizes the workforce based on their mental well-being and their work engagement. Robertson and colleagues, for example, call for the addition of mental well-being to work engagement to obtain “full engagement” [
4,
21,
29]. They argue that while many engagement scales already include items describing well-being, they are not comprehensive enough to sufficiently capture mental well-being and therefore require a separate indicator of mental well-being. Full engagement is therefore a combination of being engaged and experiencing high mental well-being [
21]. However, while the authors establish that both states are moderately correlated, they do not demonstrate that they are distinct, even though both show independent associations with productivity [
4]. Previous work such as the “well-being and engagement framework”, conceptualizes the presence of a conjoint construct but, to our knowledge, has resulted in no empirically defined measurements for the proposed categories [
22]. This framework, for example, suggests that mental well-being and work engagement form a two-dimensional space in which the employees can be divided into four meaningful subgroups (flourishing, acquiescent, distressed but functional, distressed and disengaged). These subgroups characterize a workforce based on what combinations of high or low levels of mental health and work engagement employees are reporting [
22,
30,
31]. To the best of our knowledge, the extent to which this two-dimensional room can be separated into four or, indeed, any finite number of categories has not yet been demonstrated using empirical data. A further knowledge gap is the absence of evidence that any operational measure for this construct demonstrates change over time: such evidence would be needed to justify its use as an intermediate outcome in interventions studies.
The present study describes empirical efforts to operationalize a construct characterizing a workforce’s mental well-being and work engagement, which we will refer to as “engaged well-being”. These efforts will address three aims. First, based on previous research [
4], we quantify the extent to which mental well-being and work engagement are correlated and confirm that they are distinct states that can be used as two separate factors for further analysis. Second, assuming that these states are distinct, we test whether they can be divided into meaningful subgroups with distinct profiles. Although previous conceptual work hints at four subgroups, we will develop a categorization scheme that best fits data from a large database of employees. For this we use multiple statistical techniques and corresponding validation procedures to develop a robust taxonomy that identifies subgroups within a large sample that vary in potentially important ways with respect to the construct. Finally, we will use longitudinal data to test whether the newly constructed indicator can change over time.
Discussion
Previous literature has proposed a simultaneous consideration of both mental well-being and work engagement and various studies indicate that both states separately are associated with desirable outcomes for employees and employers. Using a large sample of employees, the present study added to previous research and considerations in three ways. First, additional support was provided that mental well-being and work engagement are moderately correlated. We have added to these correlations by providing evidence that both states are distinct. Second, it was tested whether these states can be divided into distinct subgroups by identifying profiles varying with respect to their average mental well-being and work engagement. The resulting subgroups can be described as 1) high mental well-being and high work engagement (engaged well-being), 2) high mental well-being and low work engagement (disengaged well-being), 3) low mental well-being and high work engagement (engaged distress), and 4) low mental well-being and low work engagement (disengaged distress). Replicability of the subgroups (or categories) was ensured by identifying and testing empirical cut-offs. The final construct is referred to as engaged well-being. Finally, we used longitudinal data to show that engaged well-being can change over time, indicating its potential use for intervention.
Our analyses mirror several conceptual considerations and previous empirical observations. First, in line with previous literature, we found that both mental well-being and work engagement are moderately and positively correlated [
4,
46]. The positive correlation between both constructs was to be expected as a) the work engagement scale includes items that describe well-being at work, such as “I feel happy when I’m working” and that b) previous studies have found significant positive associations both cross-sectionally and longitudinally [
26,
27]. Additionally, we have shown that both states are distinct, which supports a previous analysis showing that mental well-being and work engagement have distinct associations with productivity [
4].
Our findings additionally correspond with ideas formulated in the ‘engagement and well-being framework’, arguing that the dimensions mental well-being and work engagement can be comprised into four different categories occupying every corner of this two-dimensional space [
22,
31]. However, we did not predefine the number of categories for engaged well-being and instead chose established, data-driven approaches testing multiple cluster solutions. Still, the best solution provided in our analysis was the four-cluster solution. In line with the ‘engagement and well-being framework’, the resulting clusters or categories combine either higher or lower levels of mental well-being and work engagement. A notable distinction to the framework is that it defines the mental well-being dimension as a combination of mental well-being and mental illness [
22]. Our study uses the WHO-5, a generic scale with only positively formulated items that reflect mental well-being and not mental illness [
39]. The established cut-off is often used for screening in clinical depression trials, however, the WHO-5 itself has no diagnostic specificity [
39]. In line with ideas from the positive psychology approach, which argues that the good is more than the absence of the bad [
3], and based on previous literature that has shown that mental health and mental illness [
47], as well as positive and negative affect [
48] cannot be measured on a single dimension, we decided against a combination of both positive and negative health-related states in the mental-well-being dimension. Measuring mental illness separately or possibly as another dimension in engaged well-being could be valuable, especially because positive organizational psychology studies indicate that positive and negative phenomena explain unique variance of organizational outcomes [
1]. We do believe that engaged well-being could be an instrument used within the mental well-being and engagement framework.
Engaged well-being combines interests of employees and employers [
1]. While the motivational processes associated with work engagement are often the main focus for employers, taking employee health into account is not only part of the corporate social responsibility but also of business interests, as healthy employees are more productive and the image of an organization that takes care of its employees is likely to increase [
49]. One study proposes that, due to the positive association with productivity, organizations need to apply more holistic and multipronged approaches to improve work engagement and physical health by creating motivational work environments and providing health and wellness programs [
17]. Positive psychology interventions seem to be promising for enhancing both employee well-being and performance [
50]. While the present study is the first step in corroborating a simultaneous consideration of mental well-being and work engagement in the form of engaged well-being, we are aware that more work is needed to strengthen the construct and to establish it in organizational settings, especially in interventions. In the following section, we therefore discuss the application and further testing of engaged well-being.
Future research and application in organizational settings
Our operationalization of engaged well-being is easily replicable and can change over time, giving it potential to be applicable in workplace interventions. The approach we have described in measuring engaged well-being could be used in organizations to observe the distribution of employees across the different categories as well as the changes of this distribution over time. However, we still need to test, what antecedents lead to these changes. As proposed in the job-demands resources model, there are two pathways through which job characteristics influence employees – the health-impairment and the motivational pathway [
23]. The model assumes, that while job resources are thought to be predominantly positively associated with work engagement through the motivational process, job demands are mainly negatively associated with mental well-being through the health-impairment process. Several empirical studies have found support for the assumptions of this model (for reviews see [
23,
25,
51]). Based on these pathways, we assume different needs for changes between different engaged well-being categories. For example, employees in the category engaged strain might be more likely to change into the category engaged well-being if job demands (e.g., physical demands, time pressure, mobbing) are reduced, as this should increase their mental well-being (health-impairment process). Employees in the category disengaged well-being, on the other hand, might be more likely to change into the category engaged well-being if job resources (e.g., supportive leadership, organizational justice, decision-making autonomy) are increased, as this should increase their work engagement (motivational process). These assumptions also imply that not all employees within an organization would need the same type of support, depending on their engaged well-being, therefore implying multi-component interventions. However, while the direct associations described above have been shown in previous research, multiple studies using the job-demands resources model have also shown interaction effects between demands and resources [
23,
25,
49]. It is assumed that job resources do not only affect work engagement through the motivational process, but that employees with increased job resources are additionally better able to cope with the strain caused by job demands, therefore reducing their negative impact on health [
23,
25]. A more distinctive analysis of such interactions regarding engaged well-being is important to better understand the antecedents and processes that influence this new construct.
Workplace interventions can be used to test whether and how changes in job demands or resources can influence engaged well-being. By improving work conditions (e.g., increasing supervisory support or decreasing time-pressure), employers should be able to observe a shift away from disengaged strain towards engaged well-being. It is, however, important to note, that while the overall conceptual thoughts of the ‘engagement and well-being framework’ might be translated to both the employee and the organizational level [
22,
30], the use of engaged well-being within workplace interventions should be limited to observing changes within the overall workforce of an organization rather than within an employee, as the engaged well-being categories are rather broad and therefore not able to provide detailed information on changes within individuals.
Within such workplace interventions, a deeper understanding of the two categories of engaged strain and disengaged well-being needs to be developed. Why do people report being engaged while they are strained? One conclusion might be that these employees could be addicted to their work and thus risking their own mental well-being. However, work engagement has been defined as a positive state of mind and studies indicate that work engagement and workaholism are two different constructs [
13,
52]. Additionally, our results indicate that only every fifth employee that had been engaged strained in one observation reported the same in the next, and every second reported an improved change to engaged well-being, indicating that it might not be the higher levels of work engagement that result in strain. In contrast, the category disengaged well-being was more stable. Are these employees that do not care for their work and search for validation outside of the work environment? How can changes to engaged well-being still be encouraged (e.g., through better supervision)?
Additionally, a better understanding of the consequences of engaged well-being makes the indicator more attractive for use in praxis. As argued by Grant (2012) [
22], indicators that capture employee level engagement and well-being might be better indicators of organizational success than monetary business outcomes. In the short run monetary outcomes could be quickly improved by worsening working conditions (e.g., high pressure work environments). In the long run, engaged well-being should lead to more organizational success, as employees should have better resources to reach organizational goals and are less likely to ‘burn out’. Previous studies have found positive and distinct associations of mental well-being and engagement with productivity cross-sectionally [
4] or longitudinally using a physical instead of a mental health indicator [
17]. Future studies should test this assumption by analysing the long-term effects of engaged well-being on productivity and other indicators of organizational success.
We furthermore need to discuss the interpretation of the categories of engaged well-being in relation to one another. It can be assumed that it is the least desirable to have many employees in the category disengaged strain that has on average the lowest ratings of mental well-being or work engagement, while engaged well-being should be the most desirable category. Whether the category disengaged well-being or engaged strain is “preferable” cannot be clearly defined. An ordinal or metric interpretation is therefore not possible. However, because we assume that changes in different antecedents (i.e., work characteristics) have different consequences depending on the category of engaged well-being employees find themselves in, this distinction between disengaged well-being and engaged strain is necessary for employers to make informed decisions.
Strengths and limitations
A strength of the present study is its use of established and validated indicators of mental well-being and work engagement [
39,
41]. However, the generalization of our work using the UWES-9 for work engagement is somewhat limited due to the data including a shortened scale compared to that of the original work. Future work needs to test, whether similar findings can be found using the original scaling. The use of established clustering procedures that are accompanied by several sensitivity analyses (e.g., double-cross validation, within-sum-of-squares statistics) is another strength. Because our complete case analysis was based on only two indices, we were able to include 96.7% of all respondents in our cross-sectional analyses and we can assume that the selection bias due to missing data is rather small [
53]. The bias might be larger for the longitudinal analysis, as we face sample attrition (e.g., due to a healthy worker bias [
53]). Additionally, while on the employer level the LPP is representative for private, moderate- to large-sized German companies in the manufacturing and service sectors and employees from a wide variety of sectors and business sizes are included, the employee sample itself is primarily male, older, and working full-time, and results should therefore be interpreted carefully as they might not be representative for certain working populations. Future studies should therefore test our cut-offs using study populations with different sociodemographic characteristics. A bias due to common method variance cannot be excluded, as all items were measured subjectively and based on self-reports [
54]. Therefore, we propose to further test engaged well-being against objective indicators, such as biomarkers that are associated with stress or objective indicators for productivity.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.